Key result
A hybrid architecture integrating hidden Markov models and deep learning automatically classified clinical EEG patterns with a sensitivity above 90% and a specificity below 5%.
Why the study?
Automatic EEG analysis can aid real-time brain monitoring and diagnosis, but clinical acceptance requires high sensitivity with low specificity.
Does a hybrid deep learning architecture accurately classify clinical EEG patterns?
Observational (n=518)
No
Does a hybrid deep learning architecture accurately classify clinical EEG patterns?
A hybrid deep learning architecture for automated EEG analysis achieves >90% sensitivity and <5% specificity, approaching clinical requirements for real-time brain monitoring.
Should not yet alter clinical EEG monitoring; leaves open prospective validation of hybrid models.
Brain monitoring combined with automatic analysis of EEGs provides a clinical decision support tool that can reduce time to diagnosis and assist clinicians in real-time monitoring applications (e.g., neurological intensive care units). Clinicians have indicated that a sensitivity of 95% with specificity below 5% was the minimum requirement for clinical acceptance. In this study, a high-performance automated EEG analysis system based on principles of machine learning and big data is proposed. This hybrid architecture integrates hidden Markov models (HMMs) for sequential decoding of EEG events with deep learning-based post-processing that incorporates temporal and spatial context. These algorithms are trained and evaluated using the Temple University Hospital EEG, which is the largest publicly available corpus of clinical EEG recordings in the world. This system automatically processes EEG records and classifies three patterns of clinical interest in brain activity that might be useful in diagnosing brain disorders: (1) spike and/or sharp waves, (2) generalized periodic epileptiform discharges, (3) periodic lateralized epileptiform discharges. It also classifies three patterns used to model the background EEG activity: (1) eye movement, (2) artifacts, and (3) background. Our approach delivers a sensitivity above 90% while maintaining a specificity below 5%. We also demonstrate that this system delivers a low false alarm rate, which is critical for any spike detection application.
No takes yet. Share an insight, caveat, or question.
Golmohammadi et al. (2019) conducted an observational in Brain disorders (EEG monitoring) (n=518). Hybrid deep learning architecture (HMMs and Stacked denoising Autoencoders) vs. Manual scanning and interpretation (reference standard) was evaluated on Sensitivity and specificity for automatic classification of EEG patterns. A hybrid architecture integrating hidden Markov models and deep learning automatically classified clinical EEG patterns with a sensitivity above 90% and a specificity below 5%.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: