Using electronic health records to identify computed tomography brain features associated with post-stroke dementia:...

Area of Science:

• Neurology
• Radiology
• Data Science

Background:

• Investigating the link between acute stroke imaging and post-stroke dementia.
• Assessing the feasibility of integrating research data with electronic health records.

Purpose of the Study:

• To identify acute stroke computed tomography (CT) brain features associated with post-stroke dementia.
• To evaluate the utility of electronic health record linkage for dementia outcome research.

Main Methods:

• Linked data from two stroke research datasets with expert-coded CT brain scan features.
• Utilized electronic health records for participant follow-up.
• Performed survival analyses to identify dementia risk factors.

Main Results:

• Identified 21 participants (11%) with post-stroke dementia after a median follow-up of 9.7 years.
• Cerebral atrophy on acute CT scans was linked to increased dementia risk.
• Moderate-to-severe white matter hyperintensities on acute CT scans were associated with higher dementia risk.

Conclusions:

• Electronic health record linkage is a viable method for studying post-stroke dementia.
• This approach can be scaled to larger stroke cohorts for detailed imaging predictor analysis.