Datasets:

Falah

/

research_paper2023

like 1

Fuzzy Attention Neural Network to Tackle Discontinuity in Airway Segmentation

Airway segmentation is crucial for the examination, diagnosis, and prognosis of lung diseases, while its manual delineation is unduly burdensome. To alleviate this time-consuming and potentially subjective manual procedure, researchers have proposed methods to automatically segment airways from computerized tomography (CT) images. However, some small-sized airway branches (e.g., bronchus and terminal bronchioles) significantly aggravate the difficulty of automatic segmentation by machine learning models. In particular, the variance of voxel values and the severe data imbalance in airway branches make the computational module prone to discontinuous and false-negative predictions. especially for cohorts with different lung diseases. Attention mechanism has shown the capacity to segment complex structures, while fuzzy logic can reduce the uncertainty in feature representations. Therefore, the integration of deep attention networks and fuzzy theory, given by the fuzzy attention layer, should be an escalated solution for better generalization and robustness. This paper presents an efficient method for airway segmentation, comprising a novel fuzzy attention neural network and a comprehensive loss function to enhance the spatial continuity of airway segmentation. The deep fuzzy set is formulated by a set of voxels in the feature map and a learnable Gaussian membership function. Different from the existing attention mechanism, the proposed channel-specific fuzzy attention addresses the issue of heterogeneous features in different channels. Furthermore, a novel evaluation metric is proposed to assess both the continuity and completeness of airway structures. The efficiency, generalization and robustness of the proposed method have been proved by training on normal lung disease while testing on datasets of lung cancer, COVID-19 and pulmonary fibrosis.

Microstructure estimation from diffusion-MRI: Compartmentalized models in permeable cellular tissue

Diffusion-weighted magnetic resonance imaging (DW-MRI) is used to characterize brain tissue microstructure employing tissue-specific biophysical models. A current limitation, however, is that most of the proposed models are based on the assumption of negligible water exchange between the intra- and extracellular compartments, which might not be valid in various brain tissues, including unmyelinated axons, gray matter, and tumors. The purpose of this work is to quantify the effect of membrane permeability on the estimates of two popular models neglecting exchange, and compare their performance with a model including exchange. To this aim, DW-MRI experiments were performed in controlled environments with Monte-Carlo simulations. The DW-MRI signals were generated in numerical substrates mimicking biological tissue made of spherical cells with permeable membranes like cancerous tissue or the brain gray matter. From these signals, the substrates properties were estimated using SANDI and VERDICT, the two compartment-based models neglecting exchange, and CEXI, a new model which includes exchange. Our results show that, in cellular permeable tissue, the model with exchange outperformed models without exchange in the estimation of the tissue properties by providing more stable estimates of cell size, intracellular volume fraction and extracellular diffusion coefficient. Moreover, the model with exchange estimated accurately the exchange time in the range of permeability reported for cellular tissue. Finally, the simulations performed in this work showed that the exchange between the intracellular and the extracellular space cannot be neglected in permeable tissue with a conventional PGSE sequence, to obtain accurate estimates. Consequently, existing compartmentalized models of impermeable tissue cannot be used for microstructure estimation of cellular permeable tissue.

Patient-specific mean teacher UNet for enhancing PET image and low-dose PET reconstruction on RefleXion X1 biology-guided radiotherapy system

The RefleXion X1 is the first biology-guided radiotherapy (BgRT) system. Its dual 90-degree PET detector collects fewer pair production events compared to a full-ring diagnostic PET system. In the proposed BgRT workflow, a short scan is acquired before treatment delivery to ensure image quality and consistency. The shorter scan time, a quarter of the simulation scan time, also leads to fewer coincidence events and hence reduced image quality. In this study, we proposed a patient-specific mean teacher UNet (MT-UNet) to enhance PET image quality and low-dose PET reconstruction on RefleXion X1. PET/CT scans of nine cancer patients were acquired using RefleXion X1. Every patient had one simulation scan. Five patients had additional scans acquired during the first and the final treatment fractions. Treatment scans were acquired using the same imaging protocol as the simulation scan. For each scan, we reconstructed a full-dose image and evenly split coincidence events into four sessions to reconstruct four quarter-dose PET images. For each patient, our proposed MT-UNet was trained using quarter-dose and full-dose images of the simulation scan. For the image quality enhancement task, we applied nine trained MT-UNets to full-dose simulation PET images of the nine patients to generate enhanced images, respectively. The enhanced images were compared with the original full-dose images using CNR and SNR. For the low-dose image reconstruction task, we applied five trained MT-UNets to ten quarter-dose treatment images of five patients to predict full-dose images, respectively. The predicted and ground truth full-dose images were compared using SSIM and PSNR. We also trained and evaluated patient-specific UNets for model comparison. Our proposed patient-specific MT-UNet achieved better performance in improving the quality of RefleXion low-dose and full-dose images compared to the patient-specific UNet.

Sub-second photon dose prediction via transformer neural networks

Fast dose calculation is critical for online and real time adaptive therapy workflows. While modern physics-based dose algorithms must compromise accuracy to achieve low computation times, deep learning models can potentially perform dose prediction tasks with both high fidelity and speed. We present a deep learning algorithm that, exploiting synergies between Transformer and convolutional layers, accurately predicts broad photon beam dose distributions in few milliseconds. The proposed improved Dose Transformer Algorithm (iDoTA) maps arbitrary patient geometries and beam information (in the form of a 3D projected shape resulting from a simple ray tracing calculation) to their corresponding 3D dose distribution. Treating the 3D CT input and dose output volumes as a sequence of 2D slices along the direction of the photon beam, iDoTA solves the dose prediction task as sequence modeling. The proposed model combines a Transformer backbone routing long-range information between all elements in the sequence, with a series of 3D convolutions extracting local features of the data. We train iDoTA on a dataset of 1700 beam dose distributions, using 11 clinical volumetric modulated arc therapy (VMAT) plans (from prostate, lung and head and neck cancer patients with 194-354 beams per plan) to assess its accuracy and speed. iDoTA predicts individual photon beams in ~50 milliseconds with a high gamma pass rate of 97.72% (2 mm, 2%). Furthermore, estimating full VMAT dose distributions in 6-12 seconds, iDoTA achieves state-of-the-art performance with a 99.51% (2 mm, 2%) pass rate. Offering the sub-second speed needed in online and real-time adaptive treatments, iDoTA represents a new state of the art in data-driven photon dose calculation. The proposed model can massively speed-up current photon workflows, reducing calculation times from few minutes to just a few seconds.

A probabilistic deep learning model of inter-fraction anatomical variations in radiotherapy

In radiotherapy, the internal movement of organs between treatment sessions causes errors in the final radiation dose delivery. Motion models can be used to simulate motion patterns and assess anatomical robustness before delivery. Traditionally, such models are based on principal component analysis (PCA) and are either patient-specific (requiring several scans per patient) or population-based, applying the same deformations to all patients. We present a hybrid approach which, based on population data, allows to predict patient-specific inter-fraction variations for an individual patient. We propose a deep learning probabilistic framework that generates deformation vector fields (DVFs) warping a patient's planning computed tomography (CT) into possible patient-specific anatomies. This daily anatomy model (DAM) uses few random variables capturing groups of correlated movements. Given a new planning CT, DAM estimates the joint distribution over the variables, with each sample from the distribution corresponding to a different deformation. We train our model using dataset of 312 CT pairs from 38 prostate cancer patients. For 2 additional patients (22 CTs), we compute the contour overlap between real and generated images, and compare the sampled and ground truth distributions of volume and center of mass changes. With a DICE score of 0.86 and a distance between prostate contours of 1.09 mm, DAM matches and improves upon PCA-based models. The distribution overlap further indicates that DAM's sampled movements match the range and frequency of clinically observed daily changes on repeat CTs. Conditioned only on a planning CT and contours of a new patient without any pre-processing, DAM can accurately predict CTs seen during following treatment sessions, which can be used for anatomically robust treatment planning and robustness evaluation against inter-fraction anatomical changes.

Application of the nnU-Net for automatic segmentation of lung lesion on CT images, and implication on radiomic models

Lesion segmentation is a crucial step of the radiomic workflow. Manual segmentation requires long execution time and is prone to variability, impairing the realisation of radiomic studies and their robustness. In this study, a deep-learning automatic segmentation method was applied on computed tomography images of non-small-cell lung cancer patients. The use of manual vs automatic segmentation in the performance of survival radiomic models was assessed, as well. METHODS A total of 899 NSCLC patients were included (2 proprietary: A and B, 1 public datasets: C). Automatic segmentation of lung lesions was performed by training a previously developed architecture, the nnU-Net, including 2D, 3D and cascade approaches. The quality of automatic segmentation was evaluated with DICE coefficient, considering manual contours as reference. The impact of automatic segmentation on the performance of a radiomic model for patient survival was explored by extracting radiomic hand-crafted and deep-learning features from manual and automatic contours of dataset A, and feeding different machine learning algorithms to classify survival above/below median. Models' accuracies were assessed and compared. RESULTS The best agreement between automatic and manual contours with DICE=0.78 +(0.12) was achieved by averaging predictions from 2D and 3D models, and applying a post-processing technique to extract the maximum connected component. No statistical differences were observed in the performances of survival models when using manual or automatic contours, hand-crafted, or deep features. The best classifier showed an accuracy between 0.65 and 0.78. CONCLUSION The promising role of nnU-Net for automatic segmentation of lung lesions was confirmed, dramatically reducing the time-consuming physicians' workload without impairing the accuracy of survival predictive models based on radiomics.

$^{18}$F-PSMA-1007 salivary gland dosimetry: Comparison between different methods for dose calculation and assessment of inter- and intra-patient variability

Dosimetry of salivary glands (SGs) is usually implemented using simplified calculation approaches and approximated geometries. Our aims were to compare different dosimetry methods to calculate SGs absorbed doses (ADs) following 18F-PSMA-1007 injection, and to assess the AD variation across patients and single SG components. Five patients with prostate cancer recurrence underwent PET/CT acquisitions of the head and neck, 0.5, 2 and 4 hours after 18F-PSMA-1007 injection. Parotid and submandibular glands were segmented on CT to derive SGs volumes and masses, while PETs were used to derive Time-Integrated Activity Coefficients. Average ADs to single SG components or total SG (tSG) were calculated with the following methods: i) direct Monte Carlo (MC) simulation with GATE/GEANT4; ii) spherical model (SM) of OLINDA/EXM 2.1, adopting either patient-specific or standard ICRP89 organ masses (SMstd); iii) ellipsoidal model (EM); iv) MIRD approach with organ S-factors from OLINDA/EXM 2.1 and OpenDose collaboration, with or without contribution from cross irradiation originating outside the SGs. The maximum percent AD difference across SG components ({\delta}max) and across patients ({\Delta}max) were calculated. Compared to MC, ADs to single SG components were significantly underestimated by all methods (average relative differences between -14.5% and -30.4%). Using MC, SM and EM, {\delta}max were never below 25% (up to 113%). {\delta}max up to 702% were obtained with SMstd. Concerning tSG, results within 10% of the MC were obtained only if cross irradiation from the remainder of the body or from the remainder of the head was accounted for. The {\Delta}max ranged between 58% and 78% across patients. Specific masses of single SG components should always be considered given their large intra- and inter- patient variability.

Dosimetric Evaluation of a New Rotating Gamma System for Stereotactic Radiosurgery

Purpose: A novel rotating gamma stereotactic radiosurgery (SRS) system (Galaxy RTi) with real-time image guidance technology has been developed for high-precision SRS and frameless fractionated stereotactic radiotherapy (SRT). This work investigated the dosimetric quality of Galaxy by comparing both the machine treatment parameters and plan dosimetry parameters with those of the widely used Leksell Gamma Knife (LGK) systems for SRS. Methods: The Galaxy RTi system uses 30 cobalt-60 sources on a rotating gantry to deliver non-coplanar, non-overlapping arcs simultaneously while the LGK 4C uses 201 static cobalt-60 sources to deliver noncoplanar beams. Ten brain cancer patients were unarchived from our clinical database, which were previously treated on the LGK 4C. The lesion volume for these cases varied from 0.1 cm3 to 15.4 cm3. Galaxy plans were generated using the Prowess TPS (Prowess, Concord, CA) with the same dose constraints and optimization parameters. Treatment quality metrics such as target coverage (%volume receiving the prescription dose), conformity index (CI), cone size, shots number, beam-on time were compared together with DVH curves and dose distributions. Results: Superior treatment plans were generated for the Galaxy system that met our clinical acceptance criteria. For the 10 patients investigated, the mean CI and dose coverage for Galaxy was 1.77 and 99.24 compared to 1.94 and 99.19 for LGK, respectively. The beam-on time for Galaxy was 17.42 minutes compared to 21.34 minutes for LGK (both assuming dose rates at the initial installation). The dose fall-off is much faster for Galaxy, compared with LGK. Conclusion: The Galaxy RTi system can provide dose distributions with similar quality to that of LGK with less beam-on time and faster dose fall-off. The system is also capable of real-time image guidance at treatment position to ensure accurate dose delivery for SRS.

Range margin reduction in carbon ion therapy: potential benefits of using radioactive ion beams

Radiotherapy with heavy ions, in particular, 12C beams, is one of the most advanced forms of cancer treatment. Sharp dose gradients and high biological effectiveness in the target region make them an ideal tool to treat deep-seated and radioresistant tumors, however, at the same time, sensitive to small errors in the range prediction. Safety margins are added to the tumor volume to mitigate these uncertainties and ensure its uniform coverage, but during the irradiation they lead to unavoidable damage to the surrounding healthy tissue. To fully exploit the benefits of a sharp Bragg peak, a large effort is put into establishing precise range verification methods for the so-called image-guided radiotherapy. Despite positron emission tomography being widely in use for this purpose in 12C ion therapy, the low count rates, biological washout, and broad shape of the activity distribution still limit its precision to a few millimeters. Instead, radioactive beams used directly for treatment would yield an improved signal and a closer match with the dose fall-off, potentially enabling precise in vivo beam range monitoring. We have performed a treatment planning study to estimate the possible impact of the reduced range uncertainties, enabled by radioactive 11C beams treatments, on sparing critical organs in the tumor proximity. We demonstrate that (i) annihilation maps for 11C ions can in principle reflect even millimeter shifts in dose distributions in the patient, (ii) outcomes of treatment planning with 11C beams are significantly improved in terms of meeting the constraints for the organs at risk compared to 12C plans, and (iii) less severe toxicities for serial and parallel critical organs can be expected following 11C treatment with reduced range uncertainties, compared to 12C treatments.

Another view of sequential sampling in the birth process with immigration

Models of counts-of-counts data have been extensively used in the biological sciences, for example in cancer, population genetics, sampling theory and ecology. In this paper we explore properties of one model that is embedded into a continuous-time process and can describe the appearance of certain biological data such as covid DNA sequences in a database. More specifically, we consider an evolving model of counts-of-counts data that arises as the family size counts of samples taken sequentially from a Birth process with Immigration (BI). Here, each family represents a type or species, and the family size counts represent the type or species frequency spectrum in the population. We study the correlation of $S(a,b)$ and $S(c,d)$, the number of families observed in two disjoint time intervals $(a,b)$ and $(c,d)$. We find the expected sample variance and its asymptotics for $p$ consecutive sequential samples $\mathbf{S}_p:=(S(t_0,t_1),\dots, S(t_{p-1},t_p))$, for any given $0=t_0<t_1<\dots<t_p$. By conditioning on the sizes of the samples, we provide a connection between $\mathbf{S}_p$ and $p$ sequential samples of sizes $n_1,n_2,\dots,n_p$, drawn from a single run of a Chinese Restaurant Process. The properties of the latter were studied in da Silva et al. (2022). We show how the continuous-time framework helps to make asymptotic calculations easier than its discrete-time counterpart. As an application, for a specific choice of $t_1,t_2,\dots, t_p$, we revisit Fisher's 1943 multi-sampling problem and give another explanation of what Fisher's model could have meant in the world of sequential samples drawn from a BI process.

