Artificial intelligence

As a group we are looking into ways of how recent advances in artificial intelligence (AI) can be leveraged into clinical healthcare.

AI in psychiatry

One of my first explorations of the use of AI in clinical care was to assess the usefulness of different machine learners (SVMs) to detect the metabolic syndrome in patients treated with second-generate psychotics. In contrast to the literature by then, I demonstrated that the accuracy of the different machine learning strongly depended on the population at study as a specific subpopulation was easily classified, artificially increasing accuracy. See my paper for more details.

Will AI change MS care in the next ten years?

We have recently contribute to both the YES and NO statement in an MSJ controversy series on whether AI will change MS healthcare within 10 years.

Federated learning

We strongly believe that the future of AI does not lie within large centralised databases, but rather that we have to leverage clinical datasets through federated learning. A first proof-of-concept paper by Stjn Denissen can be found here.

Key-words: AI; federated learning; healthcare in MS