Deep learning based 3D video-EEG effective and feasible, finds study
In a recent study, Tamas Karacsony and peers utilized infrared (IR) and depth (3D) videos to investigate the feasibility of a deep learning (DL) system for 24/7 seizure type detection and classification in patients with frontal lobe epilepsy (FLE), temporal lobe epilepsy (TLE), and non-epileptic events. The DL system in the study outperformed all previously published methods for seizure type detection and classification, the findings were published in Scientific Reports.
Seizure semiology is a method used to classify epileptic seizure types based on their characteristics and symptoms. However, this method requires a significant amount of resources, as it involves long-term video-Electroencephalography monitoring and visual analysis by trained professionals. To address this issue, researchers are exploring the use of computer vision-based diagnosis support tools as an alternative approach.
The highlights of the study were:
The DL system was trained on the largest 3D video-EEG database containing around 115 seizures/+680,000 video-frames/427GB.
The researchers found that their DL system achieved a promising cross-subject validation f1-score of 0.833±0.061 for the 2-class (FLE vs. TLE) case and 0.763±0.083 for the 3-class (FLE vs. TLE vs. non-epileptic) case, based on 2-second samples.
The DL system was able to accurately differentiate between seizure types with the use of an automated, semi-specialized depth-based cropping pipeline (accuracy of 95.65%) and a Mask R-CNN-based cropping pipeline (accuracy of 96.52%).
The Authors found that these results demonstrate the feasibility of using DL for 24/7 epilepsy monitoring and suggest that it may be a promising alternative to traditional seizure semiology methods. Overall, this study highlights the potential of computer vision-based tools to support the diagnosis and management of epilepsy.
Source:
Karácsony, T., Loesch-Biffar, A. M., Vollmar, C., Rémi, J., Noachtar, S., & Cunha, J. P. S. (2022). Novel 3D video action recognition deep learning approach for near real time epileptic seizure classification. In Scientific Reports (Vol. 12, Issue 1). Springer Science and Business Media LLC. https://doi.org/10.1038/s41598-022-23133-9
Disclaimer: This website is primarily for healthcare professionals. The content here does not replace medical advice and should not be used as medical, diagnostic, endorsement, treatment, or prescription advice. Medical science evolves rapidly, and we strive to keep our information current. If you find any discrepancies, please contact us at corrections@medicaldialogues.in. Read our Correction Policy here. Nothing here should be used as a substitute for medical advice, diagnosis, or treatment. We do not endorse any healthcare advice that contradicts a physician's guidance. Use of this site is subject to our Terms of Use, Privacy Policy, and Advertisement Policy. For more details, read our Full Disclaimer here.
NOTE: Join us in combating medical misinformation. If you encounter a questionable health, medical, or medical education claim, email us at factcheck@medicaldialogues.in for evaluation.