Development and applications of a smartphone-based mobile electroencephalography (EEG) system

Bateson, Anthony David

Abstract

Electroencephalography (EEG) is a clinical and research technique used to non-invasively acquire brain activity. EEG is performed using static systems in specialist laboratories where participant mobility is constrained. It is desirable to have EEG systems which enable acquisition of brain activity outside such settings. Mobile systems seek to reduce the constraining factors of EEG device and participant mobility to enable recordings in various environments but have had limited success due to various factors including low system specification. The main aim of this thesis was to design, build, test and validate a novel smartphone-based mobile EEG system.

A literature review found that the term ‘mobile EEG’ has an ambiguous meaning as researchers have used it to describe many differing degrees of participant and device mobility. A novel categorisation of mobile EEG (CoME) scheme was derived from thirty published EEG studies which defined scores for participant and device mobilities, and system specifications. The CoME scheme was subsequently applied to generate a specification for the proposed mobile EEG system which had 24 channels, sampled at 24 bit at a rate of 250 Hz. Unique aspects of the EEG system were the introduction of a smartphone into the specification, along with the use of Wi-Fi for communications. The smartphone’s processing power was used to remotely control the EEG device so as to enable EEG data capture and storage as well as electrode impedance checking via the app. This was achieved by using the Unity game engine to code an app which provided the flexibility for future development possibilities with its multi-platform support.

The prototype smartphone-based waist-mounted mobile EEG system (termed ‘io:bio’) was validated against a commercial FDA clinically approved mobile system (Micromed). The power spectral frequency, amplitude and area of alpha frequency waves were determined in participants with their eyes closed in various postures: lying, sitting, standing and standing with arms raised. Since a correlation analysis to compare two systems has interpretability problems, Bland and Altman plots were utilised with a priori justified limits of agreement to statistically assess the agreement between the two EEG systems. Overall, the results found similar agreements between the io:bio and Micromed systems indicating that the systems could be used interchangeably. Utilising the io:bio and Micromed systems in a walking configuration, led to contamination of EEG channels with artifacts thought to arise from movement and muscle-related sources, and electrode displacement.

To enable an event related potential (ERP) capability of the EEG system, additional coding of the smartphone app was undertaken to provide stimulus delivery and associated data marking. Using the waist-mounted io:bio system, an auditory oddball paradigm was also coded into the app, and delivery of auditory tones (standard and deviant) to the participant (sitting posture) achieved via headphones connected to the smartphone. N100, N200 and P300 ERP components were recorded in participants sitting, and larger amplitudes were found for the deviant tones compared to the standard ones. In addition, when the paradigm was tested in individual participants during walking, movement-related artifacts impacted negatively upon the quality of the ERP components, although components were discernible in the grand mean ERP.

The io:bio system was redesigned into a head-mounted configuration in an attempt to reduce EEG artifacts during participant walking. The initial approach taken to redesign the system involved using electronic components populated onto a flexible PCB proved to be non-robust. Instead, the rigid PCB form of the circuitry was taken from the io:bio waist-mounted system and placed onto the rear head section of the electrode cap via a bespoke cradle. Using this head-mounted system, in a preliminary auditory oddball paradigm study, ERP responses were obtained in participants whilst walking. Initial results indicate that artifacts are reduced in this head-mounted configuration, and N100, N200 and P300 components are clearly identifiable in some channels .

Citation

Bateson, A. D. (2018). Development and applications of a smartphone-based mobile electroencephalography (EEG) system. (Thesis). Hull York Medical School, the University of Hull and the University of York. Retrieved from https://hull-repository.worktribe.com/output/4222739

• Thesis Type: Thesis
• Deposit Date: Nov 11, 2020
• Publicly Available Date: Feb 23, 2023
• Keywords: Medicine
• Public URL: https://hull-repository.worktribe.com/output/4222739
• Additional Information: Hull York Medical School, The University of Hull and the University of York
• Award Date: Sep 1, 2018