Differential effects of demographics and risk factors on the nonlinear orthotropic mechanical properties of human femoropopliteal arteries.

Understanding how demographics and risk factors differentially affect the nonlinear orthotropic mechanical properties of human femoropopliteal arteries (FPAs) is critical for improving computational models of device-artery interactions. However, directly assessing these effects is challenging due to their intricate interrelationships and multifaceted impact on arterial mechanics. This study aimed to quantify the independent and combined effects of aging, diabetes, obesity, hypertension, and coronary artery disease on the mechanical behavior of human FPAs using machine learning (ML). FPAs from 450 tissue donors 12-99 years old (average age 51 ± 18 years, 66% male) were evaluated using multi-ratio planar biaxial extension. ML models, optimized through hyperparameter tuning and cross-validation, were used to describe subject-specific nonlinear orthotropic arterial properties based on demographics and risk factors. Age accounted for 60% of the variability in arterial mechanics and had a stronger influence longitudinally than circumferentially. The presence of diabetes and coronary artery disease each added 13 and 11 years to vascular age circumferentially but < 3 years longitudinally. Obesity and hypertension each added 4 years to vascular age circumferentially and less than 3 years longitudinally. Compound effects of diabetes, hypertension, and obesity aged the artery more than 21 years circumferentially and 7 years longitudinally. These findings highlight the differential impact of risk factors on orthotropic arterial mechanics and demonstrate the potential of our approach for predicting subject-specific vascular properties. Incorporating these findings into computational models can enhance the accuracy of device-artery interaction simulations by accounting for individual-specific vascular characteristics.