An automated LSTM-CNN model for EEG signal analysis achieved 100% accuracy, 100% sensitivity, and 100% specificity in distinguishing between healthy individuals and patients with seizures.
Does a combined LSTM-CNN model accurately detect epileptic seizures from EEG signals?
A proposed LSTM-CNN model demonstrates perfect accuracy, sensitivity, and specificity in detecting epileptic seizures from EEG signals, offering a potential automated diagnostic tool.
Neurologists visually inspect electroencephalogram (EEG) reports to get the epilepsy diagnosis. Scholars have suggested automated techniques to detect the ailment due to the lengthy process and global shortage of specialists. Most research in the past years has been conducted utilizing machine learning methods. But following the development of deep learning methods, many groups are employing it to make computer-aided diagnostic (CAD) systems. In this work, the authors have proposed a model comprising of Convolutional Neural Network (CNN) and long short-term memory (LSTM) to detect seizures. It focuses on extracting temporal and spatial features by integrating CNN and LSTM models. The advantage of this automated system is it extracts spatial as well as temporal features from EEG signals with less trainable parameters and gives good accuracy. This makes the system suile for real-time processing applications. It has achieved a maximum of 100% accuracy, 100% sensitivity, and 100% specificity to distinguish between healthy and seizure patients.
Jiwani et al. (Tue,) conducted a other in Epileptic seizures. LSTM-CNN model was evaluated on Distinguishing between healthy and seizure patients. An automated LSTM-CNN model for EEG signal analysis achieved 100% accuracy, 100% sensitivity, and 100% specificity in distinguishing between healthy individuals and patients with seizures.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: