Robust Brain Extraction Tool for Non-enhanced CT and CT Angiography: Cta-bet.

Purpose To develop and evaluate a deep learning-based brain extraction model, CTA-BET, capable of providing accurate brain segmentation for CT angiography (CTA) and noncontrast-enhanced (NCCT) images. Materials and Methods In this retrospective study, CTA-BET was trained using CTA data from multi-institutional cohorts ( = 100 patients) and validated on an external CTA dataset ( = 50 patients). NCCT validation was performed using the publicly available CQ500 dataset ( = 132 patients). The model's performance was compared with five benchmark non-commercial brain extraction tools. Dice score, Hausdorff distance, and z-score normalized histograms were used to evaluate segmentation performance. Results The CTA-BET model outperformed all benchmark models, achieving a mean Dice score of 0.99 (95% CI: 0.99, 0.99) on CTA data ( < .001 for all comparisons) and 0.98 (95% CI: 0.98, 0.99) on NCCT images ( < .001 for all comparisons). In terms of Hausdorff distance, CTA-BET demonstrated higher performance compared with other benchmark tools on CTA images ( < .001 for all comparisons. Conclusion CTA-BET outperformed benchmark brain extraction tools on both CTA and NCCT images, providing a robust and accurate solution that could enhance automated imaging analysis in clinical and research settings. ©RSNA, 2025.

PMID: 41147859