Towards Effective Telerehabilitation: Assessing Effects of Applying Augmented Reality in Remote Rehabilitation of Patients Suffering from Multiple Sclerosis
Magdalena Pruszyńska, Marta Milewska-Jędrzejczak, Igor Bednarski, Piotr Szpakowski, Andrzej Głąbiński, Sławomir Konrad Tadeja · 2022 · ACM Transactions on Accessible Computing · doi:10.1145/3560822
Summary
This paper presents a randomized controlled trial evaluating augmented reality-based telerehabilitation for patients with multiple sclerosis (MS). The study involved 30 patients with relapsing-remitting MS, randomly assigned to either an AR intervention group or a control group. The AR group used Neuroforma, a commercially available system that displays the user's mirror reflection on screen overlaid with virtual objects they must interact with through hand and arm movements. The four-week intervention required participants to complete 20 sessions (five per week), each lasting 40-45 minutes. The AR system included eight upper limb exercises: pressing virtual lights, collecting chestnuts, following tracks, chasing butterflies, wiping stains, and following balloon paths. Exercises targeted both bilateral and unilateral arm movements. The system provided real-time visual and audio feedback, tracked exercise completion, and automatically notified therapists if patients missed sessions for more than 48 hours. The study was conducted during the COVID-19 pandemic, which shaped the methodology: all exercises were performed at home, and the control group was advised to exercise independently without the AR system or monitoring. Outcomes were measured at baseline and after four weeks using the 9-hole peg test (fine motor coordination), a ball-pulling task (upper limb efficiency), grip strength dynamometry, and blood tests for neurotrophins (BDNF and PDGF) as markers of potential brain plasticity changes.
Key findings
The AR intervention group showed statistically significant improvements across all motor function measures. On the 9-hole peg test, completion times decreased from 3.34 to 3.19 seconds for the weaker hand and from 3.25 to 3.10 seconds for the stronger hand (both p ≤ 0.01). Ball-pulling times improved similarly, dropping from 2.55 to 2.39 seconds (weaker) and 2.47 to 2.35 seconds (stronger). Grip strength increased substantially—from 23.87 to 27.53 kg for the weaker hand and from 25.87 to 29.67 kg for the stronger hand. Critically, the control group showed no significant changes on any measure, suggesting the improvements were attributable to the AR-based intervention rather than natural variation or general activity. There were no significant differences between weaker and stronger limbs within the study group, indicating balanced bilateral improvement. However, the study found no significant changes in neurotrophin levels (BDNF or PDGF) in either group. The authors suggest the four-week duration may have been too short to induce measurable neuroplasticity biomarker changes, noting that prior VR studies showing neurotrophin effects used 5-8 week interventions combined with standard occupational therapy.
Relevance
This study provides rigorous evidence that AR-based telerehabilitation can meaningfully improve motor function in MS patients—a finding with broad implications for accessible healthcare delivery. For people with MS and other chronic conditions causing motor disability, the ability to receive effective rehabilitation at home removes significant barriers: travel difficulties, fatigue management, infection risk (particularly relevant given COVID-19), and geographic access to specialized care. The use of a commercially available system (Neuroforma) rather than custom research hardware strengthens the practical relevance. The system's features—remote therapist monitoring, automated compliance reminders, gamified exercises with real-time feedback—represent a model for accessible telehealth that maintains clinical oversight while enabling independence. For accessibility practitioners and healthcare technologists, the study demonstrates that immersive technologies can be successfully deployed for populations with motor impairments, provided the interaction design accommodates varying ability levels. The exercises were performed seated for safety, and the system verified correct posture before starting. The absence of neurotrophin changes despite functional improvements suggests that behavioral outcomes and biological markers may operate on different timescales—a consideration for future study design and for setting patient expectations about rehabilitation benefits.
Tags: augmented reality · telerehabilitation · multiple sclerosis · motor rehabilitation · upper limb function · neuroplasticity · remote healthcare