Analysis of EEG Signals Using Machine Learning Algorithms

Epilepsy is a threatening neurological disease which causes sudden bursts in the electrical activities of the brain known as epileptic seizures. Due to its less availability in terms of treatment, and the data is also very complex to be dealt with, it has become a matter of immense concern. Also, the rate of patients for a single doctor is very high in low- and middle-income countries. Epilepsy can be overcome if it is diagnosed at early stages. So, to achieve the same, an automated machine learning based model is proposed in this study which uses the EEG patterns received from the patient. The model is capable of telling whether the patient is epileptic or not by using the data from the electrical activities of the patient's brain. The model uses Supervised Machine learning and the random forest (RF) classifier to do the analysis of EEG signals for seizure detection. Python library pandas is used to deal with the data in the spreadsheet form and matplotlib is used for visualization purposes. The model has predicted epilepsy with a good accuracy of 99.96%, specificity of 99.2% and the sensitivity came out to be 100%.