Methods for the computer-assisted diagnosis and prognosis of neurodegenerative diseases using computational anatomy, genetic imaging and deep-learning

Neurodegenerative diseases represent a large group of neurological disorders with heterogeneous clinical and pathological expressions affecting specific subsets of neurons in specific functional anatomic systems. Research attention has been focused on Alzheimers, Parkinsons, Huntington, and amyotrophic lateral sclerosis. This is probably because these diseases are more frequent, arise for still unknown reasons, and progress in a relentless manner towards a devastating cognitive condition. Research against neurodegeneration is approached working in the development of ways to diagnose the cause of the neurodegeneration as early as possible. Regretfully, the development of effective preventive or protective therapies has been impeded by the limitations of our knowledge of the causes and the mechanisms underlying neurodegenerative diseases. We need predictive biomarkers based on imaging and genetics together with accurate predictive models of the rates of cognitive decline in those who exhibit preclinical, prodromal, or clinical disease.

The purpose of this project is to develop computational tools for computer-aided diagnosis and prognosis of neurodegenerative diseases. This project focuses on the development of computational techniques useful in the quest of predictive biomarkers, the selection of the kind of neurodegeneration given the current patient condition, and the prediction of questions important in the assessment of disease evolution. These problems are approached using deep-learning with a focus on the comparison with conventional machine learning achievements and the interpretability of the models. In this project, we move towards personalized medicine arena, with the objective of finding reliable and stable biomarkers that, combined with powerful computational systems, will provide high sensitivity and specificity in single individuals towards the creation of patient-specific profiles for a precise assessment of the risk of disease onset, disease evolution, and response to treatment.

Referencia PID2019-104358RB-I00