Predicting language outcome and recovery after stroke. (360G-Wellcome-097720_Z_11_Z)

The main goal is to provide proof of principle for a clinically useful tool that will predict recovery of language abilities after stroke. There are three separate but interacting lines of research, to be conducted in parallel: (A) Using structural neuroimaging and behavioural assessments, we will identify lesion and non-lesion factors that are most likely, and least likely, to cause long term communication difficulties. The influence of the identified factors will be tracked over months and years following the stroke to understand how recovery unfolds over time. (B) Using functional neuroimaging and dynamic causal modelling, we will map the brain areas that are activated when the patient is speaking or comprehending speech. This will provide a mechanistic understanding of the language pathways that are able to support recovery after damage to the normal system. (C) We will develop an easy-to-use web-based system that will provide patients, their carers and clinicians with t he likely time course of recovery on the basis of probabilistic summaries of other patients in our database who have corresponding brain damage according to the carefully constrained criteria developed in (A).