Abstract #0048

Connectivity Domain Analysis of Mild Traumatic Brain Injury: A Multi-center Study to Extract Robust Imaging Biomarkers

Armin Iraji¹, Jiachen Zhuo², Natalie M. Wiseman³, Ali-Reza Mohammadi-Nejad⁴, Rao Gullapalli², Zhifeng Kou^1,5, and E. Mark Haacke^1,5

¹Department of Biomedical Engineering, Wayne State University, Detroit, MI, United States, ²Diagnostic Radiology and Nuclear Medicine, University of Maryland, ³Department of Psychiatry and Behavioral Neurosciences, Wayne State University, Detroit, MI, United States, ⁴Henry Ford Health System, ⁵Department of Radiology, Wayne State University

Identification of biomarkers for mild traumatic brain injury (mTBI) diagnosis and outcome prediction is challenging due to the heterogeneity of mTBI patients. Multi-center studies help to alleviate this, but functional MRI data can be difficult to combine across sites. Here, we applied our recent connectivity domain (CD) framework and indentified predictive features of mTBI diagnosis and one-month outcome. Despite high heterogeneity of predictors between and within sites, classification accuracy did not suffer due to combination of datasets. Further multi-center analyses may benefit from use of the CD for generation of classification and outcome prediction models.

This abstract and the presentation materials are available to members only; a login is required.

Join Here