Site Editor
William J. Gradishar, MD, FACP, FASCO
William J. Gradishar, MD, FACP, FASCO
Posted: Friday, January 28, 2022
The use of a virtual reality visualization tool called DIVA appears to improve the ability of breast cancer surgeons to analyze preoperative breast MRI scans quickly and accurately when compared with standard image slice–based tools. These results held true regardless of surgeon seniority, but improvement was notably greater for residents. Caroline Malhaire, MD, of the Institut Curie, Paris, and colleagues published their findings in JCO Clinical Cancer Informatics.
“Visual reality technology such as DIVA has the potential to broadly transform the way medical images are communicated in the clinical setting,” concluded the authors.
Researchers evaluated the speed and accuracy of nine residents and nine practicing surgeons, using both the DIVA tool and standard image slice–based visualization to analyze breast MRI sequences. MRI sequences were from 14 patients who underwent a mastectomy, 11 patients who had a lumpectomy, and 2 healthy control cases who required no surgery.
Surgeons were timed while analyzing a total of nine random images: three using the DIVA tool, three using slice-based visualization, and three using the radiologist report without MRI images. The researchers defined speed as the time it took to analyze and produce a conclusion on tumor localization and surgical strategy. The surgeons’ analysis of breast determination, number of lesions, and breast quadrant localization were compared with postoperative reports to determine accuracy metrics.
All surgeons were able to determine the lesion number and location more quickly with the DIVA tool compared with slice-based visualization (P = .001). The time needed to determine the lesion location was similar for using DIVA and reading the radiologist report (P = .14).
Residents and practicing surgeons both showed improved accuracy in all metrics, except for the number of lesions, when using DIVA. Experienced surgeons identified the correct number of lesions 79.2% of the time using both DIVA and slice-based visualization. The researchers concluded that overall analysis performance with DIVA improved with surgeon seniority.
Disclosure: For a full list of authors’ disclosures, visit ascopubs.org.
JCO Clinical Cancer Informatics
JNCCN Spotlights
