Deep learning segments ischemic stroke on CT scans across hospitals
Ischemic Stroke Segmentation and Net Water Uptake Quantification on Multicenter Non-Contrast CT Using Supervised Target-Domain Adaptation
Computer Vision and Pattern Recognition
Summary
Identifying damaged brain areas after a stroke on standard CT scans is usually done by experts and can be slow. This paper shows a new computer program that learns to find stroke damage automatically on CT images from different hospitals. The program also measures how much water has entered the damaged brain tissue, a sign of stroke severity, without needing advanced scans. The results varied depending on the hospital’s data but showed promise for helping doctors assess stroke patients faster.
What this means in practice
- •For radiology departments: Automatically outline stroke areas and measure tissue water uptake in CT scans from multiple hospitals to assist diagnosis without extra imaging.
- •For clinical trial teams: Use automated stroke segmentation and water uptake quantification from standard CT to collect biomarkers consistently across trial centers.
- •For emergency medical device developers: Create software tools that provide near-instant stroke damage maps and water uptake estimates from baseline CT scans for hospital use.$Commercial implications: Enables stroke imaging analysis software to be marketed to hospitals for faster stroke assessment without advanced imaging.
Authors
Linus Britt, Maximilian Nielsen, Susan Klapproth, Andre Kemmling, Michael H. Lev, Gabriel Broocks, Rene Werner, Thilo Sentker
Abstract
Objectives: Quantitative assessment of infarct hypodensity on non-contrast computed tomography (NCCT), including net water uptake (NWU), requires manual or semi-manual lesion delineation, often guided by CT perfusion or diffusion-weighted MRI, limiting clinical applicability. Automated segmentation on NCCT could enable efficient biomarker extraction such as NWU but remains challenging across heterogeneous multicenter data. This study aimed to develop and externally test a domain-aware deep learning framework for ischemic stroke segmentation on NCCT and assess its suitability for NWU quantification. Materials & Methods: In this retrospective multicenter study of 801 patients from four datasets, an nnU-Net-based model was trained on NCCT scans from the University Medical Center Hamburg-Eppendorf and the Acute Ischemic Stroke Dataset. To adapt to new domains, the model was fine-tuned on target-domain subsets from Boston (n=11) and ISLES (n=75), with evaluation on held-out cases not used for fine-tuning. Automated segmentations and NWU values were compared with expert references. Results: For lesions $\geq$ 30 mL, median Dice was 0.68 (Boston) and 0.56 (ISLES). Including smaller lesions, which predominated in ISLES, median Dice was 0.54 (interquartile range [IQR] 0.30-0.70) for acute lesion segmentation (Boston dataset) and 0.20 (IQR 0.03-0.41) for NCCT lesion segmentations when compared to post-treatment infarct (primary target of the ISLES challenge). Automated NWU mean absolute error was 1.37 percentage points (SD 1.61, Boston). Conclusion: Target-domain adaptation supported NCCT-only infarct segmentation across heterogeneous external cohorts, although performance varied across domains. The approach enabled low-error NWU quantification from baseline NCCT without advanced imaging, supporting further prospective clinical evaluation.