Graph Attention Networks Unveil Determinants of Intra- and Inter-city Health Disparity

Understanding the determinants underlying variations in urban health status is important for informing urban design and planning, as well as public health policies. Multiple heterogeneous urban features could modulate the prevalence of diseases across different neighborhoods in cities and across different cities. This study examines heterogeneous features related to socio-demographics, population activity, mobility, and the built environment and their non-linear interactions to examine intra- and inter-city disparity in prevalence of four disease types: obesity, diabetes, cancer, and heart disease. Features related to population activity, mobility, and facility density are obtained from large-scale anonymized mobility data. These features are used in training and testing graph attention network (GAT) models to capture non-linear feature interactions as well as spatial interdependence among neighborhoods. We tested the models in five U.S. cities across the four disease types. The results show that the GAT model can predict the health status of people in neighborhoods based on the top five determinant features. The findings unveil that population activity and built-environment features along with socio-demographic features differentiate the health status of neighborhoods to such a great extent that a GAT model could predict the health status using these features with high accuracy. The results also show that the model trained on one city can predict health status in another city with high accuracy, allowing us to quantify the inter-city similarity and discrepancy in health status. The model and findings provide novel approaches and insights for urban designers, planners, and public health officials to better understand and improve health disparities in cities by considering the significant determinant features and their interactions.

Discriminating between individual-based models of collective cell motion in a benchmark flow geometry using standardised spatiotemporal patterns

Collectively coordinated cell migration plays a role in tissue embryogenesis, cancer, homeostasis and healing. To study these processes, different cell-based modelling approaches have been developed, ranging from lattice-based cellular automata to lattice-free models that treat cells as point-like particles or extended detailed cell shape contours. In the spirit of what Osborne et al. [PLOS Computational Biology, (2017) 13, 1-34] did for cellular tissue structure simulation models, we here compare five simulation models of collective cell migration, chosen to be representative in increasing order of included detail. They are Vicsek-Gr\'egoire particles, Szab\'o-like particles, self-propelled Voronoi model, cellular Potts model, and multiparticle cells, where each model includes cell motility. We examine how these models compare when applied to the same biological problem, and what differences in behaviour are due to different model assumptions and abstractions. For that purpose, we use a benchmark that discriminates between complex material flow models, and that can be experimentally approached using cell cultures: the flow within a channel around a circular obstacle, that is, the geometry Stokes used in his historical 1851 experiment. For each model we explain how to best implement it; vary cell density, attraction force and alignment interaction; draw the resulting maps of velocity, density and deformation fields; and eventually discuss its respective advantages and limitations. We thus provide a recommendation on how to select a model to answer a given question, and we examine whether models of motile particles and motile cells display similar collective effects.

Three-component contour dynamics model to simulate and analyze amoeboid cell motility

Amoeboid cell motility is relevant in a wide variety of biomedical applications such as wound healing, cancer metastasis, and embryonic morphogenesis. It is characterized by pronounced changes of the cell shape associated with expansions and retractions of the cell membrane, which result in a crawling kind of locomotion. Despite existing computational models of amoeboid motion, the inference of expansion and retraction components of individual cells, the corresponding classification of cells, and the a priori specification of the parameter regime to achieve a specific motility behavior remain challenging open problems. We propose a novel model of the spatio-temporal evolution of two-dimensional cell contours comprising three biophysiologically motivated components: a stochastic term accounting for membrane protrusions and two deterministic terms accounting for membrane retractions by regularizing the shape and area of the contour. Mathematically, these correspond to the intensity of a self-exciting Poisson point process, the area-preserving curve-shortening flow, and an area adjustment flow. The model is used to generate contour data for a variety of qualitatively different, e.g., polarized and non-polarized, cell tracks that are hardly distinguishable from experimental data. In application to experimental cell tracks, we inferred the protrusion component and examined its correlation to commonly used biomarkers: the actin concentration close to the membrane and its local motion. Due to the low model complexity, parameter estimation is fast, straightforward and offers a simple way to classify contour dynamics based on two locomotion types: the amoeboid and a so-called fan-shaped type. For both types, we use cell tracks segmented from fluorescence imaging data of the model organism D. discoideum. An implementation of the model is provided within the open-source software package AmoePy.

Extracting lung function-correlated information from CT-encoded static textures

The inherent characteristics of lung tissues, which are independent of breathing manoeuvre, may provide fundamental information on lung function. This paper attempted to study function-correlated lung textures and their spatial distribution from CT. 21 lung cancer patients with thoracic 4DCT scans, DTPA-SPECT ventilation images (V), and available pulmonary function test (PFT) measurements were collected. 79 radiomic features were included for analysis, and a sparse-to-fine strategy including subregional feature discovery and voxel-wise feature distribution study was carried out to identify the function-correlated radiomic features. At the subregion level, lung CT images were partitioned and labeled as defected/non-defected patches according to reference V. At the voxel-wise level, feature maps (FMs) of selected feature candidates were generated for each 4DCT phase. Quantitative metrics, including Spearman coefficient of correlation (SCC) and Dice similarity coefficient (DSC) for FM-V spatial agreement assessments, intra-class coefficient of correlation (ICC) for FM robustness evaluations, and FM-PFT comparisons, were applied to validate the results. At the subregion level, eight function-correlated features were filtered out with medium-to-large statistical strength (effect size>0.330) to differentiate defected/non-defected lung regions. At the voxel-wise level, FMs of candidates yielded moderate-to-strong voxel-wise correlations with reference V. Among them, FMs of GLDM Dependence Non-uniformity showed the highest robust (ICC=0.96) spatial correlation, with median SCCs ranging from 0.54 to 0.59 throughout ten phases. Its phase-averaged FM achieved a median SCC of 0.60, the median DSC of 0.60/0.65 for high/low functional lung volumes, respectively, and the correlation of 0.646 between the spatially averaged feature values and PFT measurements.

Deep Learning-based Protoacoustic Signal Denoising for Proton Range Verification

Objective: Proton therapy offers an advantageous dose distribution compared to the photon therapy, since it deposits most of the energy at the end of range, namely the Bragg peak (BP). Protoacoustic technique was developed to in vivo determine the BP locations. However, it requires large dose delivery to the tissue to obtain an averaged acoustic signal with a sufficient signal to noise ratio (SNR), which is not suitable in clinics. We propose a deep learning-based technique to acquire denoised acoustic signals and reduce BP range uncertainty with much lower doses. Approach: Three accelerometers were placed on the distal surface of a cylindrical polyethylene (PE) phantom to collect protoacoustic signals. In total 512 raw signals were collected at each device. Device-specific stack autoencoder (SAE) denoising models were trained to denoise the input signals, which were generated by averaging 1, 2, 4, 8, 16, or 32 raw signals. Both supervised and unsupervised learning training strategies were tested for comparison. Mean squared error (MSE), signal-to-noise ratio (SNR) and the Bragg peak (BP) range uncertainty were used for model evaluation. Main results: After SAE denoising, the MSE was substantially reduced, and the SNR was enhanced. Overall, the supervised SAEs outperformed the unsupervised SAEs in BP range verification. For the high accuracy detector, it achieved a BP range uncertainty of 0.20 +/- 3.44 mm by averaging over 8 raw signals, while for the other two low accuracy detectors, they achieved the BP uncertainty of 1.44 +/- 6.45 mm and -0.23 +/- 4.88 mm by averaging 16 raw signals, respectively. Significance: We have proposed a deep learning based denoising method to enhance the SNR of protoacoustic measurements and improve the accuracy in BP range verification, which greatly reduces the dose and time for potential clinical applications.

TrustGAN: Training safe and trustworthy deep learning models through generative adversarial networks

Deep learning models have been developed for a variety of tasks and are deployed every day to work in real conditions. Some of these tasks are critical and models need to be trusted and safe, e.g. military communications or cancer diagnosis. These models are given real data, simulated data or combination of both and are trained to be highly predictive on them. However, gathering enough real data or simulating them to be representative of all the real conditions is: costly, sometimes impossible due to confidentiality and most of the time impossible. Indeed, real conditions are constantly changing and sometimes are intractable. A solution is to deploy machine learning models that are able to give predictions when they are confident enough otherwise raise a flag or abstain. One issue is that standard models easily fail at detecting out-of-distribution samples where their predictions are unreliable. We present here TrustGAN, a generative adversarial network pipeline targeting trustness. It is a deep learning pipeline which improves a target model estimation of the confidence without impacting its predictive power. The pipeline can accept any given deep learning model which outputs a prediction and a confidence on this prediction. Moreover, the pipeline does not need to modify this target model. It can thus be easily deployed in a MLOps (Machine Learning Operations) setting. The pipeline is applied here to a target classification model trained on MNIST data to recognise numbers based on images. We compare such a model when trained in the standard way and with TrustGAN. We show that on out-of-distribution samples, here FashionMNIST and CIFAR10, the estimated confidence is largely reduced. We observe similar conclusions for a classification model trained on 1D radio signals from AugMod, tested on RML2016.04C. We also publicly release the code.

Automated Deep Aberration Detection from Chromosome Karyotype Images

Chromosome analysis is essential for diagnosing genetic disorders. For hematologic malignancies, identification of somatic clonal aberrations by karyotype analysis remains the standard of care. However, karyotyping is costly and time-consuming because of the largely manual process and the expertise required in identifying and annotating aberrations. Efforts to automate karyotype analysis to date fell short in aberration detection. Using a training set of ~10k patient specimens and ~50k karyograms from over 5 years from the Fred Hutchinson Cancer Center, we created a labeled set of images representing individual chromosomes. These individual chromosomes were used to train and assess deep learning models for classifying the 24 human chromosomes and identifying chromosomal aberrations. The top-accuracy models utilized the recently introduced Topological Vision Transformers (TopViTs) with 2-level-block-Toeplitz masking, to incorporate structural inductive bias. TopViT outperformed CNN (Inception) models with >99.3% accuracy for chromosome identification, and exhibited accuracies >99% for aberration detection in most aberrations. Notably, we were able to show high-quality performance even in "few shot" learning scenarios. Incorporating the definition of clonality substantially improved both precision and recall (sensitivity). When applied to "zero shot" scenarios, the model captured aberrations without training, with perfect precision at >50% recall. Together these results show that modern deep learning models can approach expert-level performance for chromosome aberration detection. To our knowledge, this is the first study demonstrating the downstream effectiveness of TopViTs. These results open up exciting opportunities for not only expediting patient results but providing a scalable technology for early screening of low-abundance chromosomal lesions.

Deep neuroevolution for limited, heterogeneous data: proof-of-concept application to Neuroblastoma brain metastasis using a small virtual pooled image collection

Artificial intelligence (AI) in radiology has made great strides in recent years, but many hurdles remain. Overfitting and lack of generalizability represent important ongoing challenges hindering accurate and dependable clinical deployment. If AI algorithms can avoid overfitting and achieve true generalizability, they can go from the research realm to the forefront of clinical work. Recently, small data AI approaches such as deep neuroevolution (DNE) have avoided overfitting small training sets. We seek to address both overfitting and generalizability by applying DNE to a virtually pooled data set consisting of images from various institutions. Our use case is classifying neuroblastoma brain metastases on MRI. Neuroblastoma is well-suited for our goals because it is a rare cancer. Hence, studying this pediatric disease requires a small data approach. As a tertiary care center, the neuroblastoma images in our local Picture Archiving and Communication System (PACS) are largely from outside institutions. These multi-institutional images provide a heterogeneous data set that can simulate real world clinical deployment. As in prior DNE work, we used a small training set, consisting of 30 normal and 30 metastasis-containing post-contrast MRI brain scans, with 37% outside images. The testing set was enriched with 83% outside images. DNE converged to a testing set accuracy of 97%. Hence, the algorithm was able to predict image class with near-perfect accuracy on a testing set that simulates real-world data. Hence, the work described here represents a considerable contribution toward clinically feasible AI.

Realistic 3D printed imaging tumor phantoms for validation of image processing algorithms

Medical imaging phantoms are widely used for validation and verification of imaging systems and algorithms in surgical guidance and radiation oncology procedures. Especially, for the performance evaluation of new algorithms in the field of medical imaging, manufactured phantoms need to replicate specific properties of the human body, e.g., tissue morphology and radiological properties. Additive manufacturing (AM) technology provides an inexpensive opportunity for accurate anatomical replication with customization capabilities. In this study, we proposed a simple and cheap protocol to manufacture realistic tumor phantoms based on the filament 3D printing technology. Tumor phantoms with both homogenous and heterogenous radiodensity were fabricated. The radiodensity similarity between the printed tumor models and real tumor data from CT images of lung cancer patients was evaluated. Additionally, it was investigated whether a heterogeneity in the 3D printed tumor phantoms as observed in the tumor patient data had an influence on the validation of image registration algorithms. A density range between -217 to 226 HUs was achieved for 3D printed phantoms; this range of radiation attenuation is also observed in the human lung tumor tissue. The resulted HU range could serve as a lookup-table for researchers and phantom manufactures to create realistic CT tumor phantoms with the desired range of radiodensities. The 3D printed tumor phantoms also precisely replicated real lung tumor patient data regarding morphology and could also include life-like heterogeneity of the radiodensity inside the tumor models. An influence of the heterogeneity on accuracy and robustness of the image registration algorithms was not found.

Towards Transcervical Ultrasound Image Guidance for Transoral Robotic Surgery

Purpose: Trans-oral robotic surgery (TORS) using the da Vinci surgical robot is a new minimally-invasive surgery method to treat oropharyngeal tumors, but it is a challenging operation. Augmented reality (AR) based on intra-operative ultrasound (US) has the potential to enhance the visualization of the anatomy and cancerous tumors to provide additional tools for decision-making in surgery. Methods: We propose and carry out preliminary evaluations of a US-guided AR system for TORS, with the transducer placed on the neck for a transcervical view. Firstly, we perform a novel MRI-transcervical 3D US registration study. Secondly, we develop a US-robot calibration method with an optical tracker and an AR system to display the anatomy mesh model in the real-time endoscope images inside the surgeon console. Results: Our AR system reaches a mean projection error of 26.81 and 27.85 pixels for the projection from the US to stereo cameras in a water bath experiment. The average target registration error for MRI to 3D US is 8.90 mm for the 3D US transducer and 5.85 mm for freehand 3D US, and the average distance between the vessel centerlines is 2.32 mm. Conclusion: We demonstrate the first proof-of-concept transcervical US-guided AR system for TORS and the feasibility of trans-cervical 3D US-MRI registration. Our results show that trans-cervical 3D US is a promising technique for TORS image guidance.

Fluorescent property of carbon dots extracted from cigarette smoke and the application in bio-imaging

Cigarette smoke is one of the six major pollution sources in the room air. It contains large number of particles with size less than 10 nm. There exist carbon dots (CDs) in cigarette smoke which have strong fluorescence and with good bio-compatibility and low toxicity. CDs in cigarette smoke can be applied in bio-imaging which has great potential applications in the integration of cancer diagnosis and treatment. In this paper, CDs were extracted from cigarette smoke. Then, sodium borohydride was added to CDs aqueous solution for reduction and the reduced CDs (R-CDs) were used for biological cell imaging. The results indicate that the CDs with the particle size $<$10 nm in cigarette smoke are self-assembled by the polymerizated polycyclic aromatic hydrocarbons (PAHs) and ammonium nitrite which are disk nano-structure composed of $sp^2$/$sp^3$ carbon and oxygen/nitrogen groups or polymers. Sodium borohydride can reduce the carbonyl group on the surface of CDs to hydroxyl group and increase the ratio of the Na 1s ratio of the CDs from 1.86 to 7.42. The CDs can emit blue fluorescence under ultraviolet irradiation. After reduction, the R-CDS have the intensity of fluorescence 7.2 times than before and the fluorescence quantum yield increase from 6.13\% to 8.86\%. The photoluminescence (PL) wavelength of R-CDS have red-shift of 7 nm which was due to the increasing of Na element ratio. The onion epidermal cells labeled with R-CDs show that the CDs could pass through the cell wall into the cell and reach the nucleus. The cell wall and the nucleus could be clearly visualized. CDs also shows low toxicity to human bronchial epithelial cells (BEAS-2B) with good biological activity. The obtained results indicate that the CDs and R-CDs have good fluorescent property which could be used as bio-imaging agent.

