Neurologists develop software application to help identify subtle epileptic lesions

"Our method automatically identified abnormal areas in MRI scans in 92 percent of the patients sampled, which were previously identified by expert radiologists reviewing multiple images," said first-author Thomas Thesen, PhD, assistant professor, Department of Neurology, NYU Langone Medical Center. "Based on these findings, we will focus on the ability of our application to detect the more subtle epileptic malformations that are not easily detectable by the human eye. We believe this could lead to new tools to greatly help radiologists provide more accurate and faster results with objective measures for standardizing readings."

The proof-of-concept study, entitled "Detection of epileptogenic cortical malformations with surface-based MRI morphometry," demonstrates that non-invasive and automated detection of known epileptogenic structural abnormalities in cortex is possible, and supports its potential use as a tool for diagnosis and planning of epilepsy surgery.

The researchers are encouraged by the initial results and have already started evaluating the applications ability to determine undetected lesions in previously negative MRI scans, with findings to be published later this year.

The study's other authors include Chad Carlson, Brian T. Quinn, Orrin Devinsky, Jonathan DuBois, Jacqueline French, Olga Felsovalyi, Xiuyuan Wang and Ruben Kuzniecky from NYU Langone Medical Center; Carrie R. McDonald and Eric Halgren from the University of California, San Diego and Richard Leventer from the Royal Children's Hospital, Murdoch Children's Research Institute, University of Melbourne, Australia.