Translational XR and AI Technologies with Dynamic Soft Robotic Hearts for Immersive Clinical Training and Future Intraoperative Guidance in Interventional Cardiology - Project Summary This K25 project aims to revolutionize structural heart intervention training and procedural guidance by developing an advanced extended reality (XR) platform integrated with AI, soft robotics, and fluoroscopy- compatible systems to address critical clinical challenges in procedures like atrial septal defect (ASD) closure, left atrial appendage (LAA) closure, and transcatheter mitral valve repair (TMVR). Current 2D fluoroscopy lacks depth perception, increases radiation exposure, and poses nephrotoxicity risks from contrast agents, compromising patient safety. The proposed research builds on prior work developing a real-time 3D catheter tracking XR system with submillimeter accuracy, introducing a clinically impactful solution to enhance procedural precision and realism. Structured in two phases, the project addresses these limitations through innovative technology. Phase I (Aims 1 & 2) develops an off-cathlab, radiation-free training system. Aim 1 creates an XR platform with a flexible haptic device for realistic catheter manipulation within patient-specific 3D cardiac models, simulating anatomical resistance to reduce training-related radiation exposure. Aim 2 quantifies user learning behaviors and psychomotor strategies using motion tracking, catheter trajectories, and physiological data (e.g., GSR, HR, HRV), comparing XR to 2D training to optimize skill acquisition and reduce cognitive load, critical for clinical proficiency. Phase II (Aims 3 & 4) advances to an on-cathlab, fluoroscopy-integrated system. Aim 3 designs a soft robotic heart (SRH) model using compliant materials with pneumatic actuation to replicate cardiac dynamics, integrated with AI-driven catheter pose estimation from biplane and monoplane fluoroscopy for realistic cath- lab training. Aim 4 evaluates clinical readiness with ~25 interventional cardiologists and fellows, assessing navigation accuracy, task efficiency, and perceived realism to align with clinical workflows. The platform leverages my engineering expertise in soft robotics, AI, and XR to improve clinical outcomes in interventional cardiology by enhancing spatial awareness, reducing reliance on fluoroscopy, and minimizing procedural complications. By integrating haptic feedback, AI-enhanced imaging, and dynamic SRH models, the research aims to lower learning curves, enhance skill retention, and improve procedural safety. The anticipated clinical impact includes reduced complication rates, enhanced patient safety, and improved outcomes in structural heart interventions. This work establishes a scalable, clinically adaptable training ecosystem, paving the way for future intraoperative XR guidance systems and advancing smart minimally invasive technologies (SMIT) for cardiovascular care.