Nearby voids and their galaxies: recent progress and prospects

Voids occupy about 3/4 of the volume of the Universe and contain about 15% of its mass. Due to various observational selection effects, these structure elements and galaxies populating voids, are highly under-explored. This especially relates to the lowest mass galaxies which comprise the main void population. Studying the nearby voids allows us to improve our understanding of the most elusive void objects. We present the brief overview of the current status and the prospects of the study of the nearest voids and their galaxies. First, we summarize the pioneer study of a hundred galaxies residing in the nearby Lynx-Cancer void which clearly evidence for the slower evolution of void galaxies and finds also the unusual very metal-poor and gas-rich dwarfs. Then we describe the recently defined sample of the nearby voids within the sphere with R = 25 Mpc and a sample of 1350 galaxies residing in these voids (~20% of all galaxies within this volume). We discuss the current results obtained for several directions of the study of this sample. They include: the search for Very Young Galaxies, the study of HI properties, the clustering of void galaxies and its relation to the void substructures, and the unbiased study of 260 void galaxies within the Local Volume (R < 11 Mpc). Altogether, this opens a perspective way to address the suggested peculiarities of the void galaxy formation and evolution. Finally, we briefly overview the expected advancements in the void galaxy studies related to the upcoming new facilities.

Explainable AI for Bioinformatics: Methods, Tools, and Applications

Artificial intelligence (AI) systems utilizing deep neural networks (DNNs) and machine learning (ML) algorithms are widely used for solving important problems in bioinformatics, biomedical informatics, and precision medicine. However, complex DNNs or ML models, which are often perceived as opaque and black-box, can make it difficult to understand the reasoning behind their decisions. This lack of transparency can be a challenge for both end-users and decision-makers, as well as AI developers. Additionally, in sensitive areas like healthcare, explainability and accountability are not only desirable but also legally required for AI systems that can have a significant impact on human lives. Fairness is another growing concern, as algorithmic decisions should not show bias or discrimination towards certain groups or individuals based on sensitive attributes. Explainable artificial intelligence (XAI) aims to overcome the opaqueness of black-box models and provide transparency in how AI systems make decisions. Interpretable ML models can explain how they make predictions and the factors that influence their outcomes. However, most state-of-the-art interpretable ML methods are domain-agnostic and evolved from fields like computer vision, automated reasoning, or statistics, making direct application to bioinformatics problems challenging without customization and domain-specific adaptation. In this paper, we discuss the importance of explainability in the context of bioinformatics, provide an overview of model-specific and model-agnostic interpretable ML methods and tools, and outline their potential caveats and drawbacks. Besides, we discuss how to customize existing interpretable ML methods for bioinformatics problems. Nevertheless, we demonstrate how XAI methods can improve transparency through case studies in bioimaging, cancer genomics, and text mining.

Protein Co-Enrichment Analysis of Extracellular Vesicles

Extracellular Vesicles (EVs) carry cell-derived proteins that confer functionality and selective cell uptake. However, whether proteins are packaged stochastically or co-enriched within individual EVs, and whether co-enrichment fluctuates under homeostasis and disease, has not been measured. EV abundance and protein global relative expression have been qualified by bulk analysis. Meanwhile, co-enrichment is not directly accessible via bulk measurement and has not been reported for single EV analysis. Here, we introduce the normalized index of co-enrichment (NICE) to measure protein co-enrichment. NICE was derived by (i) capturing EVs based on the expression of a membrane-bound protein, (ii) probing for the co-expression of a second protein at the population level - EV integrity underwrites the detection of single EV co-expression without the need to resolve single EVs - and (iii) normalizing measured values using two universal normalization probes. Axiomatically, NICE = 1 for stochastic inclusion or no overall co-enrichment, while for positive and negative co-enrichment NICE > 1 or < 1, respectively. We quantified the NICE of tetraspanins, growth factor receptors and integrins in EVs of eight breast cancer cell lines of varying metastatic potential and organotropism, combinatorially mapping up to 104 protein pairs. Our analysis revealed protein enrichment and co-expression patterns consistent with previous findings. For the organotropic cell lines, most protein pairs were co-enriched on EVs, with the majority of NICE values between 0.2 to 11.5, and extending from 0.037 to 80.4. Median NICE were either negative, neutral or positive depending on the cells. NICE analysis is easily multiplexed and is compatible with microarrays, bead-based and single EV assays. Additional studies are needed to deepen our understanding of the potential and significance of NICE for research and clinical uses.

Multi-domain stain normalization for digital pathology: A cycle-consistent adversarial network for whole slide images

The variation in histologic staining between different medical centers is one of the most profound challenges in the field of computer-aided diagnosis. The appearance disparity of pathological whole slide images causes algorithms to become less reliable, which in turn impedes the wide-spread applicability of downstream tasks like cancer diagnosis. Furthermore, different stainings lead to biases in the training which in case of domain shifts negatively affect the test performance. Therefore, in this paper we propose MultiStain-CycleGAN, a multi-domain approach to stain normalization based on CycleGAN. Our modifications to CycleGAN allow us to normalize images of different origins without retraining or using different models. We perform an extensive evaluation of our method using various metrics and compare it to commonly used methods that are multi-domain capable. First, we evaluate how well our method fools a domain classifier that tries to assign a medical center to an image. Then, we test our normalization on the tumor classification performance of a downstream classifier. Furthermore, we evaluate the image quality of the normalized images using the Structural similarity index and the ability to reduce the domain shift using the Fr\'echet inception distance. We show that our method proves to be multi-domain capable, provides the highest image quality among the compared methods, and can most reliably fool the domain classifier while keeping the tumor classifier performance high. By reducing the domain influence, biases in the data can be removed on the one hand and the origin of the whole slide image can be disguised on the other, thus enhancing patient data privacy.

On the Feasibility of Machine Learning Augmented Magnetic Resonance for Point-of-Care Identification of Disease

Early detection of many life-threatening diseases (e.g., prostate and breast cancer) within at-risk population can improve clinical outcomes and reduce cost of care. While numerous disease-specific "screening" tests that are closer to Point-of-Care (POC) are in use for this task, their low specificity results in unnecessary biopsies, leading to avoidable patient trauma and wasteful healthcare spending. On the other hand, despite the high accuracy of Magnetic Resonance (MR) imaging in disease diagnosis, it is not used as a POC disease identification tool because of poor accessibility. The root cause of poor accessibility of MR stems from the requirement to reconstruct high-fidelity images, as it necessitates a lengthy and complex process of acquiring large quantities of high-quality k-space measurements. In this study we explore the feasibility of an ML-augmented MR pipeline that directly infers the disease sidestepping the image reconstruction process. We hypothesise that the disease classification task can be solved using a very small tailored subset of k-space data, compared to image reconstruction. Towards that end, we propose a method that performs two tasks: 1) identifies a subset of the k-space that maximizes disease identification accuracy, and 2) infers the disease directly using the identified k-space subset, bypassing the image reconstruction step. We validate our hypothesis by measuring the performance of the proposed system across multiple diseases and anatomies. We show that comparable performance to image-based classifiers, trained on images reconstructed with full k-space data, can be achieved using small quantities of data: 8% of the data for detecting multiple abnormalities in prostate and brain scans, and 5% of the data for knee abnormalities. To better understand the proposed approach and instigate future research, we provide an extensive analysis and release code.

BRAIxDet: Learning to Detect Malignant Breast Lesion with Incomplete Annotations

Methods to detect malignant lesions from screening mammograms are usually trained with fully annotated datasets, where images are labelled with the localisation and classification of cancerous lesions. However, real-world screening mammogram datasets commonly have a subset that is fully annotated and another subset that is weakly annotated with just the global classification (i.e., without lesion localisation). Given the large size of such datasets, researchers usually face a dilemma with the weakly annotated subset: to not use it or to fully annotate it. The first option will reduce detection accuracy because it does not use the whole dataset, and the second option is too expensive given that the annotation needs to be done by expert radiologists. In this paper, we propose a middle-ground solution for the dilemma, which is to formulate the training as a weakly- and semi-supervised learning problem that we refer to as malignant breast lesion detection with incomplete annotations. To address this problem, our new method comprises two stages, namely: 1) pre-training a multi-view mammogram classifier with weak supervision from the whole dataset, and 2) extending the trained classifier to become a multi-view detector that is trained with semi-supervised student-teacher learning, where the training set contains fully and weakly-annotated mammograms. We provide extensive detection results on two real-world screening mammogram datasets containing incomplete annotations, and show that our proposed approach achieves state-of-the-art results in the detection of malignant breast lesions with incomplete annotations.

IMPORTANT-Net: Integrated MRI Multi-Parameter Reinforcement Fusion Generator with Attention Network for Synthesizing Absent Data

Magnetic resonance imaging (MRI) is highly sensitive for lesion detection in the breasts. Sequences obtained with different settings can capture the specific characteristics of lesions. Such multi-parameter MRI information has been shown to improve radiologist performance in lesion classification, as well as improving the performance of artificial intelligence models in various tasks. However, obtaining multi-parameter MRI makes the examination costly in both financial and time perspectives, and there may be safety concerns for special populations, thus making acquisition of the full spectrum of MRI sequences less durable. In this study, different than naive input fusion or feature concatenation from existing MRI parameters, a novel $\textbf{I}$ntegrated MRI $\textbf{M}$ulti-$\textbf{P}$arameter reinf$\textbf{O}$rcement fusion generato$\textbf{R}$ wi$\textbf{T}$h $\textbf{A}$tte$\textbf{NT}$ion Network (IMPORTANT-Net) is developed to generate missing parameters. First, the parameter reconstruction module is used to encode and restore the existing MRI parameters to obtain the corresponding latent representation information at any scale level. Then the multi-parameter fusion with attention module enables the interaction of the encoded information from different parameters through a set of algorithmic strategies, and applies different weights to the information through the attention mechanism after information fusion to obtain refined representation information. Finally, a reinforcement fusion scheme embedded in a $V^{-}$-shape generation module is used to combine the hierarchical representations to generate the missing MRI parameter. Results showed that our IMPORTANT-Net is capable of generating missing MRI parameters and outperforms comparable state-of-the-art networks. Our code is available at https://github.com/Netherlands-Cancer-Institute/MRI_IMPORTANT_NET.

Probabilistic Attention based on Gaussian Processes for Deep Multiple Instance Learning

Multiple Instance Learning (MIL) is a weakly supervised learning paradigm that is becoming increasingly popular because it requires less labeling effort than fully supervised methods. This is especially interesting for areas where the creation of large annotated datasets remains challenging, as in medicine. Although recent deep learning MIL approaches have obtained state-of-the-art results, they are fully deterministic and do not provide uncertainty estimations for the predictions. In this work, we introduce the Attention Gaussian Process (AGP) model, a novel probabilistic attention mechanism based on Gaussian Processes for deep MIL. AGP provides accurate bag-level predictions as well as instance-level explainability, and can be trained end-to-end. Moreover, its probabilistic nature guarantees robustness to overfitting on small datasets and uncertainty estimations for the predictions. The latter is especially important in medical applications, where decisions have a direct impact on the patient's health. The proposed model is validated experimentally as follows. First, its behavior is illustrated in two synthetic MIL experiments based on the well-known MNIST and CIFAR-10 datasets, respectively. Then, it is evaluated in three different real-world cancer detection experiments. AGP outperforms state-of-the-art MIL approaches, including deterministic deep learning ones. It shows a strong performance even on a small dataset with less than 100 labels and generalizes better than competing methods on an external test set. Moreover, we experimentally show that predictive uncertainty correlates with the risk of wrong predictions, and therefore it is a good indicator of reliability in practice. Our code is publicly available.

A network-based biomarkers discovery of Cold/Hot ZHENG chronic gastritis and Cold/Hot herbs of formulae

Objective: To discover biomarkers and uncover the mechanism of Cold/Hot ZHENG (syndrome in traditional Chinese medicine) chronic gastritis (CG) and Cold/Hot herbs in traditional Chinese medicine (TCM) formulae on systematic biology. Background: CG is a common inflammatory disease and the diagnosis of CG in TCM can be classified into Cold ZHENG (Asthenic Cold) and Hot ZHENG (Excess Hot). However, the molecular features of Cold/Hot ZHENG in CG and the mechanism of Cold/Hot herbs in formulae for CG remained unclear. Methods: Based on data of 35 patients of Cold/Hot ZHENG CG and 3 scRNA-seq CG samples, we conduct analysis with transcriptomics datasets and algorithms, to discover biomarkers for Cold/Hot ZHENG CG. And we collected 25 formulae (with traditional effects related to Cold/Hot ZHENG) for CG and corresponding 89 Cold/Hot herbs (including Warm/Cool herbs) to discover features and construct target networks of Cold/Hot herbs on the basis of network target and enrichment analysis. Results: Biomarkers of Cold/Hot ZHENG CG represented by CCL2 and LEP suggested that Hot ZHENG CG might be characterized by over-inflammation and exuberant metabolism, and Cold ZHENG CG showed a trend of suppression in immune regulation and energy metabolism. And biomarkers of Cold/Hot ZHENG showed also significant changes in the progression of gastric cancer. And biomarkers and pathways of Hot herbs intend to regulate immune responses and energy metabolism, while those of Cold herbs were likely to participate in anti-inflammation effect. Conclusion: In this study, we found that the biomarkers and mechanism of Cold/Hot ZHENG CG and those of Cold/Hot herbs were closely related to the regulation of immune and metabolisms. These findings may reflect the mechanism, build bridges between multiple views of Cold/Hot ZHENG and Cold/Hot herbs, and provide a research paradigm for further achieving precision TCM.

Target Specific De Novo Design of Drug Candidate Molecules with Graph Transformer-based Generative Adversarial Networks

Discovering novel drug candidate molecules is one of the most fundamental and critical steps in drug development. Generative deep learning models, which create synthetic data given a probability distribution, have been developed with the purpose of picking completely new samples from a partially known space. Generative models offer high potential for designing de novo molecules; however, in order for them to be useful in real-life drug development pipelines, these models should be able to design target-specific molecules, which is the next step in this field. In this study, we propose DrugGEN, for the de novo design of drug candidate molecules that interact with selected target proteins. The proposed system represents compounds and protein structures as graphs and processes them via serially connected two generative adversarial networks comprising graph transformers. DrugGEN is trained using a large dataset of compounds from ChEMBL and target-specific bioactive molecules, to design effective and specific inhibitory molecules against the AKT1 protein, which has critical importance for developing treatments against various types of cancer. On fundamental benchmarks, DrugGEN models have either competitive or better performance against other methods. To assess the target-specific generation performance, we conducted further in silico analysis with molecular docking and deep learning-based bioactivity prediction. Results indicate that de novo molecules have high potential for interacting with the AKT1 protein structure in the level of its native ligand. DrugGEN can be used to design completely novel and effective target-specific drug candidate molecules for any druggable protein, given target features and a dataset of experimental bioactivities. Code base, datasets, results and trained models of DrugGEN are available at https://github.com/HUBioDataLab/DrugGEN

3D PETCT Tumor Lesion Segmentation via GCN Refinement

Whole-body PET/CT scan is an important tool for diagnosing various malignancies (e.g., malignant melanoma, lymphoma, or lung cancer), and accurate segmentation of tumors is a key part for subsequent treatment. In recent years, CNN-based segmentation methods have been extensively investigated. However, these methods often give inaccurate segmentation results, such as over-segmentation and under-segmentation. Therefore, to address such issues, we propose a post-processing method based on a graph convolutional neural network (GCN) to refine inaccurate segmentation parts and improve the overall segmentation accuracy. Firstly, nnUNet is used as an initial segmentation framework, and the uncertainty in the segmentation results is analyzed. Certainty and uncertainty nodes establish the nodes of a graph neural network. Each node and its 6 neighbors form an edge, and 32 nodes are randomly selected for uncertain nodes to form edges. The highly uncertain nodes are taken as the subsequent refinement targets. Secondly, the nnUNet result of the certainty nodes is used as label to form a semi-supervised graph network problem, and the uncertainty part is optimized through training the GCN network to improve the segmentation performance. This describes our proposed nnUNet-GCN segmentation framework. We perform tumor segmentation experiments on the PET/CT dataset in the MICCIA2022 autoPET challenge. Among them, 30 cases are randomly selected for testing, and the experimental results show that the false positive rate is effectively reduced with nnUNet-GCN refinement. In quantitative analysis, there is an improvement of 2.12 % on the average Dice score, 6.34 on 95 % Hausdorff Distance (HD95), and 1.72 on average symmetric surface distance (ASSD). The quantitative and qualitative evaluation results show that GCN post-processing methods can effectively improve tumor segmentation performance.

