AUTHOR=Rother  Anne , Spiliopoulou Myra TITLE=Virtual Reality for Medical Annotation Tasks: A Systematic Review JOURNAL=Frontiers in Virtual Reality VOLUME=3 YEAR=2022 URL=https://www.frontiersin.org/journals/virtual-reality/articles/10.3389/frvir.2022.717383 DOI=10.3389/frvir.2022.717383 ISSN=2673-4192 ABSTRACT=

Virtual reality technologies are broadly used in medicine, including medical educational tasks like surgery training. Annotations are an inseparable part of many medical research and educational tasks. In this systematic review, we investigate the potential of VR for medical tasks with focus on annotation. The questions we pursue are as follows. (Q1) For which healthcare-associated tasks do we find VR-associated investigations and which involve a crowd worker-based annotation? (Q2) To what extent are there gender-specific differences in the usage of VR? To address these questions, we formulated a keyword list and inclusion/exclusion criteria for the collection of recent scientific articles according to the PRISMA Statement 2020. We queried the Medline database and included 59 free full articles available in English and published from 2017 upward. We inspected the abstracts of the retained articles and organized them into 6 categories that referred to VR in the medical context. We identified categories of medicine-related tasks, for which VR is used, and one category associated to cybersickness. We traced technologies used with a higher priority for some tasks, and we found that gender-related investigations are more widespread for some categories than for others. The main findings of our investigation on the role of VR for medical annotation tasks are as follows: VR was used widely for tasks associated with medicine, including medical research and healthcare, but the use of VR for annotation purposes in that context was very limited. Many of the relevant studies concerned VR in education, where annotations may refer to labeling or other enhancements of materials or may refer to exercises. The investigation of gender-related aspects was typically found in studies that encompassed the usage of VR on patients and controls, or on healthy participants in order to assess the potential and limitations of VR for specific tasks/medical assessments or treatments. To fully exploit the VR potential for tasks of medical annotation, especially for the creation of ground truth datasets and similar resources, more research is needed, especially on the interplay of annotator demographics and accessibility to VR technologies.