LightMedSeg-ISLES: Stroke Lesion Segmentation with 81x Fewer Parameters than nnU-Net

2026-09-08

AI Summary

本文提出LightMedSeg-ISLES，用比nnU-Net少81倍的参数实现脑卒中病灶分割，同时保持相近的Dice系数并提高病灶F1分数。

Abstract

Large networks and ensembles often lead medical image segmentation challenges, but their storage and inference demands complicate deployment. We present LightMedSeg-ISLES, a 1.26-million-parameter pipeline for T1-weighted stroke lesion segmentation in ISLES'26. On a 146-case held-out cohort, flip test-time augmentation produces 0.618 mean Dice and 0.599 lesion-wise F1. A 102.35-million-parameter nnU-Net ResEnc-L produces 0.634 Dice and 0.544 lesion-wise F1 after size filtering. LightMedSeg therefore retains 97.5\% of nnU-Net's Dice with 81.4$\times$ fewer parameters while improving lesion-wise F1 by 0.055. Its four-pass TTA operating point requires 4.7$\times$ fewer FLOPs per standardized patch than nnU-Net. It also slightly exceeds filtered UNETR++ and nnFormer. Longer training and stronger augmentation add 0.0358 Dice without increasing capacity, establishing a strong single-checkpoint alternative to much larger models.

Problem

• Medical Image Segmentation
• Stroke Lesion
• Model Deployment
• Parameter Efficiency

Innovation

• LightMedSeg-ISLES
• fewer parameters
• stroke lesion segmentation
• test-time augmentation
• FLOPs reduction