Medical Image Segmentation on Medical Segmentation Benchmark Suite (LIDC, AMOS, MSD) (test)
0.752Avg Dice ScoreBoltzFormer+FocalL
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| BoltzFormer+FocalLDecoder=BoltzFormer, Backbone=FocalL, Prompting protocol=Customized text prompts2025.03 | 0.752 | 0.754 | 0.927 | 0.902 | 0.702 | 0.64 | 0.67 | 0.667 | |
| BoltzFormer+Heira-BPDecoder=BoltzFormer, Backbone=Heira-BP, Prompting protocol=Customized text prompts2025.03 | 0.745 | 0.736 | 0.92 | 0.893 | 0.714 | 0.61 | 0.667 | 0.667 | |
| SEEM+Heira-BP (text)Decoder=SEEM, Backbone=Heira-BP, Prompting protocol=Customized text prompts2025.03 | 0.738 | 0.729 | 0.92 | 0.893 | 0.693 | 0.646 | 0.657 | 0.633 | |
| BoltzFormer+Heira-SDecoder=BoltzFormer, Backbone=Heira-S, Prompting protocol=Customized text prompts2025.03 | 0.738 | 0.733 | 0.913 | 0.887 | 0.704 | 0.637 | 0.644 | 0.652 | |
| BiomedParseType=Pre-trained biomedical foundation model2025.03 | 0.73 | 0.738 | 0.919 | 0.876 | 0.661 | 0.602 | 0.648 | 0.664 | |
| SEEM+Heira-S (text)Decoder=SEEM, Backbone=Heira-S, Prompting protocol=Customized text prompts2025.03 | 0.715 | 0.721 | 0.911 | 0.882 | 0.659 | 0.614 | 0.639 | 0.58 | |
| nnU-NetTraining protocol=Expert model trained for each class in each dataset separately2025.03 | 0.673 | 0.648 | 0.85 | 0.852 | 0.602 | 0.524 | 0.613 | 0.624 | |
| SAM+Heira-S (text)Decoder=SAM, Backbone=Heira-S, Prompting protocol=Customized text prompts2025.03 | 0.67 | 0.671 | 0.884 | 0.839 | 0.616 | 0.551 | 0.612 | 0.52 | |
| SAM2+Heira-BP (text)Decoder=SAM2, Backbone=Heira-BP, Prompting protocol=Customized text prompts2025.03 | 0.661 | 0.632 | 0.881 | 0.82 | 0.642 | 0.529 | 0.601 | 0.516 | |
| SAM2+Heira-S (text)Decoder=SAM2, Backbone=Heira-S, Prompting protocol=Customized text prompts2025.03 | 0.656 | 0.654 | 0.882 | 0.826 | 0.598 | 0.528 | 0.598 | 0.506 | |
| SAM+Heira-BP (text)Decoder=SAM, Backbone=Heira-BP, Prompting protocol=Customized text prompts2025.03 | 0.649 | 0.628 | 0.867 | 0.812 | 0.572 | 0.514 | 0.608 | 0.54 |