Brain connectivity measures improve prediction of functional outcome after acute ischemic stroke
Sofia Ira Ktena
Biomedical Image Analysis Group, Imperial College London, London, UK, Stroke Division & Massachusetts General Hospital, J. Philip Kistler Stroke Research Center, Harvard Medical School, Boston, USA
Corresponding Author:[email protected]
Author ProfileMarkus D. Schirmer
Department of Population Health Sciences, German Centre for Neurodegenerative Diseases (DZNE), Germany, Computer Science and Artificial Intelligence Lab, Massachusetts Institute of Technology, Boston, USA, Stroke Division & Massachusetts General Hospital, J. Philip Kistler Stroke Research Center, Harvard Medical School, Boston, USA
Abstract
The ability to predict long-term functional outcomes in the acute setting of stroke represents a highly clinically relevant problem that is currently unresolved. The field of connectomics exploits imaging and processing techniques to represent the brain's structural and/or functional connectivity as a graph. Topological properties of these brain graphs have been extensively explored, shedding light on the underlying mechanisms of brain function in health and disease. In the human connectome, a rich club organization serves as a high capacity backbone system critical for physiological neuronal connectivity. We assess the impact of ischemic stroke insults to the brain regions that constitute this rich club backbone and topological properties of functional brain networks on brain recovery in a hospital-based cohort of 41 acute ischemic stroke (AIS) patients. We demonstrate that an improved predictive model of post-stroke patient functional outcome can leverage information about brain network topology and yield a 5-fold increase in explained variance for the 90-day outcome.