Physiosolution Clinical Team · 2026-10-20 · 3 min read
How it works
Most AI movement analysis uses pose estimation. The software looks at video and identifies key points on the body, such as shoulders, hips, knees and ankles. From those points, it calculates joint angles, movement speed and symmetry.
Lab-based systems still offer the highest accuracy. Markerless camera systems trade some accuracy for convenience and cost.
What it's good at
Large, clear movements filmed from a good angle: knee bend during a squat, shoulder lift, trunk lean during walking, symmetry between left and right sides. For tracking change over time with the same setup, markerless systems can be quite useful.
They also produce visual feedback that patients understand. Seeing your knee angle on a screen can make a correction click faster than a verbal instruction.
What limits accuracy
- camera angle: movements across the camera's line of sight are harder to measure
- clothing: baggy clothes hide joint positions
- lighting and background clutter
- small joints and rotations, such as foot and wrist movements
- fast movements, where video frames blur
Accuracy also varies between products. A system that works well for knee bend may be less reliable for hip rotation.
Measurement isn't diagnosis
Knowing that a knee collapses inwards when landing is valuable. Understanding why requires assessment: strength testing, joint examination, history and watching how the person responds to changes. Movement analysis is one input among many.
Questions to ask
If a clinic or app uses AI movement analysis, reasonable questions include:
- what is it measuring, and how accurate is it for that movement?
- how will the results change my treatment?
- who reviews the results?
- where is my video stored, and who can access it?
Privacy matters. Video of your body is personal data.
A simple example
Imagine someone recovering from knee surgery doing squats at home. A movement analysis app might show that their knee bends to 80 degrees on the operated side and 110 on the other, and that their weight shifts towards the good leg. Over four weeks, it might show the gap narrowing.
That's useful. It gives objective feedback, shows progress and highlights a compensation pattern the patient may not notice.
What the app can't do is decide why the weight shift happens. It might be weakness, pain, fear or stiffness. Each needs a different response. That's where the physiotherapist's assessment comes in, using the data as a starting point rather than a conclusion.
Where it fits
Used well, movement analysis helps physiotherapists measure, explain and track change. It works best alongside clinical assessment, not instead of it. Our article on smartphone cameras for monitoring exercises looks at the home side of the same technology, and our AI recovery tracking guide explains what else can be measured.
Talk to a physiotherapist
Clinic in Indirapuram, home visits across Indirapuram, Vaishali, Vasundhara, Noida and Ghaziabad. Slots confirmed on call or WhatsApp.