Ageing-related disorders, and in particular neurological disorders, represent important challenges faced by our health systems. We need to advance early detection and diagnosis, intervention, treatment as well as the information we provide the patients and their families. We propose to use state of the art deep-learning approaches to improve detection and diagnosis of Alzheimer’s and Parkinson’s disease (AD & PD).
For AD & PD, deep-learning algorithms have not yet superseded other predictive approaches, which we attribute to insufficiently large training samples. Here, we would like to leverage the large-scale brain MRI collection from the UKBiobank to build a model of ageing and of those neurological disorders.
Despite having the highest computational cost at training, deep-learning algorithms may not require any MRI image processing, meaning they could be more easily implemented in the clinics, to assist diagnosis and evaluate response to treatment.