Designing and simulating realistic spatial frequency domain imaging systems using open-source 3D rendering software

Spatial frequency domain imaging (SFDI) is a low-cost imaging technique that can deliver real-time maps of absorption and reduced scattering coefficients. However, there are a wide range of imaging geometries that practical SFDI systems must cope with including imaging flat samples ex vivo, imaging inside tubular lumen in vivo such as in an endoscopy, and measuring tumours or polyps of varying shapes, sizes and optical properties. There is a need for a design and simulation tool to accelerate design and fabrication of new SFDI systems. We present such a system implemented using open-source 3D design and ray-tracing software Blender that is capable of simulating media with realistic optical properties (mimicking healthy and cancerous tissue), a wide variety of shapes and size, and in both planar and tubular imaging geometries. We first demonstrate quantitative agreement between Monte-Carlo simulated scattering and absorption coefficients and those measured from our Blender system. Next, we show the ability of the system to simulate absorption, scattering and shape for flat samples with small simulated tumours and show that the improved contrast associated with SFDI is reproduced. Finally, to demonstrate the versatility of the system as a design tool we show that it can be used to generate a custom look-up-table for mapping from modulation amplitude values to absorption and scattering values in a tubular geometry, simulating a lumen. As a demonstrative example we show that longitudinal sectioning of the tube, with separate look-up tables for each section, significantly improves accuracy of SFDI, representing an important design insight for future systems. We therefore anticipate our simulation system will significantly aid in the design and development of novel SFDI systems, especially as such systems are miniaturised for deployment in endoscopic and laparoscopic systems.

The Race of mRNA therapy: Evidence from Patent Landscape

mRNA therapy is gaining worldwide attention as an emerging therapeutic approach. The widespread use of mRNA vaccines during the COVID-19 outbreak has demonstrated the potential of mRNA therapy. As mRNA-based drugs have expanded and their indications have broadened, more patents for mRNA innovations have emerged. The global patent landscape for mRNA therapy has not yet been analyzed, indicating a research gap in need of filling, from new technology to productization. This study uses social network analysis with the patent quality assessment to investigate the temporal trends, citation relationship, and significant litigation for 16,101 mRNA therapy patents and summarizes the hot topics and potential future directions for this industry. The information obtained in this study not only may be utilized as a tool of knowledge for researchers in a comprehensive and integrated way but can also provide inspiration for efficient production methods for mRNA drugs. This study shows that infectious diseases and cancer are currently the primary applications for mRNA drugs. Emerging patent activity and lawsuits in this field are demonstrating that delivery technology remains one of the key challenges in the field and that drug-targeting research in combination with vector technology will be one of the major directions for the industry going forward. With significant funding, new organizations have developed novel delivery technologies in an attempt to break into the patent thicket established by companies such as Arbutus. The global mRNA therapeutic landscape is undergoing a multifaceted development pattern, and the monopoly of giant companies is being challenged.

OpenTPS -- Open-source treatment planning system for research in proton therapy

Introduction. Treatment planning systems (TPS) are an essential component for simulating and optimizing a radiation therapy treatment before administering it to the patient. It ensures that the tumor is well covered and the dose to the healthy tissues is minimized. However, the TPS provided by commercial companies often come with a large panel of tools, each implemented in the form of a black-box making it difficult for researchers to use them for implementing and testing new ideas. To address this issue, we have developed an open-source TPS. Approach. We have developed an open-source software platform, OpenTPS (opentps.org), to generate treatment plans for external beam radiation therapy, and in particular for proton therapy. It is designed to be a flexible and user-friendly platform (coded with the freely usable Python language) that can be used by medical physicists, radiation oncologists, and other members of the radiation therapy community to create customized treatment plans for educational and research purposes. Result. OpenTPS includes a range of tools and features that can be used to analyze patient anatomy, simulate the delivery of the radiation beam, and optimize the treatment plan to achieve the desired dose distribution. It can be used to create treatment plans for a variety of cancer types and was designed to be extended to other treatment modalities. Significance. A new open-source treatment planning system has been built for research in proton therapy. Its flexibility allows an easy integration of new techniques and customization of treatment plans. It is freely available for use and is regularly updated and supported by a community of users and developers who contribute to the ongoing development and improvement of the software.

Evolutionary Computation in Action: Feature Selection for Deep Embedding Spaces of Gigapixel Pathology Images

One of the main obstacles of adopting digital pathology is the challenge of efficient processing of hyperdimensional digitized biopsy samples, called whole slide images (WSIs). Exploiting deep learning and introducing compact WSI representations are urgently needed to accelerate image analysis and facilitate the visualization and interpretability of pathology results in a postpandemic world. In this paper, we introduce a new evolutionary approach for WSI representation based on large-scale multi-objective optimization (LSMOP) of deep embeddings. We start with patch-based sampling to feed KimiaNet , a histopathology-specialized deep network, and to extract a multitude of feature vectors. Coarse multi-objective feature selection uses the reduced search space strategy guided by the classification accuracy and the number of features. In the second stage, the frequent features histogram (FFH), a novel WSI representation, is constructed by multiple runs of coarse LSMOP. Fine evolutionary feature selection is then applied to find a compact (short-length) feature vector based on the FFH and contributes to a more robust deep-learning approach to digital pathology supported by the stochastic power of evolutionary algorithms. We validate the proposed schemes using The Cancer Genome Atlas (TCGA) images in terms of WSI representation, classification accuracy, and feature quality. Furthermore, a novel decision space for multicriteria decision making in the LSMOP field is introduced. Finally, a patch-level visualization approach is proposed to increase the interpretability of deep features. The proposed evolutionary algorithm finds a very compact feature vector to represent a WSI (almost 14,000 times smaller than the original feature vectors) with 8% higher accuracy compared to the codes provided by the state-of-the-art methods.

TransNetR: Transformer-based Residual Network for Polyp Segmentation with Multi-Center Out-of-Distribution Testing

Colonoscopy is considered the most effective screening test to detect colorectal cancer (CRC) and its precursor lesions, i.e., polyps. However, the procedure experiences high miss rates due to polyp heterogeneity and inter-observer dependency. Hence, several deep learning powered systems have been proposed considering the criticality of polyp detection and segmentation in clinical practices. Despite achieving improved outcomes, the existing automated approaches are inefficient in attaining real-time processing speed. Moreover, they suffer from a significant performance drop when evaluated on inter-patient data, especially those collected from different centers. Therefore, we intend to develop a novel real-time deep learning based architecture, Transformer based Residual network (TransNetR), for colon polyp segmentation and evaluate its diagnostic performance. The proposed architecture, TransNetR, is an encoder-decoder network that consists of a pre-trained ResNet50 as the encoder, three decoder blocks, and an upsampling layer at the end of the network. TransNetR obtains a high dice coefficient of 0.8706 and a mean Intersection over union of 0.8016 and retains a real-time processing speed of 54.60 on the Kvasir-SEG dataset. Apart from this, the major contribution of the work lies in exploring the generalizability of the TransNetR by testing the proposed algorithm on the out-of-distribution (test distribution is unknown and different from training distribution) dataset. As a use case, we tested our proposed algorithm on the PolypGen (6 unique centers) dataset and two other popular polyp segmentation benchmarking datasets. We obtained state-of-the-art performance on all three datasets during out-of-distribution testing. The source code of TransNetR will be made publicly available at https://github.com/DebeshJha.

End-to-end Deformable Attention Graph Neural Network for Single-view Liver Mesh Reconstruction

Intensity modulated radiotherapy (IMRT) is one of the most common modalities for treating cancer patients. One of the biggest challenges is precise treatment delivery that accounts for varying motion patterns originating from free-breathing. Currently, image-guided solutions for IMRT is limited to 2D guidance due to the complexity of 3D tracking solutions. We propose a novel end-to-end attention graph neural network model that generates in real-time a triangular shape of the liver based on a reference segmentation obtained at the preoperative phase and a 2D MRI coronal slice taken during the treatment. Graph neural networks work directly with graph data and can capture hidden patterns in non-Euclidean domains. Furthermore, contrary to existing methods, it produces the shape entirely in a mesh structure and correctly infers mesh shape and position based on a surrogate image. We define two on-the-fly approaches to make the correspondence of liver mesh vertices with 2D images obtained during treatment. Furthermore, we introduce a novel task-specific identity loss to constrain the deformation of the liver in the graph neural network to limit phenomenons such as flying vertices or mesh holes. The proposed method achieves results with an average error of 3.06 +- 0.7 mm and Chamfer distance with L2 norm of 63.14 +- 27.28.

Molecular Identifification, Antioxidant Effifficacy of Phenolic Compounds, and Antimicrobial Activity of Beta-Carotene Isolated from Fruiting Bodies of Suillus sp

Suillus species, in general, are edible mushrooms, and environmentally important that are associated mostly with pine trees in the tropics regions. These fungi considered a remarkable source of phenolic compounds that play a crucial role as antioxidants which may reduce the risk of most human chronic diseases such as cancer, diabetes, asthma, atherosclerosis, Alzheimer, and others. On the other hand, carotenoids (\b{eta} carotene) are the most popular natural pigments which play an important role to protect the plants from photo-oxidative reactions. In human, these compounds prevent oxidative stress and expects to have antimicrobial activity. Here, the phenolic compounds were extracted with Ethyl acetate from fruiting bodies of Suillus sp and analyzed by HPLC, the antioxidant activity (reducing power%) of phenolic compounds was determined at the concentrations of 1, 2.5, and 5 mg/mL. Antimicrobial activity of \b{eta} carotene pigment was measured at a concentration of 100 mg/mL against some human pathogenic bacteria such as Escherichia coli, Pseudomonas aeruginosa, Klebsiella pneumonia, and Staphylococcus aureus. The specific DNA region ITS was amplified and sequenced using ITS1 and ITS4 primers with some bioinformatics analyses. The phenolic extract isolated from fruiting bodies of Suillus sp showed a remarkable antioxidant activity by increasing the reducing power percent (from F+3 ions to F+2 ions) comparing with the industrial antioxidant (Propyl gallate) at all used concentrations. Percent of reducing power of phenolic compounds were 75.5, 84.9 and 95.7% at concentrations of 1, 2.5, and 5 mg/mL respectively; comparing with PG were 65.9, 81.3, and 93.3 at 1, 2.5, and 5 mg/mL respectively. The \b{eta} carotene pigment revealed a significant antimicrobial activity at a concentration of 100 mg/mL against K. pneumonia, E. coli, and S. aureus.

Translating Radiology Reports into Plain Language using ChatGPT and GPT-4 with Prompt Learning: Promising Results, Limitations, and Potential

The large language model called ChatGPT has drawn extensively attention because of its human-like expression and reasoning abilities. In this study, we investigate the feasibility of using ChatGPT in experiments on using ChatGPT to translate radiology reports into plain language for patients and healthcare providers so that they are educated for improved healthcare. Radiology reports from 62 low-dose chest CT lung cancer screening scans and 76 brain MRI metastases screening scans were collected in the first half of February for this study. According to the evaluation by radiologists, ChatGPT can successfully translate radiology reports into plain language with an average score of 4.27 in the five-point system with 0.08 places of information missing and 0.07 places of misinformation. In terms of the suggestions provided by ChatGPT, they are general relevant such as keeping following-up with doctors and closely monitoring any symptoms, and for about 37% of 138 cases in total ChatGPT offers specific suggestions based on findings in the report. ChatGPT also presents some randomness in its responses with occasionally over-simplified or neglected information, which can be mitigated using a more detailed prompt. Furthermore, ChatGPT results are compared with a newly released large model GPT-4, showing that GPT-4 can significantly improve the quality of translated reports. Our results show that it is feasible to utilize large language models in clinical education, and further efforts are needed to address limitations and maximize their potential.

A Data Augmentation Method and the Embedding Mechanism for Detection and Classification of Pulmonary Nodules on Small Samples

Detection of pulmonary nodules by CT is used for screening lung cancer in early stages.omputer aided diagnosis (CAD) based on deep-learning method can identify the suspected areas of pulmonary nodules in CT images, thus improving the accuracy and efficiency of CT diagnosis. The accuracy and robustness of deep learning models. Method:In this paper, we explore (1) the data augmentation method based on the generation model and (2) the model structure improvement method based on the embedding mechanism. Two strategies have been introduced in this study: a new data augmentation method and a embedding mechanism. In the augmentation method, a 3D pixel-level statistics algorithm is proposed to generate pulmonary nodule and by combing the faked pulmonary nodule and healthy lung, we generate new pulmonary nodule samples. The embedding mechanism are designed to better understand the meaning of pixels of the pulmonary nodule samples by introducing hidden variables. Result: The result of the 3DVNET model with the augmentation method for pulmonary nodule detection shows that the proposed data augmentation method outperforms the method based on generative adversarial network (GAN) framework, training accuracy improved by 1.5%, and with embedding mechanism for pulmonary nodules classification shows that the embedding mechanism improves the accuracy and robustness for the classification of pulmonary nodules obviously, the model training accuracy is close to 1 and the model testing F1-score is 0.90.Conclusion:he proposed data augmentation method and embedding mechanism are beneficial to improve the accuracy and robustness of the model, and can be further applied in other common diagnostic imaging tasks.

Medical diffusion on a budget: textual inversion for medical image generation

Diffusion-based models for text-to-image generation have gained immense popularity due to recent advancements in efficiency, accessibility, and quality. Although it is becoming increasingly feasible to perform inference with these systems using consumer-grade GPUs, training them from scratch still requires access to large datasets and significant computational resources. In the case of medical image generation, the availability of large, publicly accessible datasets that include text reports is limited due to legal and ethical concerns. While training a diffusion model on a private dataset may address this issue, it is not always feasible for institutions lacking the necessary computational resources. This work demonstrates that pre-trained Stable Diffusion models, originally trained on natural images, can be adapted to various medical imaging modalities by training text embeddings with textual inversion. In this study, we conducted experiments using medical datasets comprising only 100 samples from three medical modalities. Embeddings were trained in a matter of hours, while still retaining diagnostic relevance in image generation. Experiments were designed to achieve several objectives. Firstly, we fine-tuned the training and inference processes of textual inversion, revealing that larger embeddings and more examples are required. Secondly, we validated our approach by demonstrating a 2\% increase in the diagnostic accuracy (AUC) for detecting prostate cancer on MRI, which is a challenging multi-modal imaging modality, from 0.78 to 0.80. Thirdly, we performed simulations by interpolating between healthy and diseased states, combining multiple pathologies, and inpainting to show embedding flexibility and control of disease appearance. Finally, the embeddings trained in this study are small (less than 1 MB), which facilitates easy sharing of medical data with reduced privacy concerns.

How to design a MAMS-ROCI (aka DURATIONS) randomised trial: the REFINE-Lung case study

Background. The DURATIONS design has been recently proposed as a practical alternative to a standard two-arm non-inferiority design when the goal is to optimise some continuous aspect of treatment administration, e.g. duration or frequency, preserving efficacy but improving on secondary outcomes such as safety, costs or convenience. The main features of this design are that (i) it randomises patients to a moderate number of arms across the continuum and (ii) it uses a model to share information across arms. While papers published to date about the design have focused on analysis aspects, here we show how to design such a trial in practice. We use the REFINE-Lung trial as an example; this is a trial seeking the optimal frequency of immunotherapy treatment for non-small cell lung cancer patients. Because the aspect of treatment administration to optimise is frequency, rather than duration, we propose to rename the design as Multi-Arm Multi-Stage Response Over Continuous Intervention (MAMS-ROCI). Methods. We show how simulations can be used to design such a trial. We propose to use the ADEMP framework to plan such simulations, clearly specifying aims, data generating mechanisms, estimands, methods and performance measures before coding and analysing the simulations. We discuss the possible choices to be made using the REFINE-Lung trial as an example. Results. We describe all the choices made while designing the REFINE-Lung trial, and the results of the simulations performed. We justify our choice of total sample size based on these results. Conclusions. MAMS-ROCI trials can be designed using simulation studies that have to be carefully planned and conducted. REFINE-Lung has been designed using such an approach and we have shown how researchers could similarly design their own MAMS-ROCI trial.

