Thomas M.H. Hope

This study benchmarks predictive models for long-term stroke outcomes using a global consortium, revealing mechanistic diversity and key predictors across motor, cognitive, and emotional domains.

This study shows that using continuous lesion images improves outcome prediction after stroke compared to binary lesion images.

The paper uses dynamic causal modelling to study effective connectivity during auditory repetition, finding bidirectional and inhibitory interactions between Wernickes and Brocas areas with individual variability.

This study develops a prognostic model predicting treatment response for anomia after stroke using lesion location and therapy time.

The paper simulates functional recovery from different lesion types using a generative model, showing that complete disconnection of a system invokes the alternative system, while partial damage has more severe long-term effects.

This study investigates brain-behavior relationships in post-stroke anomic aphasia, finding that higher-order cognitive functions, not just language-specific areas, are linked to language production deficits.

The paper challenges the proportional recovery rule by showing that baseline score correlations can mislead about outcome predictability in stroke recovery.