A paper of the Neuroimage Analysis & Machine Learning group was recently published in Neurology Open Access!
In this work, Myrthe van Haaften (first author) and Esther Bron (last author) clustered MRI-derived brain volumes in patients with dementia, in order to identify atrophy patterns. They evaluated how the resulting imaging clusters related to the dementia type and other clinical characteristics. The clustering was performed on data from the Alzheimer Center Erasmus MC, and externally validated on the OASIS-4 memory clinic cohort from Washington, USA.
The findings show that that different dementia types can present with overlapping imaging patterns. Also within dementia types, a large spectrum of atrophy patterns can be present.
The open-access article can be found here.