Deep-Learning-based Fast and Accurate 3D CT Deformable Image Registration in Lung Cancer

Purpose: In some proton therapy facilities, patient alignment relies on two 2D orthogonal kV images, taken at fixed, oblique angles, as no 3D on-the-bed imaging is available. The visibility of the tumor in kV images is limited since the patient's 3D anatomy is projected onto a 2D plane, especially when the tumor is behind high-density structures such as bones. This can lead to large patient setup errors. A solution is to reconstruct the 3D CT image from the kV images obtained at the treatment isocenter in the treatment position. Methods: An asymmetric autoencoder-like network built with vision-transformer blocks was developed. The data was collected from 1 head and neck patient: 2 orthogonal kV images (1024x1024 voxels), 1 3D CT with padding (512x512x512) acquired from the in-room CT-on-rails before kVs were taken and 2 digitally-reconstructed-radiograph (DRR) images (512x512) based on the CT. We resampled kV images every 8 voxels and DRR and CT every 4 voxels, thus formed a dataset consisting of 262,144 samples, in which the images have a dimension of 128 for each direction. In training, both kV and DRR images were utilized, and the encoder was encouraged to learn the jointed feature map from both kV and DRR images. In testing, only independent kV images were used. The full-size synthetic CT (sCT) was achieved by concatenating the sCTs generated by the model according to their spatial information. The image quality of the synthetic CT (sCT) was evaluated using mean absolute error (MAE) and per-voxel-absolute-CT-number-difference volume histogram (CDVH). Results: The model achieved a speed of 2.1s and a MAE of <40HU. The CDVH showed that <5% of the voxels had a per-voxel-absolute-CT-number-difference larger than 185 HU. Conclusion: A patient-specific vision-transformer-based network was developed and shown to be accurate and efficient to reconstruct 3D CT images from kV images.

Benchmarking ChatGPT-4 on ACR Radiation Oncology In-Training (TXIT) Exam and Red Journal Gray Zone Cases: Potentials and Challenges for AI-Assisted Medical Education and Decision Making in Radiation Oncology

The potential of large language models in medicine for education and decision making purposes has been demonstrated as they achieve decent scores on medical exams such as the United States Medical Licensing Exam (USMLE) and the MedQA exam. In this work, we evaluate the performance of ChatGPT-4 in the specialized field of radiation oncology using the 38th American College of Radiology (ACR) radiation oncology in-training (TXIT) exam and the 2022 Red Journal gray zone cases. For the TXIT exam, ChatGPT-3.5 and ChatGPT-4 have achieved the scores of 63.65% and 74.57%, respectively, highlighting the advantage of the latest ChatGPT-4 model. Based on the TXIT exam, ChatGPT-4's strong and weak areas in radiation oncology are identified to some extent. Specifically, ChatGPT-4 demonstrates good knowledge of statistics, CNS & eye, pediatrics, biology, and physics but has limitations in bone & soft tissue and gynecology, as per the ACR knowledge domain. Regarding clinical care paths, ChatGPT-4 performs well in diagnosis, prognosis, and toxicity but lacks proficiency in topics related to brachytherapy and dosimetry, as well as in-depth questions from clinical trials. For the gray zone cases, ChatGPT-4 is able to suggest a personalized treatment approach to each case with high correctness and comprehensiveness. Most importantly, it provides novel treatment aspects for many cases, which are not suggested by any human experts. Both evaluations demonstrate the potential of ChatGPT-4 in medical education for the general public and cancer patients, as well as the potential to aid clinical decision-making, while acknowledging its limitations in certain domains. Because of the risk of hallucination, facts provided by ChatGPT always need to be verified.

Prediction of brain tumor recurrence location based on multi-modal fusion and nonlinear correlation learning

Brain tumor is one of the leading causes of cancer death. The high-grade brain tumors are easier to recurrent even after standard treatment. Therefore, developing a method to predict brain tumor recurrence location plays an important role in the treatment planning and it can potentially prolong patient's survival time. There is still little work to deal with this issue. In this paper, we present a deep learning-based brain tumor recurrence location prediction network. Since the dataset is usually small, we propose to use transfer learning to improve the prediction. We first train a multi-modal brain tumor segmentation network on the public dataset BraTS 2021. Then, the pre-trained encoder is transferred to our private dataset for extracting the rich semantic features. Following that, a multi-scale multi-channel feature fusion model and a nonlinear correlation learning module are developed to learn the effective features. The correlation between multi-channel features is modeled by a nonlinear equation. To measure the similarity between the distributions of original features of one modality and the estimated correlated features of another modality, we propose to use Kullback-Leibler divergence. Based on this divergence, a correlation loss function is designed to maximize the similarity between the two feature distributions. Finally, two decoders are constructed to jointly segment the present brain tumor and predict its future tumor recurrence location. To the best of our knowledge, this is the first work that can segment the present tumor and at the same time predict future tumor recurrence location, making the treatment planning more efficient and precise. The experimental results demonstrated the effectiveness of our proposed method to predict the brain tumor recurrence location from the limited dataset.

Using Spatio-Temporal Dual-Stream Network with Self-Supervised Learning for Lung Tumor Classification on Radial Probe Endobronchial Ultrasound Video

The purpose of this study is to develop a computer-aided diagnosis system for classifying benign and malignant lung lesions, and to assist physicians in real-time analysis of radial probe endobronchial ultrasound (EBUS) videos. During the biopsy process of lung cancer, physicians use real-time ultrasound images to find suitable lesion locations for sampling. However, most of these images are difficult to classify and contain a lot of noise. Previous studies have employed 2D convolutional neural networks to effectively differentiate between benign and malignant lung lesions, but doctors still need to manually select good-quality images, which can result in additional labor costs. In addition, the 2D neural network has no ability to capture the temporal information of the ultrasound video, so it is difficult to obtain the relationship between the features of the continuous images. This study designs an automatic diagnosis system based on a 3D neural network, uses the SlowFast architecture as the backbone to fuse temporal and spatial features, and uses the SwAV method of contrastive learning to enhance the noise robustness of the model. The method we propose includes the following advantages, such as (1) using clinical ultrasound films as model input, thereby reducing the need for high-quality image selection by physicians, (2) high-accuracy classification of benign and malignant lung lesions can assist doctors in clinical diagnosis and reduce the time and risk of surgery, and (3) the capability to classify well even in the presence of significant image noise. The AUC, accuracy, precision, recall and specificity of our proposed method on the validation set reached 0.87, 83.87%, 86.96%, 90.91% and 66.67%, respectively. The results have verified the importance of incorporating temporal information and the effectiveness of using the method of contrastive learning on feature extraction.

eXplainable Artificial Intelligence on Medical Images: A Survey

Over the last few years, the number of works about deep learning applied to the medical field has increased enormously. The necessity of a rigorous assessment of these models is required to explain these results to all people involved in medical exams. A recent field in the machine learning area is explainable artificial intelligence, also known as XAI, which targets to explain the results of such black box models to permit the desired assessment. This survey analyses several recent studies in the XAI field applied to medical diagnosis research, allowing some explainability of the machine learning results in several different diseases, such as cancers and COVID-19.

On-line Dose Calculation Using Deep Learning for Beams Selection in Non-Coplanar Radiotherapy

Non-coplanar Intensity-Modulated Radiation Therapy (IMRT) goes a step further by orienting the gantry carrying the radiation beam and the patient couch in a non-coplanar manner to accurately target the cancer region and better avoid organs-at-risk. The use of a non-coplanar treatment trajectory significantly enhances the degree of freedom and flexibility but increases drastically the complexity of the optimization. In inverse planning optimization the dose contribution for all potential beam directions is usually pre-calculates and pre-loads into the Treatment Planning System (TPS). The size the dose matrix becomes more critical when moving from coplanar IMRT to non-coplanar IMRT since the number of beams increases drastically. A solution would be to calculate "on-the-fly" the dose contribution to each new candidate beam during optimization. This is only possible if a dose calculation engine is fast enough to be used online during optimization iterations, which is not the case in standard method. Therefore, in this work we propose an IMRT optimization scheme using deep learning based dose engine to compute the dose matrix on-line. The proposed deep learning approach will be combined into a simulated-annealing-based optimization method for non-coplanar IMRT. Since the dose engine will compute the dose contribution on-line during the optimization, the final main optimization method requires to keep in memory a very lightweight dose matrix. The proposed method was compared with clinical data showing a good agreement considering dosimetry of the treatment plans. The main advantage of the proposed method was the reduction of the memory storage from 9GB to 10MB during the optimization process.

AdaMSS: Adaptive Multi-Modality Segmentation-to-Survival Learning for Survival Outcome Prediction from PET/CT Images

Survival prediction is a major concern for cancer management. Deep survival models based on deep learning have been widely adopted to perform end-to-end survival prediction from medical images. Recent deep survival models achieved promising performance by jointly performing tumor segmentation with survival prediction, where the models were guided to extract tumor-related information through Multi-Task Learning (MTL). However, these deep survival models have difficulties in exploring out-of-tumor prognostic information. In addition, existing deep survival models are unable to effectively leverage multi-modality images. Empirically-designed fusion strategies were commonly adopted to fuse multi-modality information via task-specific manually-designed networks, thus limiting the adaptability to different scenarios. In this study, we propose an Adaptive Multi-modality Segmentation-to-Survival model (AdaMSS) for survival prediction from PET/CT images. Instead of adopting MTL, we propose a novel Segmentation-to-Survival Learning (SSL) strategy, where our AdaMSS is trained for tumor segmentation and survival prediction sequentially in two stages. This strategy enables the AdaMSS to focus on tumor regions in the first stage and gradually expand its focus to include other prognosis-related regions in the second stage. We also propose a data-driven strategy to fuse multi-modality information, which realizes adaptive optimization of fusion strategies based on training data during training. With the SSL and data-driven fusion strategies, our AdaMSS is designed as an adaptive model that can self-adapt its focus regions and fusion strategy for different training stages. Extensive experiments with two large clinical datasets show that our AdaMSS outperforms state-of-the-art survival prediction methods.

JulianA: An automatic treatment planning platform for intensity-modulated proton therapy

Creating high quality treatment plans is crucial for a successful radiotherapy treatment. However, it demands substantial effort and special training for dosimetrists. Existing automated treatment planning systems typically require either an explicit prioritization of planning objectives, human-assigned objective weights, large amounts of historic plans to train an artificial intelligence or long planning times. Many of the existing auto-planning tools are difficult to extend to new planning goals. A new spot weight optimisation algorithm, called JulianA, was developed. The algorithm minimises a scalar loss function that is built only based on the prescribed dose to the tumour and organs at risk (OARs), but does not rely on historic plans. The objective weights in the loss function have default values that do not need to be changed for the patients in our dataset. The system is a versatile tool for researchers and clinicians without specialised programming skills. Extending it is as easy as adding an additional term to the loss function. JulianA was validated on a dataset of 19 patients with intra- and extracerebral neoplasms within the cranial region that had been treated at our institute. For each patient, a reference plan which was delivered to the cancer patient, was exported from our treatment database. Then JulianA created the auto plan using the same beam arrangement. The reference and auto plans were given to a blinded independent reviewer who assessed the acceptability of each plan, ranked the plans and assigned the human-/machine-made labels. The auto plans were considered acceptable in 16 out of 19 patients and at least as good as the reference plan for 11 patients. Whether a plan was crafted by a dosimetrist or JulianA was only recognised for 9 cases. The median time for the spot weight optimisation is approx. 2 min (range: 0.5 min - 7 min).

At-Admission Prediction of Mortality and Pulmonary Embolism in COVID-19 Patients Using Statistical and Machine Learning Methods: An International Cohort Study

By September, 2022, more than 600 million cases of SARS-CoV-2 infection have been reported globally, resulting in over 6.5 million deaths. COVID-19 mortality risk estimators are often, however, developed with small unrepresentative samples and with methodological limitations. It is highly important to develop predictive tools for pulmonary embolism (PE) in COVID-19 patients as one of the most severe preventable complications of COVID-19. Using a dataset of more than 800,000 COVID-19 patients from an international cohort, we propose a cost-sensitive gradient-boosted machine learning model that predicts occurrence of PE and death at admission. Logistic regression, Cox proportional hazards models, and Shapley values were used to identify key predictors for PE and death. Our prediction model had a test AUROC of 75.9% and 74.2%, and sensitivities of 67.5% and 72.7% for PE and all-cause mortality respectively on a highly diverse and held-out test set. The PE prediction model was also evaluated on patients in UK and Spain separately with test results of 74.5% AUROC, 63.5% sensitivity and 78.9% AUROC, 95.7% sensitivity. Age, sex, region of admission, comorbidities (chronic cardiac and pulmonary disease, dementia, diabetes, hypertension, cancer, obesity, smoking), and symptoms (any, confusion, chest pain, fatigue, headache, fever, muscle or joint pain, shortness of breath) were the most important clinical predictors at admission. Our machine learning model developed from an international cohort can serve to better regulate hospital risk prioritisation of at-risk patients.

A Scintillator Beam Monitor for Real-Time FLASH Radiotherapy

FLASH Radiotherapy (RT) is a potentially new cancer radiotherapy technique where an entire therapeutic dose is delivered in about 0.1 s and at ~1000 times higher dose rate than in conventional RT. For clinical trials to be conducted safely, precise and fast beam monitoring that can generate an out-of-tolerance beam interrupt is required. A FLASH Beam Scintillator Monitor (FBSM) is being developed based in part on two novel proprietary scintillator materials: an organic polymeric material (PM) and inorganic hybrid (HM). The FBSM provides large area coverage, low mass profile, linear response over a broad dynamic range, radiation tolerance, and real-time analysis IEC-compliant fast beam-interrupt signal. This paper includes the design concept and test results from prototype devices in radiation beams that include heavy ions, low energy protons at nA currents, FLASH level dose per pulse electron beams, and in a hospital radiotherapy clinic with electron beams. Results include image quality, response linearity, radiation hardness, spatial resolution, and real-time data processing. PM and HM scintillator exhibited no measurable drop in signal after a cumulative dose of 9 kGy and 20 kGy respectively. HM showed a small -0.02%/kGy signal decrease after a 212 kGy cumulative dose resulting from continuous exposure for 15 minutes at a high FLASH dose rate of 234 Gy/s. These tests established the linear response of the FBSM with respect to beam currents, dose per pulse, and material thickness. Comparison with commercial Gafchromic film indicates that the FBSM produces a high resolution 2D beam image and can reproduce a nearly identical beam profile, including primary beam tails. At 20 kfps or 50 microsec/frame, the real-time FPGA based computation and analysis of beam position, beam shape, and beam dose takes < 1 microsec.

AlGaN/AlN Stranski-Krastanov quantum dots for highly efficient electron beam pumped emitters: The role of miniaturization and composition to attain far UV-C emission

Conventional ultraviolet (UV) lamps for disinfection emit radiation in the 255-270 nm range, which poses a high risk of causing cancer and cataracts. To address these concerns, solid-state far UV-C sources emitting below 240 nm are gaining attention as a safe and sustainable disinfection solution for occupied spaces. Here, we delve into the extension of the AlxGa1-xN/AlN quantum dot (QD) technology towards the far UV-C range, which presents various challenges associated with the reduction of the lattice mismatch and band offset when Al is incorporated in the QDs. We explore the structural and optical impact of increasing the Al content through the increase of the Al flux and eventual correction of the Ga flux to maintain a constant metal/N ratio. We also examine the impact of extreme miniaturization of the QDs, achieved through a reduction of their growth time, on the spectral behavior and internal quantum efficiency (IQE). The high Al content results in QDs with a reduced aspect ratio (height/diameter) and thicker wetting layer when compared to the GaN/AlN system. Self-assembled QDs grown with a metal/N ratio ranging from 0.5 to 0.8 show an IQE around 50%, independent of the Al content (up to 65%) or emission wavelength (300-230 nm). However, samples emitting at wavelengths below 270 nm exhibit a bimodal luminescence associated with inhomogeneous in-plane emission attributed to fluctuations of the QD shape associated with extended defects. Reducing the QD size exacerbates the bimodality without reducing the emission wavelength. The power efficiencies under electron beam pumping range from 0.4% to 1%, with clear potential for improvement through surface treatments that enhance light extraction efficiency.

Joint regional uptake quantification of Thorium-227 and Radium-223 using a multiple-energy-window projection-domain quantitative SPECT method

Thorium-227-based alpha-particle radiopharmaceutical therapies (alpha-RPTs) are currently being investigated in several clinical and pre-clinical studies. After administration, Thorium-227 decays to Radium-223, another alpha-particle-emitting isotope, which redistributes within the patient. Reliable dose quantification of both Thorium-227 and Radium-223 is clinically important, and SPECT can perform this quantification as these isotopes also emit gamma-ray photons. However, reliable quantification is challenging for several reasons: the orders-of-magnitude lower activity compared to conventional SPECT, resulting in a very low number of detected counts, the presence of multiple photopeaks and substantial overlap in the emission spectra of these isotopes. To address these issues, we propose a multiple-energy-window projection-domain quantification (MEW-PDQ) method that jointly estimates the regional activity uptake of both Thorium-227 and Radium-223 directly using the SPECT projection data from multiple energy windows. We evaluated the method with realistic simulation studies conducted with anthropomorphic digital phantoms, including a virtual imaging trial in the context of imaging patients with bone metastases of prostate cancer who were treated with Thorium-227-based alpha-RPTs. The proposed method yielded reliable regional uptake estimates of both isotopes and outperformed state-of-art methods across different lesion sizes, contrasts, and varying levels of intra-lesion heterogeneity. This superior performance was also observed in the virtual imaging trial. Additionally, the variance of the estimated uptake approached the Cram\'er-Rao lower bound-defined theoretical limit. These results provide strong evidence in support of this method for reliable uptake quantification in Thorium-227-based alpha-RPTs.

