Acquisition and Documentation of Digital Range of Motion using Mixed Reality Headsets
Regina WK Leung, MD, MHSc, B.Eng1; Caitlin J Symonette, MD, MSc, FRCSC2; Douglas Ross, MD, MEd, FRCSC3
1Western University, Schulich School of Medicine, London, ON, Canada; 2University of Western Ontario, Schulich School of Medicine, London, ON, Canada; 3Plastic and Reconstructive Surgery, St. Joseph's Health Care Center, Western University, Schulich School of Medicine, London, ON, Canada
Introduction: Assessing finger joint range of motion (ROM) using a hand-held goniometer is fundamental to guiding diagnosis and management in hand surgery. However, the acquisition and documentation of these measurements is tedious, time consuming, and error prone. We postulated that recent advancements in hand-tracking capabilities within advanced mixed reality headsets such as the Microsoft HoloLens™ could provide a means to automate ROM measurements and subsequent documentation. This pilot study aims to develop a mixed reality tool using the HoloLens 2™ to automatically track, measure, and document finger ROM measurements at the MCP, PIP, and DIP joints with a goal of condensing a manual 5-10 minute workflow to an accurate and precise 5-10 second automated process.
Methods: The tool/app was developed in Unity using the MRTK software package for hand tracking on the HoloLens 2 and written in C# programming language. To assess accuracy and repeatability of ROM measurements, repeat samples (10) were taken at each joint (MCP, PIP, DIP) at various reference joint angles (ie 10, 30, 60 deg to max limit) held in position by a splint. All measurements were repeated for the left hand index and ring finger with 2 views (oblique and palmar). Mean differences from the reference angle and repeatability (ie difference between two measurements will be less than this value 95% of the time) were the primary outcomes.
Results: Finger ROM angles were automatically measured and labeled on a captured image of the hand by the HoloLens tool. (Figure 1) Our pilot study showed that repeatability of all HoloLens ROM measurements was less than known goniometer inter-rater repeatability of 7-9°. Mean differences and repeatability (MD, R) for the MCP, PIP, and DIP of the index finger oblique view were (1.09°, 4.73°), (1.83°, 3.64°), (1.65°, 4.16°) respectively and (1.08°, 4.86°), (1.56°, 3.78°), (1.38°, 5.14°) respectively for the ring finger oblique view. Palmar view measurements produced similar results. The labelled image and results were then automatically transmitted from within the tool via email and copy/pasted into electronic medical records.
Conclusions: We successfully developed an automated mixed reality tool using the HoloLens 2 to acquire and document finger ROM measurements thereby reducing the overall workflow to 5-10 seconds. Furthermore, our pilot study showed that mixed reality HoloLens ROM measurements are as accurate and reliable as hand-held goniometer measurements.
Back to 2023 Abstracts