Benefit of mixed reality assistance on CT-based ear anatomy spatial recognition: a preliminary study

To evaluate the benefit of a mixed reality (MR) platform on CT-based ear anatomy spatial recognition and annotation performance by comparing with traditional learning.

A prospective study (N=8) assigned undergraduate medical students to the MR group and traditional group (textbooks and images). Each completed identification exercises of 8 key anatomical structures on 20 temporal bone CT cases over three learning phases. Outcomes included annotation accuracy, task completion time and learning curve. Participants completed a Likert-style questionnaire evaluating the learning experience and the adverse health effects experienced.

The MR group achieved an earlier learning plateau and better performance than the traditional group (6.0±3.5 vs. 8.0±4.9 cases). Mixed-design ANOVA demonstrated consistent, large-magnitude performance advantages for the MR group in annotation accuracy (η²ₚ=0.605) and task completion time (η²ₚ=0.875). MR teaching method was most advantageous for complex structures such as the ossicular chain and facial nerve. Learning experience scores were higher with MR, though 50% of MR users reported mild visual fatigue.

In this preliminary study, MR-assisted learning shortened the learning curve and improved performance on a CT-based temporal bone annotation task compared with textbook-based learning, with the most pronounced advantages observed for spatially complex structures. These findings indicate the potential of MR as an adjunct in CT-based ear anatomy training and warrant evaluation in larger cohorts.