Investigating the use of an open source wearable as a tool to assess sports related concussion (SRC)

Purpose: Concussion is common in rugby and returning to play before complete recovery increases risk of secondary injury. Pitch-side medical staff make return-to-play decisions based on brief post-injury subjective observations of motor tasks, e.g. gait and balance. Subtle deficits that last for long periods may be missed. It is proposed that high resolution digital technologies could capture objective inertial (accelerometer and gyroscope) data to aid motor screening. Here, we investigate an open source inertial wearable to monitor free-living/habitual gait as a possible assessment tool. The challenges of deploying this technology in amateur rugby athletes has not been explored. We present initial findings in using the wearable for gait and balance assessment in sports related concussion (SRC) for more informed player management including: (i) acceptance, (ii) deployment, (iii) data capture and (iv) data handling/processing.

Methods: 18 male university rugby union players were asked to wear a single research-grade wearable (AX6, Axivity: 100 Hz) on the lower back (direct to skin with tape) during laboratory-based supervised tasks within 48-h period after matches: timed-up-and-go, 2-min walk and Balance Error Scoring System. Additionally, players were asked to continuously wear the AX6 for ≈7-days to gather free-living/habitual data. Player feedback was captured via semi-structured participant responses.

Results: Wearable data were gathered on 18 males (19.71 yrs ± 1.02 yrs). One physiotherapist was able to successfully prepare and deploy multiple AX6's during laboratory and free-living assessment.
i)
Acceptance: There was no reported discomfort due to wearable location or attachment method during laboratory or free-living assessment.

ii)
Deployment: For laboratory assessments it took approx. 5–10 min to configure, attach devices and complete tasks. Devices did not impact or hinder participant's ability to complete any task. For free-living it took 2–3 min to configure and attach devices.

iii)
Data capture: The AX6 captured objective data for laboratory tasks and 7-day free-living assessment with no loss of data due to memory or power/battery failure.

iv)
Data handling: Laboratory data files (≈10 mb) were quick to download via USB (<1-min) with the accompanying software (OmGUI). Free-living files (≈500 mb) took significantly longer (>15 min). Preliminary analyses found the AX6 provided robust gait and balance data but required separate Matlab-based algorithms.


Conclusion(s): Digital SRC assessment could be useful in providing insight beyond acute timeframes and objective measurements to better inform SRC diagnosis and return to play protocols. The open source AX6 is able to provide objective and robust gait and balance data from laboratory and free-living assessment without player concern and in a timely manner. However, generation of gait and balance data require specialist engineering knowledge through Matlab which limits use. Our future work is investigating data capture and handling by streamlining through the Cloud on a larger number of squads for multidisciplinary use.

Impact: Wearables are a novel technology, that can be deployed in a range of different environments which could have broad applications across clinical and allied health professions. Adopting these new technologies could improve current assessment techniques by providing more objective data in a range of pathological cohort such as SRC.