AutoClip, an artificial intelligence system using Attention U-Net architecture, achieved simultaneous segmentation of mitral valve leaflets and repair device components during transcatheter edge-to-edge repair (TEER) procedures. Testing on 987 echocardiography frames from three patients undergoing MitraClip implantation, the system reached a mean intersection-over-union score of 0.46 for identifying clip arms, grippers, delivery shafts, and valve anatomy in real-time surgical imaging. This represents the first TEER-specific AI segmentation framework designed to assist cardiac surgeons during complex valve repair procedures that currently rely heavily on operator experience and manual image interpretation. The technology could potentially standardize these intricate procedures and reduce variability between surgeons, though current accuracy remains modest due to the small sample size. For patients with mitral regurgitation, improved procedural guidance could translate to better outcomes and reduced complications during minimally invasive valve repairs. However, this proof-of-concept preprint study awaits peer review, and the limited three-patient dataset suggests results may change significantly with larger validation studies. The work establishes a reproducible framework for developing AI-assisted surgical guidance systems in structural heart interventions.
AI System AutoClip Segments Heart Valve Repair Devices During Surgery
📄 Based on research published in medRxiv preprint
Read the original research →⚠️ This is a preprint — it has not yet been peer-reviewed. Results should be interpreted with caution and may change following peer review.
For informational, non-clinical use. Synthesized analysis of published research — may contain errors. Not medical advice. Consult original sources and your physician.