Artificial Intelligence and Material Modeling in Biomechanics

This research focuses on the development and implementation of advanced material models as User-Defined Elements (UEL) in Abaqus, based on Finite Strain Theory and incorporating Hyperelasticity, Plasticity, Nonlinear Hardening, Phase-Field Fracture, Fatigue, Chemical Reactions, and Anisotropy. These models are employed to simulate the interaction between coronary stents and arterial tissue during crimping, expansion with balloon, and long-term fatigue induced by continuous blood pulsation.

To enhance and reduce the computational effort of these simulations, physics-based deep learning models are developed within the TensorFlow–Python framework. In addition, various stent families are designed using Solidworks and analyzed based on KPIs to investigate how geometrical design influences mechanical performance and durability.

The long-term goal of this work is the development of a Smart Stent, an innovative medical device that combines advanced materials, computational modelling, biosensors, and artificial intelligence. Unlike conventional stents, the Smart Stent aims not only to open blocked arteries but also to monitor its own performance and the patient’s healing process in real time.

Equipped with miniature MEMS-based blood pressure and flow sensors and biosensors capable of detecting key cellular activity (such as smooth muscle cells and macrophages), the device will continuously collect physiological data. This information will be transmitted wirelessly to external devices (e.g., smartphones or medical units), where dedicated microchips optimized for physics-based deep learning will process the data in real time.

By merging structural monitoring with biological feedback, the Smart Stent represents a paradigm shift in cardiovascular treatment, a stent that not only treats but also provides proactive, patient-specific insights into recovery and long-term vascular health.