Automated assessment of cortical mastoidectomy performance in virtual reality

Area of Science:

• Surgical Education
• Medical Simulation
• Otolaryngology

Background:

• Cortical mastoidectomy is a fundamental surgical skill for Otolaryngology trainees.
• Objective assessment of surgical competency is crucial for reducing subjectivity and trainer workload.

Purpose of the Study:

• To develop and validate an automated competency assessment system for cortical mastoidectomy.
• To evaluate the system's generalizability and comparability to expert human assessment.

Main Methods:

• Trainees performed cortical mastoidectomies on a virtual reality temporal bone simulator.
• Simulator metrics were developed to align with the Melbourne Mastoidectomy Scale.
• Automated scores were validated against blinded expert assessments.

Main Results:

• The automated system achieved high accuracy (0.9450), sensitivity (0.9547), and specificity (0.9343) compared to expert scores in Group 1.
• Strong correlations were found between automated and expert total scores for both validation groups (r=0.9144 and r=0.7224).

Conclusions:

• A virtual reality-based automated assessment method for cortical mastoidectomy competency has been established.
• This system provides an objective and reliable evaluation comparable to expert surgical assessment.