An Investigation into the Effects of Pre-training Data Distributions for Pathology Report Classification

Pre-trained transformer models have demonstrated success across many natural language processing (NLP) tasks. In applying these models to the clinical domain, a prevailing assumption is that pre-training language models from scratch on large-scale biomedical data results in substantial improvements. We test this assumption with 4 pathology classification tasks on a corpus of 2907 prostate cancer pathology reports. We evaluate 5 transformer pre-trained models that are the same size but differ in pre-training corpora. Specifically, we analyze 3 categories of models: 1)General-domain: BERT and Turing Natural Language Representation (TNLR) models, which use general corpora for pre-training, 2)Mixed-domain: BioBERT which is obtained from BERT by including PubMed abstracts in pre-training and Clinical BioBERT which additionally includes MIMIC-III clinical notes and 3)Domain-specific: PubMedBERT which is pre-trained from scratch on PubMed abstracts. We find the mixed-domain and domain-specific models exhibit faster feature disambiguation during fine-tuning. However, the domain-specific model, PubMedBERT, can overfit to minority classes when presented with class imbalance, a common scenario in pathology report data. At the same time, the mixed-domain models are more resistant to overfitting. Our findings indicate that the use of general natural language and domain-specific corpora in pre-training serve complementary purposes for pathology report classification. The first enables resistance to overfitting when fine-tuning on an imbalanced dataset while the second allows for more accurate modelling of the fine-tuning domain. An expert evaluation is also conducted to reveal common outlier modes of each model. Our results could inform better fine-tuning practices in the clinical domain, to possibly leverage the benefits of mixed-domain models for imbalanced downstream datasets.

A Comparison of Mutation and Amplification-Driven Resistance Mechanisms and Their Impacts on Tumor Recurrence

Tumor recurrence, driven by the evolution of drug resistance is a major barrier to therapeutic success in cancer. Resistance is often caused by genetic alterations such as point mutation, which refers to the modification of a single genomic base pair, or gene amplification, which refers to the duplication of a region of DNA that contains a gene. Here we investigate the dependence of tumor recurrence dynamics on these mechanisms of resistance, using stochastic multi-type branching process models. We derive tumor extinction probabilities and deterministic estimates for the tumor recurrence time, defined as the time when an initially drug sensitive tumor surpasses its original size after developing resistance. For models of amplification-driven and mutation-driven resistance, we prove law of large numbers results regarding the convergence of the stochastic recurrence times to their mean. Additionally, we prove sufficient and necessary conditions for a tumor to escape extinction under the gene amplification model, discuss behavior under biologically relevant parameters, and compare the recurrence time and tumor composition in the mutation and amplification models both analytically and using simulations. In comparing these mechanisms, we find that the ratio between recurrence times driven by amplification vs. mutation depends linearly on the number of amplification events required to acquire the same degree of resistance as a mutation event, and we find that the relative frequency of amplification and mutation events plays a key role in determining the mechanism under which recurrence is more rapid. In the amplification-driven resistance model, we also observe that increasing drug concentration leads to a stronger initial reduction in tumor burden, but that the eventual recurrent tumor population is less heterogeneous, more aggressive, and harbors higher levels of drug-resistance.

Lensless polarimetric coded ptychography (pol-CP) for high-resolution, high-throughput birefringence imaging on a chip

Polarimetric imaging provides valuable insights into the polarization state of light interacting with a sample. It can infer crucial birefringence properties of bio-specimens without using any labels, thereby facilitating the diagnosis of diseases such as cancer and osteoarthritis. In this study, we introduce a novel polarimetric coded ptychography (pol-CP) approach that enables high-resolution, high-throughput birefringence imaging on a chip. Our platform deviates from traditional lens-based polarization systems by employing an integrated polarimetric coded sensor for lensless diffraction data acquisition. Utilizing Jones calculus, we quantitatively determine the birefringence retardance and orientation information of bio-specimens from four recovered intensity images. Our portable pol-CP prototype can resolve the 435-nm linewidth on the resolution target and the imaging field of view for a single acquisition is limited only by the detector size of 41 mm^2. The prototype allows for the acquisition of gigapixel birefringence images with a 180-mm^2 field of view in ~3.5 minutes, achieving an imaging throughput comparable to that of a conventional whole slide scanner. To demonstrate its biomedical applications, we perform high-throughput imaging of malaria-infected blood smears, locating parasites using birefringence contrast. We also generate birefringence maps of label-free thyroid smears to identify thyroid follicles. Notably, the recovered birefringence maps emphasize the same regions as autofluorescence images, indicating the potential for rapid on-site evaluation of label-free biopsies. The reported approach offers a portable, turnkey solution for high-resolution, high-throughput polarimetric analysis without using lenses, with potential applications in disease diagnosis, sample screening, and label-free chemical imaging.

3DSAM-adapter: Holistic Adaptation of SAM from 2D to 3D for Promptable Medical Image Segmentation

Despite that the segment anything model (SAM) achieved impressive results on general-purpose semantic segmentation with strong generalization ability on daily images, its demonstrated performance on medical image segmentation is less precise and not stable, especially when dealing with tumor segmentation tasks that involve objects of small sizes, irregular shapes, and low contrast. Notably, the original SAM architecture is designed for 2D natural images, therefore would not be able to extract the 3D spatial information from volumetric medical data effectively. In this paper, we propose a novel adaptation method for transferring SAM from 2D to 3D for promptable medical image segmentation. Through a holistically designed scheme for architecture modification, we transfer the SAM to support volumetric inputs while retaining the majority of its pre-trained parameters for reuse. The fine-tuning process is conducted in a parameter-efficient manner, wherein most of the pre-trained parameters remain frozen, and only a few lightweight spatial adapters are introduced and tuned. Regardless of the domain gap between natural and medical data and the disparity in the spatial arrangement between 2D and 3D, the transformer trained on natural images can effectively capture the spatial patterns present in volumetric medical images with only lightweight adaptations. We conduct experiments on four open-source tumor segmentation datasets, and with a single click prompt, our model can outperform domain state-of-the-art medical image segmentation models on 3 out of 4 tasks, specifically by 8.25%, 29.87%, and 10.11% for kidney tumor, pancreas tumor, colon cancer segmentation, and achieve similar performance for liver tumor segmentation. We also compare our adaptation method with existing popular adapters, and observed significant performance improvement on most datasets.

A two-sample comparison of mean survival times of uncured sub-populations

Comparing the survival times among two groups is a common problem in time-to-event analysis, for example if one would like to understand whether one medical treatment is superior to another. In the standard survival analysis setting, there has been a lot of discussion on how to quantify such difference and what can be an intuitive, easily interpretable, summary measure. In the presence of subjects that are immune to the event of interest (`cured'), we illustrate that it is not appropriate to just compare the overall survival functions. Instead, it is more informative to compare the cure fractions and the survival of the uncured sub-populations separately from each other. Our research is mainly driven by the question: if the cure fraction is similar for two available treatments, how else can we determine which is preferable? To this end, we estimate the mean survival times in the uncured fractions of both treatment groups ($MST_u$) and develop permutation tests for inference. In the first out of two connected papers, we focus on nonparametric approaches. The methods are illustrated with medical data of leukemia patients. In Part II we adjust the mean survival time of the uncured for potential confounders, which is crucial in observational settings. For each group, we employ the widely used logistic-Cox mixture cure model and estimate the $MST_u$ conditionally on a given covariate value. An asymptotic and a permutation-based approach have been developed for making inference on the difference of conditional $MST_u$'s between two groups. Contrarily to available results in the literature, in the simulation study we do not observe a clear advantage of the permutation method over the asymptotic one to justify its increased computational cost. The methods are illustrated through a practical application to breast cancer data.

Novel Pipeline for Diagnosing Acute Lymphoblastic Leukemia Sensitive to Related Biomarkers

Acute Lymphoblastic Leukemia (ALL) is one of the most common types of childhood blood cancer. The quick start of the treatment process is critical to saving the patient's life, and for this reason, early diagnosis of this disease is essential. Examining the blood smear images of these patients is one of the methods used by expert doctors to diagnose this disease. Deep learning-based methods have numerous applications in medical fields, as they have significantly advanced in recent years. ALL diagnosis is not an exception in this field, and several machine learning-based methods for this problem have been proposed. In previous methods, high diagnostic accuracy was reported, but our work showed that this alone is not sufficient, as it can lead to models taking shortcuts and not making meaningful decisions. This issue arises due to the small size of medical training datasets. To address this, we constrained our model to follow a pipeline inspired by experts' work. We also demonstrated that, since a judgement based on only one image is insufficient, redefining the problem as a multiple-instance learning problem is necessary for achieving a practical result. Our model is the first to provide a solution to this problem in a multiple-instance learning setup. We introduced a novel pipeline for diagnosing ALL that approximates the process used by hematologists, is sensitive to disease biomarkers, and achieves an accuracy of 96.15%, an F1-score of 94.24%, a sensitivity of 97.56%, and a specificity of 90.91% on ALL IDB 1. Our method was further evaluated on an out-of-distribution dataset, which posed a challenging test and had acceptable performance. Notably, our model was trained on a relatively small dataset, highlighting the potential for our approach to be applied to other medical datasets with limited data availability.

Adaptive Region Selection for Active Learning in Whole Slide Image Semantic Segmentation

The process of annotating histological gigapixel-sized whole slide images (WSIs) at the pixel level for the purpose of training a supervised segmentation model is time-consuming. Region-based active learning (AL) involves training the model on a limited number of annotated image regions instead of requesting annotations of the entire images. These annotation regions are iteratively selected, with the goal of optimizing model performance while minimizing the annotated area. The standard method for region selection evaluates the informativeness of all square regions of a specified size and then selects a specific quantity of the most informative regions. We find that the efficiency of this method highly depends on the choice of AL step size (i.e., the combination of region size and the number of selected regions per WSI), and a suboptimal AL step size can result in redundant annotation requests or inflated computation costs. This paper introduces a novel technique for selecting annotation regions adaptively, mitigating the reliance on this AL hyperparameter. Specifically, we dynamically determine each region by first identifying an informative area and then detecting its optimal bounding box, as opposed to selecting regions of a uniform predefined shape and size as in the standard method. We evaluate our method using the task of breast cancer metastases segmentation on the public CAMELYON16 dataset and show that it consistently achieves higher sampling efficiency than the standard method across various AL step sizes. With only 2.6\% of tissue area annotated, we achieve full annotation performance and thereby substantially reduce the costs of annotating a WSI dataset. The source code is available at https://github.com/DeepMicroscopy/AdaptiveRegionSelection.

Measurement of the Neutron Radius of 208Pb Through Parity-Violation in Electron Scattering

We report the first measurement of the parity-violating asymmetry A_PV in the elastic scattering of polarized electrons from 208Pb. A_PV is sensitive to the radius of the neutron distribution (Rn). The result A_PV = 0.656 \pm 0.060 (stat) \pm 0.014 (syst) ppm corresponds to a difference between the radii of the neutron and proton distributions Rn - Rp = 0.33 +0.16 -0.18 fm and provides the first electroweak observation of the neutron skin which is expected in a heavy, neutron-rich nucleus.

DHCAL with Minimal Absorber: Measurements with Positrons

In special tests, the active layers of the CALICE Digital Hadron Calorimeter prototype, the DHCAL, were exposed to low energy particle beams, without being interleaved by absorber plates. The thickness of each layer corresponded approximately to 0.29 radiation lengths or 0.034 nuclear interaction lengths, defined mostly by the copper and steel skins of the detector cassettes. This paper reports on measurements performed with this device in the Fermilab test beam with positrons in the energy range of 1 to 10 GeV. The measurements are compared to simulations based on GEANT4 and a standalone program to emulate the detailed response of the active elements.

Weakly Bound Neutron-Rich Nuclei and Cosmic Phenomena

The single particle and bulk properties of the neutron-rich nuclei constrain fundamental issues in nuclear physics and nuclear astrophysics like the limits of existence of quantum many body systems (atomic nuclei), the equation of state of neutron-rich matter, neutron star, nucleosynthesis, evolution of stars, neutron star merging etc.. The state of the art of Coulomb breakup of the neutron-rich nuclei has been used to explore those properties. Unambiguous information on detailed components of the ground-state wave-function along with quantum numbers of the valence neutron of the nuclei have been obtained from the measurement of threshold strength along with the $\gamma$-rays spectra of the core following Coulomb breakup. The shape of this threshold strength is a finger-print of the quantum numbers of the nucleon. We investigated the ground-state properties of the neutron-rich Na, Mg, Al nuclei around N $\sim$ 20 using this method at GSI, Darmstadt. Very clear evidence has been observed for melting and merging of long cherished magic shell gaps at N = 20, 28. The evanescent neutron-rich nuclei imprint their existence in stellar explosive scenarios (r-process etc.). Coulomb dissociation (CD) is one of the important indirect measurements of the capture cross-section which may provide valuable input to the model for star evolution process, particularly the r-process. Some valuable bulk properties of the neutron-rich nuclei like the density dependent symmetry energy,neutron skin etc. play a key role in understanding cosmic phenomena and these properties have been studied via electromagnetic excitation. Preliminary results of electromagnetic excitation of the neutron-rich nucleus, $^{32}$Mg are presented.

Properties of slowly rotating asteroids from the Convex Inversion Thermophysical Model

Results from the TESS mission showed that previous studies strngly underestimated the number of slow rotators, revealing the importance of studying those asteroids. For most slowly rotating asteroids (P > 12), no spin and shape model is available because of observation selection effects. This hampers determination of their thermal parameters and accurate sizes. We continue our campaign in minimising selection effects among main belt asteroids. Our targets are slow rotators with low light-curve amplitudes. The goal is to provide their scaled spin and shape models together with thermal inertia, albedo, and surface roughness to complete the statistics. Rich multi-apparition datasets of dense light curves are supplemented with data from Kepler and TESS. In addition to data in the visible range, we also use thermal data from infrared space observatories (IRAS, Akari and WISE) in a combined optimisation process using the Convex Inversion Thermophysical Model (CITPM). This novel method has so far been applied to only a few targets, and in this work we further validate the method. We present the models of 16 slow rotators. All provide good fits to both thermal and visible data. The obtained sizes are on average accurate at the 5% precision, with diameters in the range from 25 to 145 km. The rotation periods of our targets range from 11 to 59 hours, and the thermal inertia covers a wide range of values, from 2 to <400 SI units, not showing any correlation with the period. With this work we increase the sample of slow rotators with reliable spin and shape models and known thermal inertia by 40%. The thermal inertia values of our sample do not display a previously suggested increasing trend with rotation period, which might be due to their small skin depth.

Precision Determination of the Neutral Weak Form Factor of $^{48}$Ca

We report a precise measurement of the parity-violating asymmetry $A_{\rm PV}$ in the elastic scattering of longitudinally polarized electrons from $^{48}{\rm Ca}$. We measure $A_{\rm PV} =2668\pm 106\ {\rm (stat)}\pm 40\ {\rm (syst)}$ parts per billion, leading to an extraction of the neutral weak form factor $F_{\rm W} (q=0.8733$ fm$^{-1}) = 0.1304 \pm 0.0052 \ {\rm (stat)}\pm 0.0020\ {\rm (syst)}$ and the charge minus the weak form factor $F_{\rm ch} - F_{\rm W} = 0.0277\pm 0.0055$. The resulting neutron skin thickness $R_n-R_p=0.121 \pm 0.026\ {\rm (exp)} \pm 0.024\ {\rm (model)}$~fm is relatively thin yet consistent with many model calculations. The combined CREX and PREX results will have implications for future energy density functional calculations and on the density dependence of the symmetry energy of nuclear matter.

