A machine learning model based on the post-stroke cognitive impairment network can predict long-term cognitive impairment after minor stroke.
Why this matters
Although many stroke sufferers experience long-lasting cognitive impairment, accurate long-term predictors of diminished cognition are lacking. These findings suggest that machine learning algorithms that search for combinations of features can reliably predict cognitive decline based on clinical and imaging data. This model can therefore accurately predict subsequent cognitive decline post-stroke and enable interventions in at-risk individuals.