Application of Augmented Reality in Robot-Assisted Mitral Valve Repair Surgery: A Feasibility Study

Area of Science:

• Robotic Surgery
• Medical Imaging
• Augmented Reality

Background:

• Intraoperative visualization of hidden anatomy is crucial for safe and effective mitral valve surgery.
• Minimally invasive approaches like robotic mitral valve repair require advanced visualization tools.

Purpose of the Study:

• To develop and assess semi-automated intraoperative 3-dimensional augmented reality (3D-AR) overlays for robotic mitral valve repair.
• To visualize critical anatomical structures during robot-assisted mitral valve surgery.

Main Methods:

• Generated 3D point clouds from intraoperative stereoscopic images captured during robot-assisted mitral valve repair in 5 patients.
• Created 3D-AR overlays using a scale-adaptive iterative closest point algorithm and mitral valve annulus landmarks.
• Validated accuracy by calculating 3D point cloud accuracy and landmark registration error (LRE).

Main Results:

• Successfully created accurate 3D point clouds and 3D-AR overlays for all 5 patients.
• Achieved a median 3D point cloud error of -0.92 mm and an LRE of 5.12 mm.
• Overlay generation took approximately 5 minutes, with visualization possible directly within the robotic console.

Conclusions:

• Presented a novel algorithm for accurate, semi-automated 3D-AR overlay generation in robot-assisted mitral valve repair.
• Demonstrated the potential to visualize essential anatomical structures, including the circumflex artery and aortic valve leaflets.
• This technology may advance automated intraoperative 3D-AR vision in robotic cardiac surgery, improving safety, accuracy, and efficiency.