Aim

This study aims to compare the accuracy of 3D markerless movement data captured with a single smartphone camera with the current gold standard (3D Marker based movement data).  For this study healthy participants will be performing rehabilitation exercises used in current clinical physiotherapy practice. Additionally, the movement data will be used to explore the development of a quality movement index, which summarises the quality of an exercise in one single number.

Background

Musculoskeletal conditions like osteoarthritis affect over 1.7 billion people globally. Rehabilitation programmes such as “Good Life with Osteoarthritis in Denmark” (GLA:D®), have proven to reduce pain, improve mobility and quality of life through education and specific exercises. However, after patients go home and have left the supervised setting of private practice, one question remains: How well are the exercises actually being performed? Existing digital tools can remind patients what to do but cannot assess how well they do it, meaning there is little feedback on exercise quality.

New technologies such as 3D movement analysis based on a video from a single smartphone have demonstrated to capture and assess the quality of gait and specific assessments. These markerless motion capture technologies could pave the way for movement analysis in real life settings, such as at home. However, the validity and accuracy of assessing rehabilitation exercises with this technology is still unknown.

In addition, standardised ways to interpret rehabilitation exercise quality in the home setting is also not yet explored. The markerless motion capture technologies could also make this possible. The idea: a new quality measure – provisionally called the Exercise Deviation Index (EDI) – summarises the performance of an exercise in a single number. This would allow you to see at a glance whether an exercise has been performed correctly or whether there are deviations from the optimal movement pattern. Patients would thus receive clear, objective feedback – presented in an understandable way. This study is part of our vision which aims to build a real-time feedback system based on markerless movement data, which people who are prescribed exercises can use at home.

Project content and Procedure

This project comprises three smaller parts:

1) Selection of the most important exercises

In the first part, we want to find out which exercises in the GLA:D® programme are best suited for individual feedback. To do this, we are developing a survey for physiotherapists who work with the programme. They will be asked to describe how they implement GLA:D® in practice – and, above all, which exercises they believe are particularly well suited for feedback at home. This selection will form the basis for the next step in the project.

2) Validation in the movement laboratory

The second part focuses on direct comparison: How well does a single smartphone measure up to the gold standard in motion analysis? To find out, healthy participants come to the motion lab and perform four selected GLA:D® exercises. The movements are recorded simultaneously using the VICON system (gold standard) and several markerless systems, including a smartphone. The data is then evaluated and compared.

3) Development of the EDI

The third part explores the development of the Exercise Deviation Index (EDI). First, experienced physiotherapists assess which exercises from part 2 are ‘correctly’ or ‘incorrectly’ executed. Then, the correctly performed exercises are used as a reference to create a ‘standard profile’ – a template for ideal execution of the movement. Next, the less well-executed or incorrectly performed exercises are used to test whether the EDI reliably detects these deviations from the standard profile.

Innovation

This project will lay the foundation for a practical, accessible tool that enables remote, feedback-driven rehabilitation. By validating the accuracy of a low-cost, smartphone-based system, we take a critical step toward personalisation of exercises, improving adherence, and clinical outcomes in osteoarthritis care.

More broadly, our work is part of a growing movement towards digital, scalable, and patient-centred rehabilitation solutions that not only improve the quality of life for those affected, but also significantly reduce the burden on healthcare systems.

© Photo: Christoph Böhm