Digital sensor motion analysis

Wearable sensors track joint angle, movement speed, symmetry and repetition count during exercise, feeding it back live. The patient sees whether the movement was performed correctly at the moment they perform it, rather than being corrected afterwards.

It also counts. Prescribed exercise adherence is routinely over-reported, and a sensor record replaces self-report with data 1.

Home exercise programme monitoring, post-surgical range and progression tracking, sports rehabilitation, and any programme where exercise technique determines whether the exercise works — which is most of them.

It is particularly useful in the transition from supervised to independent exercise, which is where most programmes lose their effect.

Motor learning depends on knowing the result of a movement close to when it happened. Feedback given at the next appointment, a week later, does very little.

Real-time feedback also identifies the patient who is doing the exercise diligently and incorrectly — a common and frustrating pattern that looks like non-response to treatment.

Sensors record how much load was actually applied, how many repetitions were completed and how technique changed as fatigue set in. Self-reported exercise volume is unreliable, and how far it strays from the truth varies widely from person to person 1.

The value is in progression: an exercise programme that is not measurably harder in week four than in week one is not a programme, and the data makes that visible.

Two things, both of which are otherwise guesswork. Whether the prescribed load was actually used — people habitually work below what was set once nobody is watching — and whether technique held as fatigue arrived.

The second matters more than it sounds. Repetitions performed with deteriorating form after the sixth set are not the repetitions that were prescribed, and they load joints and tissues in ways the programme did not intend.

Accurate enough to track range and repetitions, within limits. A systematic review of 42 studies comparing wearable sensors with laboratory motion capture found that accuracy is higher for simple movements than for complex ones, and better at the joints of the legs; far fewer studies have tested whether readings repeat reliably from one session to the next 2. Sensor readings are therefore most trustworthy when the same exercise is measured in the same way from week to week.

Whether feedback from sensors improves recovery is a separate question. After stroke, a review of 13 randomised trials found moderate-quality evidence that real-time sensor feedback during walking training improves balance and functional mobility, but no benefit for activities of daily living 3. The trials were small, with 304 participants in total 3. The feedback is a training aid; the exercise programme remains the treatment.