CENet: Context Enhancement Network for Medical Image Segmentation

Official pretrained weights for CENet (MICCAI 2025, early accept, top 9%), a 2D segmentation network for abdominal CT, cardiac MRI and dermoscopy images.

CENet uses a PVTv2-B2 encoder and adds two modules:

• DSEB (Dual Selective Enhancement Block) in the skip connections, which sharpens boundaries and helps with small organs.
• CFAM (Context Feature Attention Module) in the decoder, a multi-scale block that keeps spatial detail and reduces redundant features.

Quick start

You do not need to clone anything. PyTorch Hub fetches the model code from GitHub, and the weights are downloaded from this repository.

pip install torch torchvision timm==1.0.16 monai==1.4.0 "numpy<2" scipy SimpleITK huggingface_hub

Segment your own images

The predictor applies the same preprocessing as in training and returns a mask with the size of your input.

import torch

# Skin lesion (dermoscopy image)
skin = torch.hub.load("xmindflow/cenet", "predictor", dataset="ham10000", trust_repo=True)
mask = skin.predict("lesion.jpg")            # (H, W): 0 = background, 1 = lesion

# Abdominal CT (NIfTI volume, processed slice by slice)
ct = torch.hub.load("xmindflow/cenet", "predictor", dataset="synapse", trust_repo=True)
mask = ct.predict("abdomen_ct.nii.gz")       # (D, H, W)
print(ct.classes)                            # label names, index = label value

predict accepts a file path (.png, .jpg, .npy, .nii, .nii.gz), a PIL image or a NumPy array. It runs on the GPU if one is available.

Use the model directly

model = torch.hub.load("xmindflow/cenet", "cenet", dataset="synapse", trust_repo=True)
x = torch.randn(1, 1, 224, 224)              # one preprocessed slice
with torch.no_grad():
    logits = model(x)                        # (1, 9, 224, 224)
pred = logits.argmax(dim=1)

The original model reported in the paper for Synapse is available as torch.hub.load("xmindflow/cenet", "cenet_org", trust_repo=True).

Models

dataset	Task	Input	Labels	File
synapse	Abdominal CT, 8 organs	1 × 224 × 224	0 background, 1 aorta, 2 gallbladder, 3 left kidney, 4 right kidney, 5 liver, 6 pancreas, 7 spleen, 8 stomach	synapse_2418962.pth
acdc	Cardiac MRI	1 × 224 × 224	0 background, 1 right ventricle, 2 myocardium, 3 left ventricle	acdc_2418732.pth
ham10000	Skin lesion	3 × 224 × 224	0 background, 1 lesion	ham10k_2418602.pth
ph2	Skin lesion	3 × 224 × 224	0 background, 1 lesion	ph2_2418571.pth
cenet_org	Abdominal CT, original paper model	1 × 224 × 224	same as synapse	synapse_cenet_org_2352272.pth

Each file is a plain PyTorch state_dict (about 130 MB).

Preprocessing

The predictor does this for you. If you call the model directly, prepare inputs the same way:

• Synapse (CT): axial slices. Clip Hounsfield units to [-125, 275], scale to [0, 1], resize to 224 × 224 (cubic), then normalize with mean 0.5 and std 0.5.
• ACDC (MRI): short-axis slices. Scale intensities to [0, 1], resize to 224 × 224 (cubic).
• HAM10000 / PH2 (dermoscopy): RGB image. Resize to 224 × 224 (bilinear), divide by 255, then min-max normalize the image to [0, 1].

Predictions are made at 224 × 224 and resized back with nearest-neighbor interpolation.

Results

To reproduce the paper's numbers, use the evaluation scripts in the GitHub repository. The README explains how to get and prepare each dataset.

Download the weights only

hf download afshinbigboy/CENet --include "*.pth" --local-dir weights

or in Python:

from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="afshinbigboy/CENet", filename="synapse_2418962.pth")

Intended use and limitations

These models are for research only. They are not a medical device and must not be used for diagnosis or treatment decisions. Each model was trained on a single public dataset and may perform poorly on images from other scanners, protocols or populations.

License

The code and weights are released under the MIT license. The training datasets have their own terms (for example, HAM10000 is CC BY-NC 4.0), so please also respect the license of each dataset.

Citation

If you use CENet, please cite:

@InProceedings{BozAfs_CENet_MICCAI2025,
  author    = {Bozorgpour, Afshin and Kolahi, Sina Ghorbani and Azad, Reza and Hacihaliloglu, Ilker and Merhof, Dorit},
  title     = {{CENet}: Context Enhancement Network for Medical Image Segmentation},
  booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2025},
  series    = {Lecture Notes in Computer Science},
  volume    = {15960},
  pages     = {120--129},
  year      = {2025},
  publisher = {Springer Nature Switzerland},
  address   = {Cham},
  doi       = {10.1007/978-3-032-04927-8_12}
}