Identifying abnormal patterns of wall shear stress in mild-to-moderate aortic dilation and tricuspid valves using...

Area of Science:

• Cardiovascular Imaging and Hemodynamics
• Biomedical Engineering
• Statistical Analysis in Medical Research

Background:

• Aortic wall shear stress (WSS) mapping aids in visualizing and quantifying WSS patterns.
• Investigating regional WSS differences requires statistical methods that account for multiple testing.
• Previous studies lacked robust statistical approaches for patch-wise WSS comparisons.

Purpose of the Study:

• To evaluate the efficacy of permutation tests for analyzing regional WSS differences.
• To compare permutation tests with common cardiovascular statistical methods.
• To assess WSS patterns in patients with mild-to-moderate aortic dilation and tricuspid valves.

Main Methods:

• Utilized 4D Flow MRI-derived WSS parameters from 46 patients with aortic dilation and 51 controls.
• Mapped 3D WSS data onto a shared geometry, dividing the aorta into 40x40 surface patches.
• Assessed permutation tests with TFCE against Student's t-test and Wilcoxon rank sum test using synthetic and real data.

Main Results:

• Permutation tests demonstrated lower false positive rates than uncorrected t-tests and Wilcoxon tests.
• Permutation tests showed lower false negative rates than Bonferroni-corrected t-tests.
• TFCE-based permutation tests revealed lower peak WSS, higher OSI, and higher diastolic WSS in patients with aortic dilation.

Conclusions:

• Permutation tests are suitable for local WSS analysis in cohort comparisons.
• These tests effectively correct for family-wise error rate while considering spatial correlations.
• Permutation tests enhance the statistical rigor of WSS analysis in cardiovascular research.