FULL BODY HAPTIC BODYSUIT - AN INSTRUMENT TO MEASURE THE RANGE AND SPEED OF MOTION IN PATIENTS WITH AXIAL SPONDYLOARTHRITIS (AXSPA) - PRELIMINARY RESULTS

BackgroundMovement of the spine is restricted in axial spondyloarthritis (axSpA) [1]. Spine function is usually assessed by the Bath Ankylosing Spondylitis Metrology Index (BASMI), which is based on a limited set of defined motions that are measured semiquantitatively in the spatial dimension but not in the temporal dimension. Sensor-based measurement of spine function in axSpA patients is in its infancy but may provide a deeper and more detailed understanding of the impact of axSpA on the impairment of spine function [2,3]. In theory, unbiased full body assessment of spine motion may open a new dimension in function analysis in axSpA.ObjectivesTo test if a of a full-body based haptic capturing of spine motion is technically feasible and can pick up the measurements of BASMI items 1-5. Furthermore, we aimed to investigate whether such measurements are accurate and reproducible comparing to BASMI scores done by rheumatologists. Lastly, we sought to measure velocity of spine movements to allow spatiotemporal analysis of motion.MethodsFor full-body haptic assessment of spine motion a full-body haptic suit (Teslasuit; VR Electronics Ltd, London) was used that consists of a smart textile two-piece bodysuit that not only captures range and speed of motion but also provides biometric and haptic feedback. This device is currently tested in clinical trials (https://teslasuit.io/rehabilitation/) but has not been tested in rheumatic diseases such as axSpA [4]. Since there is no pre-defined technical solution for measuring BASMI, we used the integrated inertial measurement units (IMUs) of the suit (Figure 1a). The suit is recording the position and rotations of its IMUs and is transforming the raw data to position coordinates and joint angle of the bones. We implemented an algorithm that is accessing the sensor data and is calculating the BASMI measurements as well as velocity. Assessment were done in healthy individuals. BASMI was obtained 3 times by teslasuit followed by standard BASMI measurement by 2 independent rheumatologists. In addition, rotational movements with their maximum torso speed to evaluate angular velocity were performed (Figure 1b). Measurements were compared using absolute values and relative standard deviation (which is the standard deviation normalized by the mean).ResultsFive healthy individuals (all males, age: 27.6 ± 1.8 years, height: 178 ± 5 cm; weight 70.0 ± 8.0 kg) were assessed. Teslasuit measurements were well tolerated. Technically, we were able to calculate BASMI item 3 and 5, finger-to-floor distance and the velocity of the spine movement using the position data of hand, talus and upper back sensors (Figure 1a, b). Due to absence of sensors at the head and the required back areas, BASMI 1, 2, 4 could only partially be captured and require further programming, which is currently performed. Only marginal differences were detected regarding the relative standard deviations of measurements between teslasuit and rheumatologists (BASMI 3: rheumatologists 8,5%: suit 10%; BASMI 5: rheumatologists 5,4%: suit 4,9%) (Figure 1c). The speed of spinal motion could be measured with an average angular velocity of 172.2 degrees/sec over the entire rotation motion and an average maximum angular velocity of 417.2 degrees/sec.ConclusionThis study shows that full-body haptic-suits can capture spinal motion including parts of the BASMI score. In addition, they allow to measure the speed of spinal movement, which might be an important and so far unrecognized factor to test the impact of axSpA on spinal function. Based on these results, full-body haptic-suits will be tested in axSpA patients in the future. Furthermore, technical solutions are currently developed to implement the remaining BASMI scores into the suit as well as connections from the suit to virtual reality devices for patients and doctors.References[1]Sieper, J. and D. Poddubnyy, Axial spondyloarthritis. Lancet 2017.[2]Gardiner, P.V., et al., Rheumatology (Oxford) 2020.[3]Kiefer, D., et al., S J Rheumatol 2022.[4]Caserman, P. et al Sensors (Basel) 2021.AcknowledgementsThis work was (partly) funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – SFB 1483 – Project-ID 442419336, EmpkinS.Disclosure of InterestsNone declared