The Laser-hybrid Accelerator for Radiobiological Applications

The `Laser-hybrid Accelerator for Radiobiological Applications', LhARA, is conceived as a novel, uniquely-flexible facility dedicated to the study of radiobiology. The technologies demonstrated in LhARA, which have wide application, will be developed to allow particle-beam therapy to be delivered in a completely new regime, combining a variety of ion species in a single treatment fraction and exploiting ultra-high dose rates. LhARA will be a hybrid accelerator system in which laser interactions drive the creation of a large flux of protons or light ions that are captured using a plasma (Gabor) lens and formed into a beam. The laser-driven source allows protons and ions to be captured at energies significantly above those that pertain in conventional facilities, thus evading the current space-charge limit on the instantaneous dose rate that can be delivered. The laser-hybrid approach, therefore, will allow the vast ``terra incognita'' of the radiobiology that determines the response of tissue to ionising radiation to be studied with protons and light ions using a wide variety of time structures, spectral distributions, and spatial configurations at instantaneous dose rates up to and significantly beyond the ultra-high dose-rate `FLASH' regime. It is proposed that LhARA be developed in two stages. In the first stage, a programme of in vitro radiobiology will be served with proton beams with energies between 10MeV and 15MeV. In stage two, the beam will be accelerated using a fixed-field accelerator (FFA). This will allow experiments to be carried out in vitro and in vivo with proton beam energies of up to 127MeV. In addition, ion beams with energies up to 33.4MeV per nucleon will be available for in vitro and in vivo experiments. This paper presents the conceptual design for LhARA and the R&D programme by which the LhARA consortium seeks to establish the facility.

Nine Recommendations for Decision Aid Implementation from the Clinician Perspective

Background: Shared decision-making (SDM) aims to empower patients to take an active role in their treatment choices, supported by clinicians and patient decision aids (PDAs). The purpose of this study is to explore barriers and possible facilitators to SDM and a PDA in the prostate cancer trajectory. In the process we identify possible actions that organizations and individuals can take to support implementation in practice. Methods: We use the Ottawa Model of Research Use as a framework to determine the barriers and facilitators to SDM and PDAs from the perspective of clinicians. Semi-structured interviews were conducted with urologists (n=4), radiation oncologists (n=3), and oncology nurses (n=2), focusing on the current decision-making process experienced by these stakeholders. Questions included their attitudes towards SDM and PDAs, barriers to implementation and possible strategies to overcome them. Results: Time pressure and patient characteristics were cited as major barriers by 55% of the clinicians we interviewed. Structural factors such as external quotas for certain treatment procedures were also considered as barriers by 44% of the clinicians. Facilitating factors involved organizational changes to em-bed PDAs in the treatment trajectory, training in using PDAs as a tool for SDM, and clinician motivation by disseminating positive clinical outcomes. Our findings also suggest a role for external stakeholders such as healthcare insurers in creating economic incentives to facilitate implementation. Conclusion: Our findings highlight the importance of a multi-faceted implementation strategy to support SDM. While clinician motivation and patient activation are essential, structural/economic barriers may hamper implementation. Action must also be taken at the administrative and policy levels to foster a collaborative environment for SDM and, in the process, for PDAs.

Secondary radiation measurements for particle therapy applications: prompt photons produced by $^{4}$He, $^{12}$C and $^{16}$O ion beams in a PMMA target

Charged particle beams are used in Particle Therapy (PT) to treat oncological patients due to their selective dose deposition in tissues and to their high biological effect in killing cancer cells with respect to photons and electrons used in conventional radiotherapy. Nowadays, protons and carbon ions are used in PT clinical routine but, recently, the interest on the potential application of helium and oxygen beams is growing due to their reduced multiple scattering inside the body and increased linear energy transfer, relative biological effectiveness and oxygen enhancement ratio. The precision of PT demands for online dose monitoring techniques, crucial to improve the quality assurance of treatments. The beam range confined in the irradiated target can be monitored thanks to the neutral or charged secondary radiation emitted by the interactions of hadron beams with matter. Prompt photons are produced by nuclear de-excitation processes and, at present, different dose monitoring and beam range verification techniques based on the prompt {\gamma} detection have been proposed. It is hence of importance to perform the {\gamma} yield measurement in therapeutical-like conditions. In this paper we report the yields of prompt photons produced by the interaction of helium, carbon and oxygen ion beams with a PMMA target. The measurements were performed at the Heidelberg Ion-beam Therapy center (HIT) with beams of different energies. A LYSO scintillator has been used as photon detector. The obtained {\gamma} yields for $^{12}$C ion beams are compared with results from literature, while no other results from $^{4}$He and $^{16}$O beams have been published yet. A discussion on the expected resolution of a slit camera detector is presented, demonstrating the feasibility of a prompt-{\gamma} based monitoring technique for PT treatments using helium, carbon and oxygen ion beams.

Deep learning to achieve clinically applicable segmentation of head and neck anatomy for radiotherapy

Over half a million individuals are diagnosed with head and neck cancer each year worldwide. Radiotherapy is an important curative treatment for this disease, but it requires manual time consuming delineation of radio-sensitive organs at risk (OARs). This planning process can delay treatment, while also introducing inter-operator variability with resulting downstream radiation dose differences. While auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying and achieving expert performance remain. Adopting a deep learning approach, we demonstrate a 3D U-Net architecture that achieves expert-level performance in delineating 21 distinct head and neck OARs commonly segmented in clinical practice. The model was trained on a dataset of 663 deidentified computed tomography (CT) scans acquired in routine clinical practice and with both segmentations taken from clinical practice and segmentations created by experienced radiographers as part of this research, all in accordance with consensus OAR definitions. We demonstrate the model's clinical applicability by assessing its performance on a test set of 21 CT scans from clinical practice, each with the 21 OARs segmented by two independent experts. We also introduce surface Dice similarity coefficient (surface DSC), a new metric for the comparison of organ delineation, to quantify deviation between OAR surface contours rather than volumes, better reflecting the clinical task of correcting errors in the automated organ segmentations. The model's generalisability is then demonstrated on two distinct open source datasets, reflecting different centres and countries to model training. With appropriate validation studies and regulatory approvals, this system could improve the efficiency, consistency, and safety of radiotherapy pathways.

Deep learning for detection and segmentation of artefact and disease instances in gastrointestinal endoscopy

The Endoscopy Computer Vision Challenge (EndoCV) is a crowd-sourcing initiative to address eminent problems in developing reliable computer aided detection and diagnosis endoscopy systems and suggest a pathway for clinical translation of technologies. Whilst endoscopy is a widely used diagnostic and treatment tool for hollow-organs, there are several core challenges often faced by endoscopists, mainly: 1) presence of multi-class artefacts that hinder their visual interpretation, and 2) difficulty in identifying subtle precancerous precursors and cancer abnormalities. Artefacts often affect the robustness of deep learning methods applied to the gastrointestinal tract organs as they can be confused with tissue of interest. EndoCV2020 challenges are designed to address research questions in these remits. In this paper, we present a summary of methods developed by the top 17 teams and provide an objective comparison of state-of-the-art methods and methods designed by the participants for two sub-challenges: i) artefact detection and segmentation (EAD2020), and ii) disease detection and segmentation (EDD2020). Multi-center, multi-organ, multi-class, and multi-modal clinical endoscopy datasets were compiled for both EAD2020 and EDD2020 sub-challenges. The out-of-sample generalization ability of detection algorithms was also evaluated. Whilst most teams focused on accuracy improvements, only a few methods hold credibility for clinical usability. The best performing teams provided solutions to tackle class imbalance, and variabilities in size, origin, modality and occurrences by exploring data augmentation, data fusion, and optimal class thresholding techniques.

Consistency checks of results from a Monte Carlo code intercomparison for emitted electron spectra and energy deposition around a single gold nanoparticle irradiated by X-rays

Organized by the European Radiation Dosimetry Group (EURADOS), a Monte Carlo code intercomparison exercise was conducted where participants simulated the emitted electron spectra and energy deposition around a single gold nanoparticle (GNP) irradiated by X-rays. In the exercise, the participants scored energy imparted in concentric spherical shells around a spherical volume filled with gold or water as well as the spectral distribution of electrons leaving the GNP. Initially, only the ratio of energy deposition with and without GNP was to be reported. During the evaluation of the exercise, however, the data for energy deposition in the presence and absence of the GNP were also requested. A GNP size of 50 nm and 100 nm diameter was considered as well as two different X-ray spectra (50 kVp and 100kVp). This introduced a redundancy that can be used to cross-validate the internal consistency of the simulation results. In this work, evaluation of the reported results is presented in terms of integral quantities that can be benchmarked against values obtained from physical properties of the radiation spectra and materials involved. The impact of different interaction cross-section datasets and their implementation in the different Monte Carlo codes is also discussed.

Prediction of the Position of External Markers Using a Recurrent Neural Network Trained With Unbiased Online Recurrent Optimization for Safe Lung Cancer Radiotherapy

During lung radiotherapy, the position of infrared reflective objects on the chest can be recorded to estimate the tumor location. However, radiotherapy systems have a latency inherent to robot control limitations that impedes the radiation delivery precision. Prediction with online learning of recurrent neural networks (RNN) allows for adaptation to non-stationary respiratory signals, but classical methods such as RTRL and truncated BPTT are respectively slow and biased. This study investigates the capabilities of unbiased online recurrent optimization (UORO) to forecast respiratory motion and enhance safety in lung radiotherapy. We used 9 observation records of the 3D position of 3 external markers on the chest and abdomen of healthy individuals breathing during intervals from 73s to 222s. The sampling frequency was 10Hz, and the amplitudes of the recorded trajectories range from 6mm to 40mm in the superior-inferior direction. We forecast the 3D location of each marker simultaneously with a horizon value between 0.1s and 2.0s, using an RNN trained with UORO. We compare its performance with an RNN trained with RTRL, LMS, and offline linear regression. We provide closed-form expressions for quantities involved in the loss gradient calculation in UORO, thereby making its implementation efficient. Training and cross-validation were performed during the first minute of each sequence. On average over the horizon values considered and the 9 sequences, UORO achieves the lowest root-mean-square (RMS) error and maximum error among the compared algorithms. These errors are respectively equal to 1.3mm and 8.8mm, and the prediction time per time step was lower than 2.8ms (Dell Intel core i9-9900K 3.60 GHz). Linear regression has the lowest RMS error for the horizon values 0.1s and 0.2s, followed by LMS for horizon values between 0.3s and 0.5s, and UORO for horizon values greater than 0.6s.

Bayesian calibration of simulation models: A tutorial and an Australian smoking behaviour model

Simulation models of epidemiological, biological, ecological, and environmental processes are increasingly being calibrated using Bayesian statistics. The Bayesian approach provides simple rules to synthesise multiple data sources and to calculate uncertainty in model output due to uncertainty in the calibration data. As the number of tutorials and studies published grow, the solutions to common difficulties in Bayesian calibration across these fields have become more apparent, and a step-by-step process for successful calibration across all these fields is emerging. We provide a statement of the key steps in a Bayesian calibration, and we outline analyses and approaches to each step that have emerged from one or more of these applied sciences. Thus we present a synthesis of Bayesian calibration methodologies that cut across a number of scientific disciplines. To demonstrate these steps and to provide further detail on the computations involved in Bayesian calibration, we calibrated a compartmental model of tobacco smoking behaviour in Australia. We found that the proportion of a birth cohort estimated to take up smoking before they reach age 20 years in 2016 was at its lowest value since the early 20th century, and that quit rates were at their highest. As a novel outcome, we quantified the rate that ex-smokers switched to reporting as a 'never smoker' when surveyed later in life; a phenomenon that, to our knowledge, has never been quantified using cross-sectional survey data.

OpenKBP-Opt: An international and reproducible evaluation of 76 knowledge-based planning pipelines

We establish an open framework for developing plan optimization models for knowledge-based planning (KBP) in radiotherapy. Our framework includes reference plans for 100 patients with head-and-neck cancer and high-quality dose predictions from 19 KBP models that were developed by different research groups during the OpenKBP Grand Challenge. The dose predictions were input to four optimization models to form 76 unique KBP pipelines that generated 7600 plans. The predictions and plans were compared to the reference plans via: dose score, which is the average mean absolute voxel-by-voxel difference in dose a model achieved; the deviation in dose-volume histogram (DVH) criterion; and the frequency of clinical planning criteria satisfaction. We also performed a theoretical investigation to justify our dose mimicking models. The range in rank order correlation of the dose score between predictions and their KBP pipelines was 0.50 to 0.62, which indicates that the quality of the predictions is generally positively correlated with the quality of the plans. Additionally, compared to the input predictions, the KBP-generated plans performed significantly better (P<0.05; one-sided Wilcoxon test) on 18 of 23 DVH criteria. Similarly, each optimization model generated plans that satisfied a higher percentage of criteria than the reference plans. Lastly, our theoretical investigation demonstrated that the dose mimicking models generated plans that are also optimal for a conventional planning model. This was the largest international effort to date for evaluating the combination of KBP prediction and optimization models. In the interest of reproducibility, our data and code is freely available at https://github.com/ababier/open-kbp-opt.

The Brain Tumor Segmentation (BraTS) Challenge 2023: Focus on Pediatrics (CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs)

Pediatric tumors of the central nervous system are the most common cause of cancer-related death in children. The five-year survival rate for high-grade gliomas in children is less than 20\%. Due to their rarity, the diagnosis of these entities is often delayed, their treatment is mainly based on historic treatment concepts, and clinical trials require multi-institutional collaborations. The MICCAI Brain Tumor Segmentation (BraTS) Challenge is a landmark community benchmark event with a successful history of 12 years of resource creation for the segmentation and analysis of adult glioma. Here we present the CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs 2023 challenge, which represents the first BraTS challenge focused on pediatric brain tumors with data acquired across multiple international consortia dedicated to pediatric neuro-oncology and clinical trials. The BraTS-PEDs 2023 challenge focuses on benchmarking the development of volumentric segmentation algorithms for pediatric brain glioma through standardized quantitative performance evaluation metrics utilized across the BraTS 2023 cluster of challenges. Models gaining knowledge from the BraTS-PEDs multi-parametric structural MRI (mpMRI) training data will be evaluated on separate validation and unseen test mpMRI dataof high-grade pediatric glioma. The CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs 2023 challenge brings together clinicians and AI/imaging scientists to lead to faster development of automated segmentation techniques that could benefit clinical trials, and ultimately the care of children with brain tumors.

The Brain Tumor Segmentation (BraTS) Challenge 2023: Glioma Segmentation in Sub-Saharan Africa Patient Population (BraTS-Africa)

Gliomas are the most common type of primary brain tumors. Although gliomas are relatively rare, they are among the deadliest types of cancer, with a survival rate of less than 2 years after diagnosis. Gliomas are challenging to diagnose, hard to treat and inherently resistant to conventional therapy. Years of extensive research to improve diagnosis and treatment of gliomas have decreased mortality rates across the Global North, while chances of survival among individuals in low- and middle-income countries (LMICs) remain unchanged and are significantly worse in Sub-Saharan Africa (SSA) populations. Long-term survival with glioma is associated with the identification of appropriate pathological features on brain MRI and confirmation by histopathology. Since 2012, the Brain Tumor Segmentation (BraTS) Challenge have evaluated state-of-the-art machine learning methods to detect, characterize, and classify gliomas. However, it is unclear if the state-of-the-art methods can be widely implemented in SSA given the extensive use of lower-quality MRI technology, which produces poor image contrast and resolution and more importantly, the propensity for late presentation of disease at advanced stages as well as the unique characteristics of gliomas in SSA (i.e., suspected higher rates of gliomatosis cerebri). Thus, the BraTS-Africa Challenge provides a unique opportunity to include brain MRI glioma cases from SSA in global efforts through the BraTS Challenge to develop and evaluate computer-aided-diagnostic (CAD) methods for the detection and characterization of glioma in resource-limited settings, where the potential for CAD tools to transform healthcare are more likely.

