Session: ASME Undergraduate Student Design Expo

Paper Number: 175461

Synthetic Brain Waves for Advancing Electroencephalography (Eeg) Source Localization 

Electroencephalography (EEG) is a widely used technique for measuring the brain’s electrical activity through non-invasive electrodes placed on the scalp. EEG provides critical insights into neural function and is frequently applied in both clinical and research settings, including monitoring sleep, diagnosing neurological disorders, and studying cognitive processes. However, EEG signals are highly sensitive to interference from a variety of noise sources, which can obscure the underlying neural activity and complicate accuracy of interpretations. Sources of noise include environmental factors, physiological artifacts, and equipment-related disturbances, all of which must be carefully accounted for in both experimental and clinical analyses.

In this study, MATLAB and Octave are used as computational platforms to simulate EEG signals under controlled conditions, allowing for systematic examination of the effects of different noise types. The objective is to create simplified representations of brain waves as detected by scalp electrodes and to explore how these signals are influenced by both environmental noise and common neural disorders such as epilepsy. The base EEG signal is modeled as a 10 Hz sine wave, which corresponds to the alpha rhythm commonly observed in awake and relaxed individuals.

Three major noise types were added to the base signal to simulate realistic recording conditions. White Gaussian noise (WGN) represents random fluctuations present in virtually all electronic recordings and provides a baseline stochastic disturbance. Electromyographic (EMG) artifacts simulate high-frequency muscle activity, such as facial or scalp muscle contractions, which can contaminate EEG recordings. Powerline interference is modeled as a 60 Hz sinusoidal component, reflecting common electrical noise from mains electricity that frequently affects EEG equipment. Standard signal processing techniques are used to generate these noise components and combine them with the base signal, producing noisy EEG waveforms suitable for analysis.

These signals are first visualized in the time domain to observe overall waveform characteristics and the temporal impact of noise. Subsequently, the Fast Fourier Transform (FFT) is applied to convert the signals into the frequency domain, highlighting dominant frequencies and revealing how different noise sources alter the spectral profile of the EEG. Beyond computational simulation, these signals are implemented in a custom-built, simplified head phantom designed for source localization studies, providing a practical testbed for evaluating electrode placement, signal recovery, and neural source detection strategies.

Overall, this approach offers a controlled, replicable framework for understanding EEG signal contamination and lays groundwork for future advancements in neural signal analysis, effective noise reduction, and brain-computer interface (BCI) applications. By systematically exploring the interaction between neural signals and noise, this study contributes to more accurate EEG interpretation and the development of technologies that rely on reliable detection of brain activity.

Presenting Author: Eugenia Jin University of Texas Arlington

Presenting Author Biography: 11th grader at Grapevine High School; currently ranked 2 out of 435
2025 AAASE MIT Summer Academic Camp
2 years of research experience
2025 NASA Earth System Science 1st place Award (ISEF Fort Worth Regional Science and Engineering Fair)
2025 UIL State: Number Sense Qualifier
2025 UIL Regionals: Number Sense Team Champion
2024-2025 Recognized for Outstanding Academic Performance

Authors:

Eugenia Jin University of Texas Arlington
Ashfaq Adnan University of Texas Arlington
Nafisa Tasfee University of Texas Arlington
Vi Pham University of Texas Arlington