AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Learning Topological Interactions for Multi-Class Medical Image Segmentation

Saumya Gupta, Xiaoling Hu, James Kaan, Michael Jin, Mutshipay Mpoy, Katherine Chung, Gagandeep Singh, Mary Saltz, Tahsin Kurc, Joel Saltz, Apostolos Tassiopoulos, Prateek Prasanna and 1 more

ECCV 2022 · Detection and segmentation › Open-set and few-shot detection · approach: Interactive and referring segmentation · 46 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add topological interaction loss L_ti (ablation)nnUNet · IVUS lumenDice: 0.893 -> 0.949Improved
add topological interaction loss L_ti (ablation)nnUNet · IVUS lumen% violations: 2.708 -> 0.157Improved
add topological interaction loss L_ti (ablation)nnUNet · IVUS mediaDice: 0.856 -> 0.910Improved
add topological interaction loss L_tinnUNet · Aorta lumen% violations: 5.424 -> 3.526Improved
add topological interaction loss L_tinnUNet · Aorta wallDice: 0.741 -> 0.758Improved
add topological interaction loss L_tinnUNet · Multi-Atlas kidneyDice: 0.931 -> 0.936No clear effect
add topological interaction loss L_tinnUNet · Multi-Atlas spleenDice: 0.950 -> 0.960Improved
add topological interaction loss L_tinnUNet · SegTHOR tracheaDice: 0.910 -> 0.929Improved
use MSE instead of CE as pixel loss in L_ti (ablation)Ours · IVUS lumenDice: 0.949 -> 0.915Hurt
use Dice loss instead of CE as pixel loss in L_ti (ablation)Ours · IVUS lumenDice: 0.949 -> 0.937Hurt
lower topological loss weight to 5.0e-5 (ablation)Ours · IVUS lumenDice: 0.949 -> 0.913Hurt
raise topological loss weight to 2.0e-4 (ablation)Ours · IVUS lumenDice: 0.949 -> 0.938Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy