Natus Medical Incorporated has announced the launch of autoSCORE, the first-of-its-kind artificial intelligence model capable of automatic and comprehensive clinical EEG interpretation, providing accuracy on par with medical experts.
The autoSCORE application was developed in Norway by Holberg EEG using a deep-learning model and trained on the world’s largest dataset of more than 30,000 expertly labeled EEG recordings. The development and validation of the model have been carefully designed to avoid typical errors and sources of bias, and the output is focused on aspects key to the interpretation of clinical EEGs [1].
“The development of the world’s first AI model capable of reading clinical EEGs at the level of the best experts in the world is a significant accomplishment,” says Harald Aurlien, co-founder of Holberg EEG and a consultant in clinical neurophysiology at Haukeland University Hospital. “The partnership and distribution agreement with Natus, one of the world’s largest EEG companies, will help bring this innovative solution to customers around the world.”
The 2023 study “Automated Interpretation of Clinical Electroencephalograms Using Artificial Intelligence,” published in JAMA, revealed autoSCORE consistently performed with accuracy, sensitivity, and specificity near or above 90% [1]. It also showed the autoSCORE application to be on par with leading human experts while outperforming similar AI models on the market.
“The autoSCORE application has the potential to improve clinical outcomes and value of care for the millions suffering from epilepsy,” states Natus Neuro Chief Executive Officer Chris Landon. “This newly launched solution represents the first of what we expect will be a series of innovative AI applications at Natus designed to deliver efficiency and quality improvements across healthcare.”
Visual EEG review is considered the gold standard for determining abnormalities, but it demands clinical resources and years of specialized training. The AI-enabled autoSCORE application is designed to allow neurocare teams to more efficiently and consistently perform EEG data analysis.
Unlike traditional spike and seizure detectors, the autoSCORE application comprehensively assesses EEG data for multiple clinically relevant abnormalities and provides study-level assessments to inform care teams whether an EEG study is normal or abnormal.
The autoSCORE application has received U.S. Food and Drug Administration clearance for use with routine EEG studies and is currently available in the United States, exclusively with Natus NeuroWorks EEG Software Version 10.
Explore AITechPark for the latest advancements in AI, IOT, Cybersecurity, AITech News, and insightful updates from industry experts!
