The helical morphology of Type B aortic dissections (TBAD) represents a potentially important geometric biomarker that may influence dissection progression. While three-dimensional surface-based quantification methods provide accurate TBAD helicity assessment, their clinical adoption remains limited by significant processing time. We developed and validated a clinically practical centerline-based helicity quantification method using routine imaging software (TeraRecon) against an extensively validated surface-based method (SimVascular). In 87 TBAD patients, we semi-automatically extracted aortic, true lumen, and branch vessel centerlines from CT imaging. Helical parameters, including true lumen helical angle and peak helical twist, were computed relative to a standardized anatomical reference, enabling classification of patients into four distinct helicity categories: left-chiral, right-chiral, non-helical, and mixed-chiral patterns. The centerline method demonstrated 92% classification accuracy with excellent agreement with surface-based measurements (Cohen’s 𝜅 = 0.88, p < 0.001). Wilcoxon signed-rank tests revealed a median difference of −0.4° (z = −1.08, p = 0.28), indicating no statistically significant systematic bias between methods. This centerline approach we have developed provides clinically feasible TBAD helicity classification while maintaining excellent agreement with the gold-standard surface-based method. This technique can integrate seamlessly with existing clinical workflows, enabling practical assessment of TBAD helical morphology for enhanced risk stratification and personalized treatment planning.
Keywords:
Aortic dissection; Helical morphology; Geometric biomarker; Computed tomography