Project Details
Projekt
Automated longitudinal characterization of the choroid plexus and glymphatic system in multiple sclerosis: facilitation of biomarker extraction and improvement of patient stratification
Applicants
Dr. Gabriel Gonzalez Escamilla, Ph.D.; Professor Dr. Sergiu Groppa; Dr.-Ing. Andrea Kronfeld; Dr. Anirban Mukhopadhyay, Ph.D.; Professor Dr. Ahmed Othman
Subject Area
Nuclear Medicine, Radiotherapy, Radiobiology
Term
since 2023
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 515302522
Multiple sclerosis (MS) is the most common chronic inflammatory and highly debilitating degenerative disease affecting the central nervous system. It leads to progressive disability in young adults and a severe life quality impairment. The treatment of MS is highly costly for both the affected individuals as well as the entire health system. Brain imaging plays a central role in the diagnosis of MS and might also grant a vast amount of quantitative information which may facilitate patient stratification for therapeutic decisions and outcome prediction. However, this is insufficiently used for clinical purposes so far. Current state-of-the-art clinical imaging mainly assesses MS lesions and atrophy but does not yet take advantage of key structures involved in neuroinflammatory pathways. Very recent studies demonstrated the central role of the choroid plexus (ChP) and the glymphatic system in ongoing neuroinflammation and neurodegeneration in MS as well as in several other neuropsychiatric disorders. Robust algorithms for the delimitation of these structures are however lacking. The aim of the herein project is to develop an automated unified workflow for quantification of choroid plexus, perivascular spaces (as part of the glymphatic system) as well as lesion topology analysis. This will be achieved by using explainable artificial intelligence (AI) models in patients with MS which will be utilized for the prediction of clinical trajectories and for patient stratification. Briefly, we will employ AI to ensure an integrative radiologist-in-the-loop framework tailored for segmentation, together with a downstream alignment analysis on three large multi-institutional cohorts (Mainz, Düsseldorf, TU München) in order to relate the imaging information to clinical parameters. To ensure synergetic effects, a strong collaboration with other partners of SPP2177, will be established to facilitate sharing of segmentation methods and validation data. With this approach, we postulate that the integrative analysis of ChP characteristics, glymphatic system, and lesion topographies will lead to the development of new imaging biomarkers that remarkably leverage clinical scenarios that are fundamental for the treatment of neuroimmunological and neurodegenerative disorders.
DFG Programme
Priority Programmes
