LibGuides: BIOL 373: Automated Methods in Seizure Detection: Our Research and Beyond

Where Our Work Started

Automated identification of multiple seizure-related and interictal epileptiform event types in the EEG of mice
This link takes you to Professor Rachel Bergstrom's Ph.D research project. The paper talks about how an algorithm that utilizes wavelet decomposition and line length calculations can successfully determine spike, seizure, and abnormal electrical activity in the EEGs recorded from mice. The course Biology 373 started from this paper and hopes to continue building on the foundation from the research of Professor Bergstrom.

Below you will find a screenshot of human EEG data obtained from iEEG.org. Being able to pull actual human EEG data was a huge step for this project and a great success. Details on how to pull information from iEEG can be obtained in the bottom right box of this page labeled "Useful Documents for the Continuation of This Project."

Current Technologies in the Field of Automated Seizure Detection

Automatic epileptic seizure detection in EEGs based on line length feature and artificial neural networks
Here is a an article that describes an algorithm much like the one used in this project. It utilizes line length, wavelet transform, and an artificial neural network to detect seizures from human EEG data. Full text through a PDF download available in upper right hand corner of website.
Prediction of Seizure Onset in an In-Vitro Hippocampal Slice Model of Epilepsy Using Gaussian-Based and Wavelet-Based Artificial Neural Networks
This article shows that it is possible to use a wavelet based algorithm to detect seizures in rats. It contains highly dynamic parameters and could be tested using our line length based algorithm. Full text through a PDF download is available in upper right hand corner of website.

What did Biology 373 Accomplish in Spring 2017?

Here is a PowerPoint presentation of the substance behind this entire project: the code for the algorithm. This presentation will describe what the algorithm created by Professor Bergstrom is, what it does, and how "the code" behind the program allows for the detection of spikes, seizures, and abnormal electrical activity in the EEGs of mice.

Documentation of and Resources for the Continuation of This Project

Automated Analysis of Human Intracranial EEGs for Seizure Detection in Epilepsy
This is a summary of the research done during summer 2019. The Matlab code that Dr. Bergstrom developed for mice EEG was applied to human intracranial EEGs. Data from IEEG.org was used, specifically Mayo Clinic patient 038. Four channels (AD1-4) from hour 17 to hour 23 were manually scored by three trained observers. These were compared to the algorithms output. The algorithm was not as successful on the human EEG compared to the mice EEGs. Future development is needed.
Process to Pull Information from iEEG
This file details the steps one must take in order to obtain raw data and documentation from the ieeg.org database.

MATLAB Academy
MATLAB is the primary coding program used in this course. The algorithm from Professor Bergstrom's Ph.D research is easily accessible and readily edited in MATLAB. This link will take you to the MATLAB Academy, a website that allows you to take introductory course that explain the basics of MATLAB (and allow you to learn how to code!) for free.

Click on "Launch MATLAB Onramp" then create an account using your Beloit email. It is free, simple, fun, and will give you a basic understanding of how to manipulate the code of this research project.

Review Paper Bibliography
This Word document contains 23 references used from the review article started during the spring semester of 2017. The references in it mainly pertain to automated methods in seizure detection technology. The articles cited range from 2003 to 2017.
Methods for Visually Scoring EEGs

PowerPoints From Spring Semester 2017

Here are files containing the PowerPoint presentations given throughout the semester on various topics in Biology 373. Hopefully they can provide some background information on general topics that pertain to this class. The presentation already located on this page is a PowerPoint that was created in the spring of 2017 for one of these Thursday Topic Presenations.

EEG Presentation
This document can also be found on the page about EEGs.
iWorx Presentation
This presentation gives a brief description of iWorx and how it can capture EEG data.
EEG Montages, Schema, Interpretation, & Automation
A presentation about EEG montages and how to interpret them.
Wavelet Transform Presentation
This presentation describes the wavelet transform and its use in the automated detection of seizures.
EEG Progress Presentation
This presentation takes you through the progression of using EEG and how to collect, analyze, and interpret real life data.