Novel network analysis technology can pinpoint seizure originating brain regions in minutes
New techniques to aid in seizure diagnosis and surgical planning benefit millions of patients with epilepsy, but the road to progress has been slow and difficult. New research by Bin He of Carnegie Mellon University and his team, in collaboration with the UPMC and Harvard Medical School, introduces a new network analysis technology that uses minimally invasive electrophysiological resting records to locate regions seizures and predict the results of seizures.
Epilepsy affects about 70 million people worldwide and more than 3.4 million Americans. About a third of those affected cannot be treated with drugs alone. For these patients, surgical removal of the tissues caused by seizures or neuromodulation procedures are possible avenues of treatment in order to maintain quality of life.
In current practice, before any surgical removal of tissue, doctors often drill holes in the skull to place recording electrodes on top of the brain. The electrodes record electrical activity in the brain over days or weeks, no matter how long it takes for the seizures to materialize, to report where the seizures are occurring. While necessary, this practice can be time consuming, costly, and inconvenient for patients to stay in the hospital for days or weeks.
He and his collaborators have developed an alternative to the current and recently published clinical routine in Advanced science. His new network analysis technique can identify the brain regions that originate from seizures and predict the outcome of a patient’s seizures before surgery, using only 10 minutes of rest-free recordings. need to wait for seizures to occur.
In a group of 27 patients, our accuracy in locating seizure-initiating brain regions was 88%, which is a fascinating result. We use machine learning and network analysis to analyze a 10-minute sleep state log to predict where the seizure will occur. While this method is still invasive, it is to a lesser extent, because we’ve taken the recording timeline of several days or even weeks to 10 minutes. “
Bin He, Professor of Biomedical Engineering at Carnegie Mellon University
He continued: “In the same group of patients, our accuracy in predicting the outcome of seizures, or the possibility of being free of seizures after surgery, was 92%. Finally, this type of data could guide patients to surgery or move away from surgery, and this information is not available today. “
The technique extracts the information flow through all the recording electrodes and makes a prediction based on the different levels of information flow. He and his colleagues found that the flow of information from the non-seizure tissue to the original tissue of the seizures is much larger than the reverse direction, and the difference significantly in the flow of information it often leads to a seizure-free result. Once implemented, this approach could have a significant impact on informing physicians and families whether a patient should undergo surgery and what the likelihood of surgical success would be.
Helping patients remains the driving force and overall goal. By focusing on non-invasive and minimally invasive approaches, he believes that both the patient and the healthcare system can benefit.
“This research will not only provide information on the likelihood of surgical success in people with epilepsy and their caregivers, but will also help us understand the underlying mechanisms of seizures through a minimally invasive approach,” said Vicky Whittemore, Ph.D. .D. ., director of the program, National Institute of Neurological Disorders and Stroke, part of the National Institutes of Health.
Faculty of Engineering, Carnegie Mellon University
Jiang, H., et al. (2022) SEIT interictal rest state connectivity locates the onset of the seizure and predicts the outcome of the seizure. Advanced science. doi.org/10.1002/advs.202200887.