Needs and requirements of a Health Data Ecosystem for Alzheimer prediction
Medical technology is evolving at a rapid pace, yielding increasing amounts of multi-modal, patient-specific data. Hospital information systems have meticulously archived these health records and are now boasting decades worth of patient data. The availability of this latter data, combined with advanced analysis techniques originating from the field of artificial intelligence, has enabled novel approaches to medicine, in which preventive strategies are increasingly employed, therapeutic decisions are taken considering predictive models of outcome, and administered therapies are ever more personalized. Despite these promises, this resource remains largely underexplored in research due to intrinsic difficulties in accessing and collecting the data.
The DIMENTIA project, funded by Innoviris, focuses on the prediction of cognitive evolution in Alzheimer’s disease patients based on multimodal patient characteristics (neurophysiology, neuroanatomy, cerebrospinal fluid, …). It explores how UZ Brussels patient data might be made available, used and governed within a “digital health platform” to reach the above goals. DIMENTIA is a collaboration between Collibra, UZ Brussel and the ETRO and SMIT – VUB research groups
The role of SMIT
With a focus on stakeholder research, SMIT looks into the (functional) needs and requirements of (a) researchers interested in appropriating this valuable hospital data, (b) healthcare professionals who log this data on a daily basis, and (c) patients and if/how they wish to have agency within such a health data ecosystem.