An Intelligent System for the Prediction of Epileptic Seizures

Overview

More than 65 million people live with epilepsy. Epilepsy is one of the most common neurological conditions that results in unexpected seizures. Due to the unpredictable nature of epileptic seizures, it drastically increases the risk of injury, especially in daily activities such as walking or driving. Epilepsy also affects a patient's mental health, often a result of the perpetual anxiety from the seizures. The purpose of this project is to develop an accurate prediction device that utilizes raw EEG data for the prediction of epileptic seizures to alert patients of an oncoming seizure beforehand to escape dangerous situations. Using the raw EEG data, features were extracted by computing the average power spectral density of different brain waves after applying the Fast Fourier Transform. The extracted features were used as the input dataset to the various machine learning algorithms. Each model is tested with new unseen data using various metrics such as accuracy, precision, recall, and F1 score. The Random Forest (RF) model showed the highest accuracy (99%) and precision (99.3%) among the models tested and therefore selected to build the prediction device. To make the model predict seizures at a reliable rate, it is also important to reduce the dimensions of the feature space by selecting the most important features. When predicting the seizure, some channels are going to be more important than others. Channel importance is calculated for RF. This analysis helped to reduce the number of channels from 22 before channel importance to only 7 channels without significant hits to performance metrics. Using the RF algorithm, an embedded program is developed to run on a portable, low-power hardware device. The hardware system is capable of collecting multi-channel, real-time EEG signals, feature extraction, predicting and transmitting results over Bluetooth to receive in smartphone devices. The hardware includes BeagleBone Black microcontroller running open-source software and a Bluetooth transmitter-receiver with analog-to-digital converters. The hardware allows for autonomous operation and provides a warning whenever a patient is about to experience a seizure. The uncertainty in the occurrence of seizures emphasizes the need to predict seizures accurately and provide a warning for the individuals. The portable, low power hardware device is capable of giving a warning before the seizure happens. This technology can help patients perform their daily activities with more reassurance and also potentially save their lives in a particularly dangerous situation.

Competition history

• AJAS 2020 Category not listed

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• AJAS None projects
• Every AJAS 2020 project