Intraoperative tracking of aortic valve plane

Area of Science:

• Medical Imaging
• Cardiovascular Surgery
• Computer Vision

Background:

• Transcatheter Aortic Valve Implantation (TAVI) requires precise visualization of the aortic valve plane.
• Intra-operative fluoroscopic imaging presents challenges for accurate valve plane localization.
• Aortic valve calcifications are key landmarks for tracking the valve plane.

Purpose of the Study:

• To develop and evaluate a novel method for tracking the aortic valve plane in real-time fluoroscopic images during TAVI.
• To improve the accuracy and safety of TAVI procedures through enhanced image analysis.
• To specifically address the challenge of tracking aortic valve calcifications.

Main Methods:

• A Tracking-Learning-Detection (TLD) approach was adapted for aortic valve calcification tracking.
• The method involves updating a recursive tracker that jointly tracks all relevant features.
• The system determines the aortic valve plane position within intra-operative TAVI fluoroscopic images.

Main Results:

• The proposed method achieved an absolute mean displacement error of less than 10 pixels (approximately 2mm).
• Performance was validated across four distinct patient databases.
• The system demonstrated high accuracy in tracking aortic valve calcifications.

Conclusions:

• The developed TLD-based method is suitable for optimizing and securing the TAVI procedure.
• Accurate tracking of the aortic valve plane using calcifications can enhance surgical precision.
• This approach offers a valuable tool for improving TAVI procedural safety and efficacy.