Filtering Electroencephalogram Data Using R: A Study on Patients with Seizures

Overview

Seizures pose an imperative risk due to their frequency rising after the COVID-19 pandemic, with between 25-30% of patients going misdiagnosed or undiagnosed for seizure related conditions such as epilepsy. Electroencephalogram data can determine the location and timing of seizures to prevent future ones from occurring, but needs to be filtered to reduce interference and noise from biosignals. Current research focuses on filters for specific frequency isolation but doesn’t evaluate resulting oscillations that distort results or smear signals of existing issues. Three filters were evaluated in R to determine impacts of filtering on ringing and distortion: a Butterworth bandpass filter tested in three different orders which emphasized signals in certain frequencies while diminishing signals in others; a moving average filter that smooths the data without specific cutoffs; and a zero-phase filter that smooths the noise while preserving original timing. These three functions were tested on pediatric EEG data from Children’s Hospital Boston. Based on the results, the Moving Average removes more noise and isolates the pre-existing signal with a simpler algorithm, making it more useful for real-world applications such as wearable IoT bracelets used to detect seizures. Likewise, the zero-phase filter is preferred when analyzing entire signals and massive datasets, since it filters through data efficiently and preserves the features and timing of each graph. Understanding and testing these different filters provides knowledge about the applications, benefits, and consequences of the various types of filters and reveals new types of filters that can be tested on data for effectiveness.

Competition history

• CSEF 2026 Medicine & Physiology (Track 2) (Senior Division) · Entry S-21-23

Related projects

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

Browse more like this

• All Biomedical and Health Sciences projects
• CSEF Biomedical and Health Sciences projects
• Every CSEF 2026 project