PATTERNS RECOGNITION OF EEG SINGAL BASED ON IAF FEATURES AND DEEP LEARNING METHODS FOR EPILEPSY DETECTION

• Main Author: Irfan Girindra, Muhammad
• Format: Final Project
• Language: Indonesia
• Subjects: Fisika
• Online Access: https://digilib.itb.ac.id/gdl/view/46060
• Institution: Institut Teknologi Bandung

Epilepsy is a disorder of the central nervous system. The main symptom of epilepsy is the emergence of seizures. Electroencephalography (EEG) is a tool that used for recording the electrical activity of the brain using electrodes that is placed on the scalp. EEG is one of the tools commonly used in studying brain activity, especially in brain anomaly disease, such as seizure. In analyzing EEG signals, feature extraction and machine learning models that can study the signal patterns are needed. In this study, a deep learning model, multi-layer perceptron, will be used to be able identify EEG signals in epilepsy patients during seizures. The signal feature that was used in this study is individual alpha frequency, a frequency that has the highest power value within the alpha wave range. The results of this study indicate that there is a decrease in the value of IAF when epilepsy patients experience seizures compared to when patients are in a normal state, and this study successfully build a deep learning model that has an accuracy of 96.55% in classifying EEG signals in normal state and in seizures using IAF features on epilepsy patients. Thus, that model can be used to detect epilepsy.