The National Institutes of Health is continuing support of research at the Wayne State University School of Medicine that will investigate a novel benefit-risk analysis for the presurgical evaluation of pediatric drug-resistant focal epilepsy.
The method being explored utilizes state-of-the-art artificial intelligence techniques called deep-learning neural network to provide an optimal resection boundary of the epileptogenic brain tissue, resulting in the maximized benefit – seizure freedom – and minimized risk to motor skills, language, hearing and vision.
Associate Professor of Pediatrics and of Neurology Justin Jeong, Ph.D., is the principal investigator on the project, launched in 2015. The NIH’s National Institute of Neurological Disorders and Stroke has renewed the study for another five years, awarding WSU $2.12 million to support “Novel DWI Methods to Minimize Postoperative Deficits in Pediatric Epilepsy Surgery” through June 30, 2026.
“We will also study if the integration of deep-learning network and diffusion MRI connectome helps decide timely surgery by providing preoperative imaging markers underlying a high likelihood of postoperative neurocognitive improvements and mechanistic insight in structural brain reorganization associated with the corresponding improvements,” Dr. Jeong said. “We expect that the results of this project will translate deep-learning network-based diffusion MRI techniques to optimize the surgical margin for pediatric epilepsy, predict the postoperative neurocognitive outcome and determine the specific mechanism of postoperative brain reorganization.”
Surgical resection of the epileptogenic brain tissue is an excellent option for children whose seizures do not respond to medical therapy. The ultimate goal of pediatric epilepsy surgery is to achieve long-term seizure freedom (often achieved following the complete resection of the epileptogenic zone) and neurocognitive improvement (a secondary benefit often associated with postoperative seizure freedom) with minimal postoperative deficits in eloquent functions, including primary motor, language, hearing and vision.
“Unfortunately, referrals for epilepsy surgery are often delayed until it is too late to prevent the detrimental psychosocial effects of refractory seizures. Many children who are offered surgery have had drug-resistant epilepsy for half of their lives and thus been exposed to potential risks of epilepsy-related morbidity/mortality such as diminished memory and cognition, poor general health, injuries, sudden unexpected death, anxiety, hopelessness and suicidal thoughts,” Dr. Jeong said.
His co-investigators include WSU faculty members Professor of Pediatrics, of Neurology and of Neurosurgery Csaba Juhász, M.D., Ph.D.; Professor of Pediatrics and of Neurology Eishi Asano, M.D., Ph.D.; Assistant Professor of Pediatrics Michael Behen, Ph.D.; Professor of Computer Science Ming Dong, Ph.D.; and Professor of Neurosurgery Sandeep Sood, M.D. His co-investigator team includes Associate Professor of Neurology Aimee Luat, M.D., who is also a faculty member at Central Michigan University; and Fang-Chen Yeh, M.D., Ph.D., of University of Pittsburgh.
The researchers believe that the stigma and reluctance toward surgery are caused by demarcation of the exact resection boundary to optimize the benefit-risk ratio, which is challenging in pediatric epilepsy surgery, partly because invasive electrical stimulation mapping, or ESM, the current gold-standard in epilepsy surgery, often fails to localize eloquent areas in children. This low sensitivity inevitably increases the fear of postoperative deficits. Moreover, ESM does not provide quantitative information as to whether a patient’s current brain development supports a high likelihood of long-term neurocognitive improvement following surgery. Thus, the secondary benefit of possible academic improvements following surgery may not sufficiently motivate parents to select surgery early, Dr. Jeong said.
The placement of intracranial electrodes inevitably poses risks of surgical complications as well, leading to permanent morbidity or even death, in up to 8% of procedures, he said, and may require additional treatment.
“Taken together, there are clinical needs for more structured, quantitative and non-invasive tools capable of helping families and providers make early surgical decisions by accurately providing the probabilities of both favorable and unfavorable postoperative outcomes from preoperative imaging analysis at the whole-brain level; replacing routine invasive ESM procedures by non-invasively providing accurate locations of intact brain pathways; and saving substantial time, effort and cost of the presurgical evaluation,” Dr. Jeong added. “Especially, these tools would help reduce the secondary risk of surgical complications in children with multiple epileptogenic zones. Our project approaches these clinical needs by developing a deep learning-based imaging tool that enables a quantitative benefit-risk analysis before epilepsy surgery.”
Dr. Jeong is a magnetic resonance imaging physicist and biomedical image/signal processing scientist with a background in MRI-oriented biomedical engineering, machine learning, neuroscience and neuroimaging research methodologies.
“I was gratified by the fact that NIH re-recognized the value of my commitment to improving the quality of life in young children suffering from epilepsy. During the first funding period, I had been fortunate to be in a stable position with excellent collaborators throughout our campus. I am now passionate about the high potential of this renewal grant to help many families and their kids who will undergo epilepsy surgery in the future,” he said. “My current grant project shows how important the collaboration of multidisciplinary expertise is for our students to incubate new fellowship awards and learn to develop their independent research projects.”
The renewal grant is sponsoring the NIH F30 fellowship grant of his mentee Nolan O’Hara, a School of Medicine M.D./Ph.D. candidate who aimed to investigate a novel multimodal imaging approach that can characterize the propagation dynamics of infantile spams using intracranial EEG and diffusion MRI. O’Hara will finish his doctorate this semester and then complete medical school.
The five-year award from the NIH is 2R01NS089659.