Preliminary investigation: Feasibility study of a virtual reality breast biopsy simulation

Ultrasound-guided breast biopsies are the most frequently performed imaging-guided breast procedures. During training, learners may lack exposure due to limited volume and sensitive anatomy. Current simulation training often involves the use of manufactured or homemade (chicken or turkey breast) phantoms. Virtual reality is an emerging technology, allowing learners to have flexibility in learning, real-life interactive experiences and measurable feedback. The purpose of this study is to assess the feasibility of a novel virtual reality breast biopsy simulation trainer.

This was a HIPAA-compliant, IRB-approved prospective feasibility study. The participants were three fellowship-trained breast radiologists with varying years of experience. Participants received an introduction to the virtual reality (VR) breast biopsy simulation and brief training session, describing how to enter the simulation and navigate controls. The participants were instructed to perform as many breast biopsies as possible within a 15-min time period. One biopsy cycle consisted of entering the breast with the biopsy needle, taking a biopsy sample and removing the needle. Time to successfully biopsy the mass was recorded by the VR program, and this data was recorded to assess improvement from start to finish of the simulation. A post-procedure survey was administered to all participants immediately after completion of the simulation. Qualitative open-ended subjective feedback was also obtained via Qualtrics.

All three breast radiologists completed the simulation. There were no complications from the procedure, including no motion sickness or fatigue resulting in simulation termination. Results of data obtained from the simulation showed decreased time to successful biopsy (slope = -19.23) with each subsequent trial for all three participant breast radiologists. A decreased time to biopsy was associated with a higher cumulative number of successful biopsies (p = 0.0037). A higher number of cumulative successful biopsies was associated with decreased number of body entries (p = 0.0332) and biopsy fires (p = 0.0221) before a successful tissue sample. Mean responses for Likert scale survey results were overall high. The radiologists found the simulator to be engaging (4.67/5.00 ± 0.47), realistic (2.67/3.00 ± 0.47) and would recommend the simulation to other healthcare professionals (2.67/3.00 ± 0.47). The radiologists participating in the trial also provided overall favorable subjective feedback.

This study presents a novel approach for ultrasound-guided breast biopsy training with a VR simulation that showed to be successfully capable of recording time-to-completion of each biopsy attempt. Future studies will be directed toward assessing the utility of the simulation in improving trainee skills.