Machine learning-based operation skills assessment with vascular difficulty index for vascular intervention surgery

Area of Science:

• Medical Engineering
• Machine Learning in Medicine
• Surgical Skill Assessment

Background:

• Accurate assessment of surgical skills is vital for improving vascular intervention outcomes and endovascular robotic surgery.
• Current assessment methods often overlook patient-specific vascular conditions, leading to evaluation inaccuracies.

Purpose of the Study:

• To propose an operation skills assessment method that includes a vascular difficulty level index for catheter insertion at the aortic arch in endovascular surgery.
• To enhance the objectivity and accuracy of surgical skill evaluation by integrating patient-specific vascular characteristics.

Main Methods:

• Machine learning models were developed to characterize vascular anatomical structures and determine the difficulty of the aortic arch.
• The vascular difficulty level was combined with surgeon operating characteristics (e.g., speed, smoothness) as objective indices.
• Skills were evaluated using machine learning based on these combined indices for catheter insertion at the aortic arch.

Main Results:

• The proposed method integrates vascular difficulty into skill assessment for endovascular surgery.
• The accuracy of surgical skill assessment improved significantly from 86.67% to 96.67% with the inclusion of the vascular difficulty index.
• This demonstrates a more objective and precise evaluation of surgical expertise.

Conclusions:

• The developed method offers a more accurate and objective approach to assessing surgical skills in endovascular procedures, particularly for aortic arch catheter insertion.
• This technique can be utilized for training novice surgeons and advancing research in vascular interventional surgery robots.