Unmet Need
Epilepsy is a devastating disease that affects over 50 million people globally. 70% of patients diagnosed with epilepsy can treat their symptoms with medication, but around 30% of patients have symptoms that do not respond to drugs. Recent studies have found that the most effective treatments for drug resistant epilepsy are interventions that surgically remove or electrically interrupt seizure initiation in the epileptogenic zone (EZ). The EZ refers to the network of regions in the brain from which seizure activity is triggered. However, due to lack of information and proper understanding of the anatomical extent of the EZ, surgical success rates of these interventions are quite low, varying from 30%-70%. Given that total diagnostic and surgical treatment costs are at least $200,000 per patient, these low success rates create a significant clinical and economic burden, and many surgical candidates opt out of this potentially curative procedure. Thus, there is need for a method that can provide better localization and guidance of the EZ zone for increased surgery success rates.
Technology Overview
Johns Hopkins inventors have created a computational algorithm that utilizes invasive EEG (iEEG) recordings to identify the seizure onset zone in focal epilepsy patients. The algorithm utilizes dynamic network models (DNMs) to leverage iEEG data for each patient to localize the EZ. DNMs are estimated from interictal (between seizure) data which will significantly decrease the required invasive monitoring time, as current clinical practice relies predominantly on data captured during seizure events. The properties derived from each DNM that have correlation to the EZ are then developed, validated, and benchmarked against state-of-the-art algorithms on a large population of drug resistant epilepsy patients. This tool has the potential to facilitate more accurate localization of the seizure onset zone by providing new information from largely ignored interictal data, while minimizing patient risk factors and hospital costs by decreasing required monitoring time.
Stage of Development
The algorithm is being developed and undergoing testing.
Publication
Kristin M Gunnarsdottir, Adam Li, Rachel J Smith, Joon-Yi Kang, Anna Korzeniewska, Nathan E Crone, Adam G Rouse, Jennifer J Cheng, Michael J Kinsman, Patrick Landazuri, Utku Uysal, Carol M Ulloa, Nathaniel Cameron, Iahn Cajigas, Jonathan Jagid, Andres Kanner, Turki Elarjani, Manuel Melo Bicchi, Sara Inati, Kareem A Zaghloul, Varina L Boerwinkle, Sarah Wyckoff, Niravkumar Barot, Jorge Gonzalez-Martinez, Sridevi V Sarma, Source-sink connectivity: a novel interictal EEG marker for seizure localization, Brain, Volume 145, Issue 11, November 2022, Pages 3901–3915, https://doi.org/10.1093/brain/awac300
