mAIcompass · Multimodal AI-based Clinical prOgnostic Models for PArkinson's Symptoms progression and Subtyping

Parkinson’s Disease (PD) is the fastest-growing neurological condition with global cases estimated to double to 20M by 2050. The spectrum, onset and progression of PD symptoms are highly heterogeneous between individuals, hampering treatment personalization through early prediction of disease course. I aim to develop mAIcompass; a multimodal AI-based clinical application, that leverages longitudinal clinical and imaging data, acquired during follow-up examinations, to facilitate real-time prognostication of PD progression, in terms of motor function deterioration, cognitive decline and neuropsychiatric impairment. It will also aid the early prediction of the overall progression subtype, based on neurobiological phenotypes of PD subgroups with similar progression pattern, identified through temporal clustering. The fellowship will be hosted by the Neuropsychiatry team of Amsterdam UMC that leads ENIGMA PD consortium, the largest existing international collection of PD datasets. I will leverage advanced expertise to develop novel algorithms for longitudinal PD data modelling and subtyping at the Center for AI and Data Science for Integrated Diagnostics, University of Pennsylvania (outgoing phase). Upon validation of expert clinicians and engineers, I will integrate the AI models into a user-friendly standalone application software intended for educational, research and experimental clinical use. Feedback for its utility will be obtained from end-users and patients focus groups at a PD clinical center. These tasks will set a roadmap for mAIcompass future clinical implementation, that is expected to aid the personalized PD clinical management, towards maximizing patients’ treatment benefit and reducing healthcare costs, proportionally varying with PD progression. mAIcompass aspires to link its expected scientific, societal and economic impacts to the Parkinson’s Manifesto for Europe, as well as to EU priorities emphasizing promotion of citizens' health and well-being.