MyBrain: a novel EEG embedded system for epilepsy monitoring. (3rd October 2017)

• Record Type: Journal Article
• Authors: Pinho, Francisco; Cerqueira, João; Correia, José; Sousa, Nuno; Dias, Nuno
• Is Part Of: Journal of medical engineering & technology. Volume 41:Number 7(2017)
• Page Start: 564
• Page End: 585
• Publication Date: 2017-10-03
• Subjects: EEG -- epilepsy -- wireless -- dry-electrodes -- embedded systems; Biomedical engineering -- Periodicals; Medical technology -- Periodicals; 610.28
• DOI: 10.1080/03091902.2017.1382585
• Languages: English
• ISSNs: 0309-1902

Abstract

Abstract: The World Health Organisation has pointed that a successful health care delivery, requires effective medical devices as tools for prevention, diagnosis, treatment and rehabilitation. Several studies have concluded that longer monitoring periods and outpatient settings might increase diagnosis accuracy and success rate of treatment selection. The long-term monitoring of epileptic patients through electroencephalography (EEG) has been considered a powerful tool to improve the diagnosis, disease classification, and treatment of patients with such condition. This work presents the development of a wireless and wearable EEG acquisition platform suitable for both long-term and short-term monitoring in inpatient and outpatient settings. The developed platform features 32 passive dry electrodes, analogue-to-digital signal conversion with 24-bit resolution and a variable sampling frequency from 250 Hz to 1000 Hz per channel, embedded in a stand-alone module. A computer-on-module embedded system runs a Linux ® operating system that rules the interface between two software frameworks, which interact to satisfy the real-time constraints of signal acquisition as well as parallel recording, processing and wireless data transmission. A textile structure was developed to accommodate all components. Platform performance was evaluated in terms of hardware, software and signal quality. The electrodes were characterised through electrochemical impedance spectroscopy and the operating system performance running an epileptic discrimination algorithm was evaluated. Signal quality was thoroughly assessed in two different approaches: playback of EEG reference signals and benchmarking with a clinical-grade EEG system in alpha-wave replacement and steady-state visual evoked potential paradigms. The proposed platform seems to efficiently monitor epileptic patients in both inpatient and outpatient settings and paves the way to new ambulatory clinical regimens as well as non-clinical EEG applications.