Med-Real2Sim: Non-Invasive Medical Digital Twins using Physics-Informed Self-Supervised Learning and Virtual Patient Breathing Simulator using VR

This paper presents two progressive phases of the med-real2sim project — an integrated framework for developing non-invasive medical digital twins in healthcare. phase 1 introduces a cardiac digital twin system that leverages physics-informed neural networks (pinns) and self-supervised learning to model patient cardiac profiles, simulate pressure–volume dynamics, and predict health outcomes from echocardiogram data. phase 2 extends the concept to respiratory physiology, presenting a virtual patient breathing simulator that employs webxr, webgl, and webassembly-based physics simulation to render immersive, browser-accessible 3d models of lung and diaphragm mechanics. together, the two phases demonstrate how digital twin technology, when combined with ai and physics-based modeling, can significantly improve diagnostic accuracy, personalize treatment, and enhance medical education through interactive virtual simulation.