What question did this study set out to answer?

Evaluate automated, topology-aware segmentation and anatomical classification of the Circle of Willis across computed tomography angiography and magnetic resonance angiography scans.

July 23, 2026NEJM AIOpen Access

The TopCoW Challenge — Topology-Aware Circle of Willis Segmentation for CT and MR Angiography

View Full Paper

Ask AI

Bookmark

Authors

KYKaiyuan YangUniversity of Zurich FMFabio MusioUniversity of Zurich YMYihui MaFirst Affiliated Hospital of Zhengzhou University

Discussion

Member takes

Implication

Benchmark study demonstrates high automated segmentation accuracy for brain arterial networks in multi-center angiographic scans, highlighting clinical potential for neurovascular disease assessment.

Key Points

Evaluate automated, topology-aware segmentation and anatomical classification of the Circle of Willis across computed tomography angiography and magnetic resonance angiography scans.
Built a benchmark dataset of 200 paired CT angiography and MR angiography patient scans with virtual reality-assisted, voxel-level annotations for 13 vessel components.
Evaluated models from over 250 registered participants using internal and external test sets totaling 226 scans across more than five centers.
Top-performing teams achieved over 90% Dice scores for vessel component segmentation across nearly all test sets.
Algorithms reached over 80% F1 scores for detecting key vessel components and over 70% balanced accuracy for classifying anatomical variants.
Best-performing models demonstrated diagnostic utility in identifying fetal-type posterior cerebral artery configurations and localizing aneurysms.

Cite This Study

Yang et al. (2026) studied this question.

synapsesocial.com/papers/6a7e15cbfe8427622e9533a6 https://doi.org/10.1056/aidbp2500994

View Full Paper

Ask AI

Bookmark

July 23, 2026NEJM AIOpen Access

The TopCoW Challenge — Topology-Aware Circle of Willis Segmentation for CT and MR Angiography

View Full Paper

Ask AI

Bookmark

Authors

KYKaiyuan YangUniversity of Zurich FMFabio MusioUniversity of Zurich YMYihui MaFirst Affiliated Hospital of Zhengzhou University

Discussion

Member takes

Implication

Benchmark study demonstrates high automated segmentation accuracy for brain arterial networks in multi-center angiographic scans, highlighting clinical potential for neurovascular disease assessment.

Key Points

Evaluate automated, topology-aware segmentation and anatomical classification of the Circle of Willis across computed tomography angiography and magnetic resonance angiography scans.
Built a benchmark dataset of 200 paired CT angiography and MR angiography patient scans with virtual reality-assisted, voxel-level annotations for 13 vessel components.
Evaluated models from over 250 registered participants using internal and external test sets totaling 226 scans across more than five centers.
Top-performing teams achieved over 90% Dice scores for vessel component segmentation across nearly all test sets.
Algorithms reached over 80% F1 scores for detecting key vessel components and over 70% balanced accuracy for classifying anatomical variants.
Best-performing models demonstrated diagnostic utility in identifying fetal-type posterior cerebral artery configurations and localizing aneurysms.

Cite This Study

Yang et al. (2026) studied this question.

synapsesocial.com/papers/6a7e15cbfe8427622e9533a6 https://doi.org/10.1056/aidbp2500994

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark