README.md

metadata

license: cc-by-nc-sa-4.0

CTransPath-based ABMIL models for metastasis detection

These are weakly-supervised, attention-based multiple instance learning models for binary metastasis detection (normal versus metastasis). The models were trained on the CAMELYON16 dataset using CTransPath embeddings.

Data

• Training set consisted of 243 whole slide images (WSIs).
  • 143 negative
  • 100 positive
    • 52 macrometastases
    • 48 micrometastases
• Validation set consisted of 27 WSIs.
  • 16 negative
  • 11 positive
    • 6 macrometastases
    • 5 micrometastases
• Test set consisted of 129 WSIs.
  • 80 negative
  • 49 positive
    • 22 macrometastases
    • 27 micrometastases

Evaluation

Below are the classification results on the test set.

Seed	Sensitivity	Specificity	BA	Precision	F1
0	0.776	0.912	0.844	0.844	0.809
1	0.857	0.912	0.885	0.857	0.857
2	0.735	0.963	0.849	0.923	0.818
3	0.776	0.925	0.850	0.864	0.817
4	0.776	0.950	0.863	0.905	0.835

How to reuse the model

The model expects 128 x 128 micrometer patches, embedded with the CTransPath model.

import torch
from abmil import AttentionMILModel

model = AttentionMILModel(in_features=768, L=512, D=384, num_classes=2, gated_attention=True)
model.eval()
state_dict = torch.load("seed1/model_best.pt", map_location="cpu", weights_only=True)
model.load_state_dict(state_dict)

# Load a bag of features
bag = torch.ones(1000, 768)
with torch.inference_mode():
    logits, attention = model(bag)