The Journal of Neural Engineering published research led by the University of Minnesota Medical School that evaluated the reliability of human experts compared to an automated algorithm in assessing the quality of intracranial electroencephalography (iEEG) data. This research aims to improve seizure detection and localization for epilepsy patients. iEEG is a procedure that measures brain activity by placing electrodes directly on or inside the brain, providing crucial information for diagnosing and treating conditions like epilepsy.
The research team enlisted 16 experts, including EEG technologists and neurologists, to rate 1,440 iEEG channels as “good” or “bad,” where good implied recording brain activity and bad implied not recording brain activity. Their evaluations were compared to each other and against the Automated Bad Channel Detection (ABCD) algorithm developed at the University of Minnesota. The ABCD algorithm demonstrated higher accuracy (95.2%) and better overall performance in identifying channels with high-frequency noise compared to human raters. This highlights potential biases and limitations in human-based EEG assessments and suggests a future where automated methods can aid clinicians in improving the accuracy and efficiency of seizure detection.
Alexander Herman, MD, PhD, an assistant professor at the U of M Medical School, emphasized the importance of the ABCD algorithm’s performance in enhancing patient care by improving seizure detection and ultimately patient outcomes. The research underscores the potential of automated solutions to enhance the reliability and efficiency of iEEG data interpretation, which is critical for seizure localization and improved patient outcomes. By reducing the workload and variability in assessments, automated algorithms can help clinicians focus more on clinical decision-making and patient care.
David Darrow, MD, MPH, an assistant professor at the U of M Medical School and neurosurgeon with M Health Fairview, highlighted the potential of automated algorithms to outperform human experts in identifying bad EEG channels. He emphasized the importance of refining these automated methods further and exploring their application in real-time clinical settings. Future research should aim to enhance these automated methods and assess their effectiveness in improving seizure detection and patient care outcomes.
The study received funding from the Institute for Translational Neuroscience and MnDRIVE Brain Conditions. This research showcases the potential of automated algorithms in improving the accuracy and efficiency of iEEG data interpretation for epilepsy patients. By leveraging automated methods to aid human raters in identifying bad EEG channels, clinicians can enhance their focus on clinical decision-making and improve patient care outcomes. Future research should continue to refine and explore the application of automated methods in real-time clinical settings to further enhance seizure detection and localization.