Database

The dataset that isn’t there yet

A large, representative, labelled clinical movement dataset does not exist today. That sounds unlikely in a field where everyone talks about data, but it is true: the only way to produce that data at the moment is a gait lab, and a gait-lab recording costs tens of euros per measurement. That puts it out of reach for almost every physiotherapy practice and outside most researchers’ project budgets. The result is that the analyses everyone would want to run (age-based norms, recovery patterns after knee surgery, reference data per diagnosis) are simply impossible because the underlying data doesn’t exist.

Movalytics costs cents per recording. The question is no longer whether the data will exist, but how to collect it cleanly. Our proposal is simple: practices and researchers use the tool for free, and we get to use the anonymised movement parameters from every recording to build a dataset. No video, no patient identifiers, just angles and parameters. The clinician gets an objective measurement; we get the data to build the normative references the same clinician benefits from. Everyone is digging for gold; we sell the shovel and keep the spoil heap.

For researchers with a concrete question — back pain, post-surgical recovery, fall prevention, sport — we analyse existing video data for free, help with the processing and publish together. Seriously designed graduate projects count too; students with a rigorous protocol are researchers as far as we are concerned.