A novel approach for detection of consciousness level in comatose patients from EEG signals with 1-D convolutional neural network

Coma is an unresponsive state of unconsciousness from which a person cannot be awakened. Glasgow Score (GCS) clinical scale for determining the depth and length coma. GCS plays important role in effective accurate patient evaluation critical planning right treatment modalities care because it shows outcomes measurement performed several times day. The universally accepted as gold standard validated assessing patient's level consciousness. However, scale's success has been questioned due to variations interobserver reliability performance. In this study, data set generated Electroencephalography (EEG) signals obtained 39 comatose patients was used training deep neural networks classification consciousness level. EEG were recorded during nurse family interaction with patients. classified proposed 1D-CNN model. Consequently, two classes that we label low high are 83.3% accuracy. To our best knowledge, no prior studies using EEG-based recording process. Our study unique other terms procedure methods.