×
×
Neurologists and epileptologists increasingly rely on AI to perform more rapid EEG analysis to help detect and prioritize abnormalities in far less time. As the leader in this category, Natus autoSCORE is the world’s first AI-based model that provides comprehensive clinical EEG interpretation, with accuracy on par with medical experts.
The intended users are suitably trained professionals who are qualified to clinically review EEG recordings.
autoSCORE is intended to be used for the analysis of adult and pediatric Routine EEG’s, Long Term Monitoring EEG’s, and Ambulatory EEG recordings using a full 1020 scheme plus an EKG channel. autoSCORE is not intended to be sued for ICU and neonatal recordings.
autoSCORE is exclusively available through Natus and is currently compatible with NeuroWorks 10 GMA5.
No, the user is required to review the EEG and exercise their clinical judgement to independently make a conclusion supporting or not supporting brain disease.
autoSCORE is able to identify normal EEGs. For abnormal recordings, autoSCORE further sub-classifies abnormalities into four clinically relevant categories:
In AI algorithms, the % values indicate the confidence level of the analysis results such as abnormality markers in the study, study level normal/abnormal conclusion and study level subcategories (studies <4h). The higher the percentages, the more users can trust the autoSCORE findings. Detections with lower probability % values require more attention and possibly expertise for review.
No, autoSCORE is a locked algorithm trained by deep learning principles to indicate EEG sections including abnormalities, thereby providing controlled performance.
autoSCORE is an analyzer running on the local EEG system just like other non-AI based analyzers in NeuroWorks. autoSCORE does not meet the definition of a cyber device: It does not transmit clinical data over the local network nor to the internet. autoSCORE does not interface or collect any protected health information.
autoSCORE provides a more comprehensive classification of EEG abnormalities than current detectors in the market. Unlike traditional spike and seizure detectors, autoSCORE assesses not only Epileptiform abnormalities but also Non-Epileptiform abnormalities, guiding the diagnosis and corresponding treatment. Spike and Seizure detectors also do not provide study-level assessment informing care teams whether an EEG is normal or abnormal. With sensitivity similar to human experts, and higher than existing AI-based detectors, autoSCORE is trained to minimize false detections.
051565 RevA