MS-PINPOINT · UCL Queen Square
We build the computational tools that read the scans and records hospitals already hold — across many hospitals, internationally — so that what happens to one person with MS can inform the care of the next.
The scale of the question
Two studies
Bringing together routine imaging and clinical data from twelve hospitals to work out what actually predicts how multiple sclerosis progresses in an individual.
This is the study with an opt-out: if you are a patient at one of the twelve hospitals, it is the one to act on.
Our next study, newly funded and beginning when the group moves to KCL. It widens the network to sixteen hospitals and takes a new direction: AI that understands free text and conversations with patients, for a future in which people hold a central role in their own care, supported by doctors both human and AI.
It has not started, so there is nothing to opt out of yet and no data is being collected for it.
If you are a patient at one of the twelve MS-PINPOINT hospitals and do not want your data included, the opt-out above is the one you want. DREAMS has not started and has no opt-out yet.
Software
Hospital MRI archives hold decades of imaging that was never analysed at scale, because the scans were acquired for clinical care rather than research. MindGlide is our deep-learning tool for measuring them anyway — turning an archive into a dataset.
Talks & lectures
Deck notes from conferences, invited lectures and teaching—newest first.
Latest news