Construction d'une plate-forme intégrée pour la cartographie de l'exposition des populations aux substances chimiques de l'environnement

L'analyse du lien entre l'environnement et la sant\'e est devenue une pr\'eoccupation majeure de sant\'e publique comme en t\'emoigne l'\'emergence des deux Plans nationaux sant\'e environnement. Pour ce faire, les d\'ecideurs sont confront\'es au besoin de d\'eveloppement d'outils n\'ecessaires \`a l'identification des zones g\'eographiques dans lesquelles une surexposition potentielle \`a des substances toxiques est observ\'ee. L'objectif du projet Syst\`eme d'information g\'eographique (SIG), facteurs de risques environnementaux et d\'ec\`es par cancer (SIGFRIED 1) est de construire une plate-forme de mod\'elisation permettant d'\'evaluer, par une approche spatiale, l'exposition de la population fran\c{c}aise aux substances chimiques et d'en identifier ses d\'eterminants. L'\'evaluation des expositions est r\'ealis\'ee par le biais d'une mod\'elisation multim\'edia probabiliste. Les probl\`emes \'epist\'emologiques li\'es \`a l'absence de donn\'ees sont palli\'es par la mise en {\oe}uvre d'outils utilisant les techniques d'analyse spatiale. Un exemple est fourni sur la r\'egion Nord-Pas-de-Calais et Picardie, pour le cadmium, le nickel et le plomb. Le calcul de l'exposition est r\'ealis\'e sur une dur\'ee de 70 ans sur la base des donn\'ees disponibles autour de l'ann\'ee 2004 sur une maille de 1 km de c\^ot\'e. Par exemple pour le Nord-Pas-de-Calais, les indicateurs permettent de d\'efinir deux zones pour le cadmium et trois zones pour le plomb. Celles-ci sont li\'ees \`a l'historique industriel de la r\'egion : le bassin minier, les activit\'es m\'etallurgiques et l'agglom\'eration lilloise. La contribution des diff\'erentes voies d'exposition varie sensiblement d'un polluant \`a l'autre. Les cartes d'exposition ainsi obtenues permettent d'identifier les zones g\'eographiques dans lesquelles conduire en priorit\'e des \'etudes environnementales de terrains. Le SIG construit constitue la base d'une plate-forme o\`u les donn\'ees d'\'emission \`a la source, de mesures environnementales, d'exposition, puis sanitaires et socio-\'economiques pourront \^etre associ\'ees. -- Analysis of the association between the environment and health has become a major public health concern, as shown by the development of two national environmental health plans. For such an analysis, policy-makers need tools to identify the geographic areas where overexposure to toxic agents may be observed. The objective of the SIGFRIED 1 project is to build a work station for spatial modeling of the exposure of the French population to chemical substances and for identifying the determinants of this exposure. Probabilistic multimedia modeling is used to assess exposure. The epistemological problems associated with the absence of data are overcome by the implementation of tools that apply spatial analysis techniques. An example is furnished for the region of Nord-Pas-de-Calais and Picardie, for cadmium, nickel and lead exposure. The calculation of exposure is performed for duration of 70 years on the basis of data collected around 2004 fora grid of squares 1 km in length. For example, for Nord-Pas-de-Calais, the indicators allow us to define two areas for cadmium and three for lead. They are linked to the region's industrial history: mining basin, metallurgy activities, and the Lille metropolitan area. The contribution of various exposure pathways varied substantially from one pollutant to another. The exposure maps thus obtained allow us to identify the geographic area where environmental studies must be conducted in priority. The GIS thus constructed is the foundation of a workstation where source emission data, environmental exposure measurements, and finally health and socioeconomic measurements can be combined.

The LUX-ZEPLIN (LZ) Experiment

We describe the design and assembly of the LUX-ZEPLIN experiment, a direct detection search for cosmic WIMP dark matter particles. The centerpiece of the experiment is a large liquid xenon time projection chamber sensitive to low energy nuclear recoils. Rejection of backgrounds is enhanced by a Xe skin veto detector and by a liquid scintillator Outer Detector loaded with gadolinium for efficient neutron capture and tagging. LZ is located in the Davis Cavern at the 4850' level of the Sanford Underground Research Facility in Lead, South Dakota, USA. We describe the major subsystems of the experiment and its key design features and requirements.

Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species

Background - The process of generating raw genome sequence data continues to become cheaper, faster, and more accurate. However, assembly of such data into high-quality, finished genome sequences remains challenging. Many genome assembly tools are available, but they differ greatly in terms of their performance (speed, scalability, hardware requirements, acceptance of newer read technologies) and in their final output (composition of assembled sequence). More importantly, it remains largely unclear how to best assess the quality of assembled genome sequences. The Assemblathon competitions are intended to assess current state-of-the-art methods in genome assembly. Results - In Assemblathon 2, we provided a variety of sequence data to be assembled for three vertebrate species (a bird, a fish, and snake). This resulted in a total of 43 submitted assemblies from 21 participating teams. We evaluated these assemblies using a combination of optical map data, Fosmid sequences, and several statistical methods. From over 100 different metrics, we chose ten key measures by which to assess the overall quality of the assemblies. Conclusions - Many current genome assemblers produced useful assemblies, containing a significant representation of their genes, regulatory sequences, and overall genome structure. However, the high degree of variability between the entries suggests that there is still much room for improvement in the field of genome assembly and that approaches which work well in assembling the genome of one species may not necessarily work well for another.

Beyond Low Earth Orbit: Biomonitoring, Artificial Intelligence, and Precision Space Health

Human space exploration beyond low Earth orbit will involve missions of significant distance and duration. To effectively mitigate myriad space health hazards, paradigm shifts in data and space health systems are necessary to enable Earth-independence, rather than Earth-reliance. Promising developments in the fields of artificial intelligence and machine learning for biology and health can address these needs. We propose an appropriately autonomous and intelligent Precision Space Health system that will monitor, aggregate, and assess biomedical statuses; analyze and predict personalized adverse health outcomes; adapt and respond to newly accumulated data; and provide preventive, actionable, and timely insights to individual deep space crew members and iterative decision support to their crew medical officer. Here we present a summary of recommendations from a workshop organized by the National Aeronautics and Space Administration, on future applications of artificial intelligence in space biology and health. In the next decade, biomonitoring technology, biomarker science, spacecraft hardware, intelligent software, and streamlined data management must mature and be woven together into a Precision Space Health system to enable humanity to thrive in deep space.

Beyond Low Earth Orbit: Biological Research, Artificial Intelligence, and Self-Driving Labs

Space biology research aims to understand fundamental effects of spaceflight on organisms, develop foundational knowledge to support deep space exploration, and ultimately bioengineer spacecraft and habitats to stabilize the ecosystem of plants, crops, microbes, animals, and humans for sustained multi-planetary life. To advance these aims, the field leverages experiments, platforms, data, and model organisms from both spaceborne and ground-analog studies. As research is extended beyond low Earth orbit, experiments and platforms must be maximally autonomous, light, agile, and intelligent to expedite knowledge discovery. Here we present a summary of recommendations from a workshop organized by the National Aeronautics and Space Administration on artificial intelligence, machine learning, and modeling applications which offer key solutions toward these space biology challenges. In the next decade, the synthesis of artificial intelligence into the field of space biology will deepen the biological understanding of spaceflight effects, facilitate predictive modeling and analytics, support maximally autonomous and reproducible experiments, and efficiently manage spaceborne data and metadata, all with the goal to enable life to thrive in deep space.

Federated Learning Enables Big Data for Rare Cancer Boundary Detection

Although machine learning (ML) has shown promise in numerous domains, there are concerns about generalizability to out-of-sample data. This is currently addressed by centrally sharing ample, and importantly diverse, data from multiple sites. However, such centralization is challenging to scale (or even not feasible) due to various limitations. Federated ML (FL) provides an alternative to train accurate and generalizable ML models, by only sharing numerical model updates. Here we present findings from the largest FL study to-date, involving data from 71 healthcare institutions across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, utilizing the largest dataset of such patients ever used in the literature (25,256 MRI scans from 6,314 patients). We demonstrate a 33% improvement over a publicly trained model to delineate the surgically targetable tumor, and 23% improvement over the tumor's entire extent. We anticipate our study to: 1) enable more studies in healthcare informed by large and diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further quantitative analyses for glioblastoma via performance optimization of our consensus model for eventual public release, and 3) demonstrate the effectiveness of FL at such scale and task complexity as a paradigm shift for multi-site collaborations, alleviating the need for data sharing.

Segment Anything Model (SAM) Meets Glass: Mirror and Transparent Objects Cannot Be Easily Detected

Meta AI Research has recently released SAM (Segment Anything Model) which is trained on a large segmentation dataset of over 1 billion masks. As a foundation model in the field of computer vision, SAM (Segment Anything Model) has gained attention for its impressive performance in generic object segmentation. Despite its strong capability in a wide range of zero-shot transfer tasks, it remains unknown whether SAM can detect things in challenging setups like transparent objects. In this work, we perform an empirical evaluation of two glass-related challenging scenarios: mirror and transparent objects. We found that SAM often fails to detect the glass in both scenarios, which raises concern for deploying the SAM in safety-critical situations that have various forms of glass.

The Change You Want to See

We live in a dynamic world where things change all the time. Given two images of the same scene, being able to automatically detect the changes in them has practical applications in a variety of domains. In this paper, we tackle the change detection problem with the goal of detecting "object-level" changes in an image pair despite differences in their viewpoint and illumination. To this end, we make the following four contributions: (i) we propose a scalable methodology for obtaining a large-scale change detection training dataset by leveraging existing object segmentation benchmarks; (ii) we introduce a co-attention based novel architecture that is able to implicitly determine correspondences between an image pair and find changes in the form of bounding box predictions; (iii) we contribute four evaluation datasets that cover a variety of domains and transformations, including synthetic image changes, real surveillance images of a 3D scene, and synthetic 3D scenes with camera motion; (iv) we evaluate our model on these four datasets and demonstrate zero-shot and beyond training transformation generalization.

Evaluating GPT as an Adjunct for Radiologic Decision Making: GPT-4 Versus GPT-3.5 in a Breast Imaging Pilot.

Despite rising popularity and performance, studies evaluating the use of large language models for clinical decision support are lacking. Here, we evaluate ChatGPT (Generative Pre-trained Transformer)-3.5 and GPT-4's (OpenAI, San Francisco, California) capacity for clinical decision support in radiology via the identification of appropriate imaging services for two important clinical presentations: breast cancer screening and breast pain. We compared ChatGPT's responses to the ACR Appropriateness Criteria for breast pain and breast cancer screening. Our prompt formats included an open-ended (OE) and a select all that apply (SATA) format. Scoring criteria evaluated whether proposed imaging modalities were in accordance with ACR guidelines. Three replicate entries were conducted for each prompt, and the average of these was used to determine final scores. Both ChatGPT-3.5 and ChatGPT-4 achieved an average OE score of 1.830 (out of 2) for breast cancer screening prompts. ChatGPT-3.5 achieved a SATA average percentage correct of 88.9%, compared with ChatGPT-4's average percentage correct of 98.4% for breast cancer screening prompts. For breast pain, ChatGPT-3.5 achieved an average OE score of 1.125 (out of 2) and a SATA average percentage correct of 58.3%, as compared with an average OE score of 1.666 (out of 2) and a SATA average percentage correct of 77.7%. Our results demonstrate the eventual feasibility of using large language models like ChatGPT for radiologic decision making, with the potential to improve clinical workflow and responsible use of radiology services. More use cases and greater accuracy are necessary to evaluate and implement such tools.

Performance of Generative Large Language Models on Ophthalmology Board Style Questions.

To investigate the ability of generative artificial intelligence models to answer ophthalmology board style questions DESIGN: Experimental study. This study evaluated three large language models (LLMs) with chat interfaces, Bing Chat (Microsoft) and ChatGPT 3.5 and 4.0 (OpenAI), using 250 questions from the Basic Science and Clinical Science (BCSC) Self-Assessment Program (SAP). While ChatGPT is trained on information last updated in 2021, Bing Chat incorporates more recently indexed internet search to generate its answers. Performance was compared to human respondents. Questions were categorized by complexity and patient care phase, and instances of information fabrication or non-logical reasoning were documented. Primary outcome: response accuracy. performance in question subcategories and hallucination frequency. Human respondents had an average accuracy of 72.2%. ChatGPT-3.5 scored the lowest (58.8%), while ChatGPT-4.0 (71.6%) and Bing Chat (71.2%) performed comparably. ChatGPT-4.0 excelled in workup-type questions (OR = 3.89, 95% CI 1.19-14.73, p = 0.03) compared with diagnostic questions, but struggled with image interpretation (OR = 0.14, 95% CI 0.05-0.33, p < 0.01) when compared with single step reasoning questions. Against single step questions, Bing Chat also faced difficulties with image interpretation (OR = 0.18, 95% CI 0.08-0.44, p < 0.01) and multi-step reasoning (OR = 0.30, 95% CI 0.11-0.84, p = 0.02). ChatGPT-3.5 had the highest rate of hallucinations or non-logical reasoning (42.4%), followed by ChatGPT-4.0 (18.0%) and Bing Chat (25.6%). LLMs (particularly ChatGPT-4.0 and Bing Chat) can perform similarly with human respondents answering questions from the BCSC SAP. The frequency of hallucinations and non-logical reasoning suggest room for improvement in the performance of conversational agents in the medical domain.

Genital and Extragenital Lichen Sclerosus et Atrophicus: A Case Series Written Using ChatGPT.

Background Lichen sclerosus et atrophicus (LSEA) is a chronic inflammatory dermatosis of genital and extragenital sites with a prevalence ranging from 9% in prepubertal patients to 50% in postmenopausal patients. Chat generative pre-trained transformer (ChatGPT) is an artificial intelligence tool designed to assist humans based on supervised and reinforcement techniques. In this study, we aimed to evaluate the characteristics of patients with LSEA using ChatGPT. Methods In this retrospective study, we included all patients who presented to the outpatient dermatology department during 2017-2022 at a tertiary care teaching hospital in South India. Information regarding demographic data, characteristics of LSEA, comorbidities, and associated autoimmune disorders was gathered using a medical chart review. Following data analysis and drafting of the manuscript, the utility of ChatGPT-3 and ChatGPT-4 in finalizing the draft was assessed. Results Of 20 patients diagnosed with LSEA, 16 (80%) and four (20%) patients were females and males, respectively. Of them, 50% of female patients had attained menopause. While 65% of patients had genital LSEA, 30% of patients had extragenital LSEA only, and 5% of patients had both genital and extragenital LSEA. Furthermore, four (20%) patients were prepubertal children. Of four male patients, two (50%) were younger than 18 years of age, and one patient was diagnosed with balanitis xerotica obliterans. The commonest associated features in LSEA included joint involvement (30%), hypertension (25%), and anemia (15%). Rare concomitant disorders included psoriasis, asthma, and basal cell carcinoma over the nose. Conclusions LSEA may be confused with other various dermatoses, such as morphea, vitiligo, and lichen planus. A high index of suspicion is required, especially in children, to diagnose it early and intervene to prevent further complications. Its relationship with autoimmune disorders and comorbidities warrants further large-scale studies. ChatGPT was unreliable in the literature search due to the provision of non-existent citations. ChatGPT-4 was better than ChatGPT-3 since it provided few true publications. ChatGPT was used in this study to summarize the articles identified by the authors during the literature search and to correct grammatical errors in the final draft of the manuscript.

Performance of ChatGPT on dermatology Specialty Certificate Examination multiple choice questions.

ChatGPT is a large language model trained on increasingly large datasets by OpenAI to perform language-based tasks. It is capable of answering multiple-choice questions, such as those posed by the dermatology SCE examination. We asked two iterations of ChatGPT: ChatGPT-3.5 and ChatGPT-4 84 multiple-choice sample questions from the sample dermatology SCE question bank. ChatGPT-3.5 achieved an overall score of 63.1%, and ChatGPT-4 scored 90.5% (a significant improvement in performance (p<0.001)). The typical pass mark for the dermatology SCE is 70-72%. ChatGPT-4 is therefore capable of answering clinical questions and achieving a passing grade in these sample questions. There are many possible educational and clinical implications for increasingly advanced artificial intelligence (AI) and its use in medicine, including in the diagnosis of dermatological conditions. Such advances should be embraced provided that patient safety is a core tenet, and the limitations of AI in the nuances of complex clinical cases are recognised.

Chat Generative Pretrained Transformer Fails the Multiple-Choice American College of Gastroenterology Self-Assessment Test.

Chat Generative Pretrained Transformer (ChatGPT) is a natural language processing model that generates human-like text. ChatGPT-3 and ChatGPT-4 were used to answer the 2022 and 2021 American College of Gastroenterology self-assessment tests. The exact questions were inputted in both versions of ChatGPT. A score of 70% or higher was required to pass the assessment. Overall, ChatGPT-3 scored 65.1% on 455 included questions and GPT-4 scored 62.4%. ChatGPT did not pass the American College of Gastroenterology self-assessment test. We do not recommend its use for medical education in gastroenterology in its current form.