Breast Magnetic Resonance Image Analysis for Surgeons Using Virtual Reality: A Comparative Study
Authors
Mohamed El Beheiry, Thomas Gaillard, Noémie Girard, Lauren Darrigues, Marie Osdoit, Jean-Guillaume Feron, Anne Sabaila, Enora Laas, Virginie Fourchotte, Fatima Laki, Fabrice Lecuru, Benoit Couturaud, Jean-Philippe Binder, Jean-Baptiste Masson, Fabien Reyal, Caroline Malhaire
Abstract
PURPOSE
The treatment of breast cancer, the leading cause of cancer and cancer mortality among women worldwide, is mainly on the basis of surgery. In this study, we describe the use of a medical image visualization tool on the basis of virtual reality (VR), entitled DIVA, in the context of breast cancer tumor localization among surgeons. The aim of this study was to evaluate the speed and accuracy of surgeons using DIVA for medical image analysis of breast magnetic resonance image (MRI) scans relative to standard image slice-based visualization tools.
MATERIALS AND METHODS
In our study, residents and practicing surgeons used two breast MRI reading modalities: the common slice-based radiology interface and the DIVA system in its VR mode. Metrics measured were compared in relation to postoperative anatomical-pathologic reports.
RESULTS
Eighteen breast surgeons from the Institut Curie performed all the analysis presented. The MRI analysis time was significantly lower with the DIVA system than with the slice-based visualization for residents, practitioners, and subsequently the entire group ( P < .001). The accuracy of determination of which breast contained the lesion significantly increased with DIVA for residents ( P = .003) and practitioners ( P = .04). There was little difference between the DIVA and slice-based visualization for the determination of the number of lesions. The accuracy of quadrant determination was significantly improved by DIVA for practicing surgeons ( P = .01) but not significantly for residents ( P = .49).
CONCLUSION
This study indicates that the VR visualization of medical images systematically improves surgeons' analysis of preoperative breast MRI scans across several different metrics irrespective of surgeon seniority.