Coronary Heart Disease Preoperative Gesture Interactive Diagnostic System Based on Augmented Reality

Area of Science:

• Medical Imaging
• Human-Computer Interaction
• Surgical Planning

Background:

• Traditional coronary heart disease diagnosis relies on invasive angiography, lacking intuitive interaction and precise measurements.
• Current methods for assessing vascular stenosis are empirical and can be inaccurate, impacting surgical planning.

Purpose of the Study:

• To develop a gesture-interactive diagnostic system using Augmented Reality (AR) for coronary heart disease preoperative diagnosis.
• To enhance the accuracy and intuitiveness of vascular stenosis diagnosis and measurement in interventional surgery planning.

Main Methods:

• Utilized Leap Motion Controller for capturing hand gesture data (position, orientation, hand shape).
• Employed K-means algorithm for gesture training, SRT-HMM for gesture recognition, and state transition models for gesture reusability.
• Incorporated gesture prejudgment and space rotation transformation to improve algorithm efficiency and trajectory integrity.

Main Results:

• The system demonstrated accurate gesture recognition with good distinguishability and generalization abilities.
• Augmented reality and gesture interaction enabled intuitive diagnosis and measurement of vascular stenosis on coronary artery models.
• Gesture interaction experiments confirmed the system's availability and reliability for preoperative diagnosis.

Conclusions:

• The proposed Augmented Reality-based gesture interactive system offers an intuitive and accurate solution for coronary heart disease preoperative diagnosis.
• This approach significantly improves upon traditional methods by providing natural interaction and precise measurements for vascular stenosis.
• The system holds promise for enhancing the planning and execution of vascular interventional surgeries.