A new human heart vessel identification, segmentation and 3D reconstruction mechanism

Area of Science:

• Medical Imaging
• Computer-Aided Surgery
• Cardiovascular Anatomy

Background:

• Accurate identification and segmentation of inhomogeneous image regions, particularly human heart surface vessels, pose significant challenges in modern medicine.
• Precise localization of cardiac vessels is crucial for surgeons to plan interventions, minimize injury risks, and enhance training for novice practitioners.

Purpose of the Study:

• To introduce a novel mechanism for identifying surgical vessel positions through image enhancement.
• To develop a 3D vessel reconstruction technique from single-view color images of the human heart obtained during open-heart surgery.

Main Methods:

• Image enhancement to identify surgical vessel positions.
• 3D reconstruction of cardiac vessels from single-view color images.
• Vessel segmentation using the RGB color space.

Main Results:

• The proposed method significantly reduces the time needed to locate the vessel region of interest (ROI), ensuring clear visualization for surgeons.
• Achieved a 94.42% accuracy rate for vessel segmentation using the RGB color space, outperforming other color spaces.
• Demonstrated a reduction in training time for novice cardiac surgeons.

Conclusions:

• The developed mechanism assists surgeons in performing procedures more efficiently, minimizing errors, and reducing surgical effort.
• The 3D vessel reconstruction from a single image facilitates learning of heart anatomy and improves cardiac surgery training for novices.
• Extensive experiments confirm the proposed method's superior performance compared to existing state-of-the-art techniques.