Mapping Harmful Brain Activity in the ICU
M. Brandon Westover
November 16, 2022, Wednesday, 3:00 PM - 4:00 PM EST
Electroencephalography (EEG) in patients who are severely ill often reveals patterns of brain activity that resemble seizures, even though these patients rarely have convulsions or other outward clinical signs of seizures. This activity is the subject of a decades long debate: If seizure-like brain activity harms the brain then it may warrant aggressive treatment; if such activity is simply an epiphenomenon, then aggressive treatment likely contributes to poor clinical outcomes. I will describe how we have begun to resolve this debate by applying machine learning to analyze massive clinical data sets.
M. Brandon Westover, MD, PhD obtained a PhD in Physics, working in the field of information theory and computer+biological vision. He is board certified practicing Neurologist and a clinical neurophysiologist / epileptologist at Massachusetts General Hospital (MGH). He directs the MGH Critical Care EEG Monitoring Service and the MGH Clinical Data Animation Center (CDAC). His research uses "Big Data" and develops Artificial Intelligence / Machine Learning methods to improve medical care for patients with anoxic brain injury, seizures and seizure-like brain states, cerebral ischemia, delirium, and sleep disorders, and to develop closed-loop control technology for precision control of anesthesia in the ICU.