Can Motion Capture Detect Disease Before Symptoms Appear?

Can we detect early signs of disease simply by analysing how people move?

Dr Matthew Banger, Gait Lab Technician at Imperial College London’s MSk Lab, is pushing motion capture beyond the lab and into the real world, using it to answer crucial questions about movement and redefine how we measure, understand and treat musculoskeletal diseases. We had the opportunity to speak to Matthew about the research at a recent Imperial Lates event.

https://www.youtube.com/watch?v=4K5Bv_WRmEM

Why Movement Matters

Long before disease symptoms become obvious, subtle changes occur in the way the body positions the skeleton and compensates for shortcomings through adjusting posture and gait.

The challenge is catching those changes early enough to make a difference.

"We're trying to work out how early you can actually start picking these things up," Matthew explains.

Traditional marker-based motion capture systems are incredibly accurate, and are indeed used in the world’s top biomechanical research labs. 

These systems use motion capture suits, retroreflective markers placed strategically on the human body and infrared cameras which feed positional data into software and create a skeleton that can be tracked to millimetre accuracy. 

However, a marker-based motion capture shoot can be lengthy and resource-intensive. It would be logistically difficult to realistically screen thousands of people with this capture method.

Markerless Motion Capture in Public

Using synchronised video cameras from OptiTrack and machine learning models from Theia trained to detect human features, the system reconstructs a skeletal model, mapping head position, shoulder and hip alignment, knee tracking and overall gait patterns, all without a single marker.

Beneath clothing and soft tissue, biomechanical modelling estimates the underlying skeletal movement. The result is a fast, scalable and surprisingly engaging screening tool. At Imperial Lates, around 100 people were captured in just three hours. At a previous two day event, the team screened over 300 participants.

The long-term ambition? 10,000 people over the next two years.

From Public Screening to Deep Profiling

If someone’s movement patterns suggest early deviation from a healthy baseline, they may be invited into the Biodynamics Lab for deeper analysis.

From there, the research expands beyond movement alone, exploring potential biochemical factors and interdisciplinary collaboration. The aim isn’t to “cure” diseases like arthritis, rather to enable earlier intervention and improve resilience before decline becomes debilitating.

Taking Research to Communities

One of the most important aspects of this project is access.

Rather than waiting for people to come to the university, the team is going to them, engaging communities that might not naturally participate in academic research. Bringing the technology to cultural centres, religious organisations and public venues grants access to far more participants, and when trying to understand how disease develops across populations, representation matters.

Markerless systems make this possible. A three minute capture and a short follow up questionnaire can build a meaningful biomechanical profile, something that simply wouldn’t be feasible with traditional lab sessions.

Why This Matters

Musculoskeletal conditions are among the most common causes of chronic pain and disability worldwide. Arthritis alone affects millions of people, and for many, the diagnosis only comes once significant damage has already been done. By that point, treatment is often about managing decline rather than preventing it.

The shift in how someone walks, how they bear weight, how their knee tracks as they move, can begin years before a person feels any pain. But without a practical way to measure those changes at scale, the medical system has largely had to wait for people to show up with symptoms. That's the gap this research is trying to close. 

"We've got a really nice picture of that person," Matthew says of the data they collect.

Movement, biomechanics and questionnaire responses all combine into a profile that could, one day, flag someone as worth a closer look before they ever set foot in a GP's waiting room.

If it works, the implications stretch well beyond arthritis. The same approach - large-scale, low-barrier movement screening - could eventually be applied to a whole range of conditions where how we move is an early warning sign that something is changing beneath the surface.

“We're just trying to put people in a better place so that they can cope better with the onset of these diseases — and potentially also react better to their treatments."

Want to Get Involved?

The MSk Lab at Imperial College London is actively recruiting participants as the project expands into communities.

If you’d like to take part, host a session, or follow where the team will be next, you can connect with Dr Matthew Banger via email, and the MSk Lab directly via their site or their X/Twitter account.