Hyperelastic constitutive model discovery for veins

Status: Manuscript in preparation.

Veins are often exposed to altered mechanical loading in clinical settings such as vein grafting, arteriovenous fistula creation, and tissue-engineered vascular graft implantation. However, venous tissue behaves differently from arterial tissue, and existing constitutive models can struggle to fit circumferential and axial stress-stretch responses while generalizing across vein types.

In this project, we developed a constitutive artificial neural network (CANN) framework to discover a compact, interpretable hyperelastic constitutive model from biaxial mechanical data across three vein types: saphenous vein, jugular vein, and inferior vena cava. The framework expands the candidate function library with structurally motivated terms for progressive fiber recruitment and fiber-family interactions, then uses Sobol sensitivity analysis, correlation-based pruning, and response-level family collapse to reduce the discovered basis to an eight-term strain-energy density function. The resulting model achieved high fitting accuracy across all three vein types and provides a candidate constitutive form for further validation across larger, independent cohorts.

Advisors and collaborators: Dr. Marisa Bazzi, Gavin Mays, Prof. Ellen Kuhl, Prof. Jay Humphrey, and Prof. Alison Marsden.