How We Won the ISLES'24 Challenge by Preprocessing

2025-05-23

AI Summary

This work addresses the challenging task of predicting follow-up MRI lesion progression in acute ischemic stroke patients using only admission non-contrast CT scans. To this end, we propose a CT-specific preprocessing and segmentation framework. Methodologically, we introduce two key components: (1) systematic validation of skull stripping and adaptive CT intensity window truncation—demonstrating their critical impact on segmentation performance in scenarios where lesions are initially invisible on CT; and (2) an end-to-end nnU-Net-based model leveraging large residual connections. Evaluated on the ISLES’24 challenge test set, our method achieves a mean Dice score of 28.5 (std 21.27), ranking first and substantially outperforming existing approaches. This represents the first clinically viable solution for longitudinal lesion assessment in the absence of follow-up imaging, offering a practical pathway for dynamic lesion monitoring under real-world constraints.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance

Abstract

Stroke is among the top three causes of death worldwide, and accurate identification of stroke lesion boundaries is critical for diagnosis and treatment. Supervised deep learning methods have emerged as the leading solution for stroke lesion segmentation but require large, diverse, and annotated datasets. The ISLES'24 challenge addresses this need by providing longitudinal stroke imaging data, including CT scans taken on arrival to the hospital and follow-up MRI taken 2-9 days from initial arrival, with annotations derived from follow-up MRI. Importantly, models submitted to the ISLES'24 challenge are evaluated using only CT inputs, requiring prediction of lesion progression that may not be visible in CT scans for segmentation. Our winning solution shows that a carefully designed preprocessing pipeline including deep-learning-based skull stripping and custom intensity windowing is beneficial for accurate segmentation. Combined with a standard large residual nnU-Net architecture for segmentation, this approach achieves a mean test Dice of 28.5 with a standard deviation of 21.27.

Problem

• Accurate stroke lesion segmentation from CT scans
• Overcoming limited visibility of lesions in CT
• Preprocessing pipeline for improved segmentation accuracy

Innovation

• Deep-learning-based skull stripping preprocessing
• Custom intensity windowing for CT scans
• Large residual nnU-Net segmentation architecture