UC Irvine

The nature of bioelectrical signals is mixed with numerous sources of electric noise and interference. With the advancement in technologies, electronic data acquisition systems and sensors have been developed with high signal-to-noise ratio, becoming indispensable tools for different research studies with biological signals, including electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG), electrooculography (EOG), and electrodermal activity (EDA) on both human and animal models. However, the majority of current systems are typically cumbersome and involve wiring systems to read out data from subjects. This becomes more difficult to apply to special subjects (e.g., aquatic animals or low amplitude) or situations such as subjects who are on heavy duty or require long-term monitoring applications with ease of operation. In the context of this research work, I wanted to explore new approaches for data acquisition design and sensors to address such problems. First, a prolonged sedation system for continuous ECG monitoring of multiple zebrafish was proposed. The prolonged continuous ECG performance of our system increased diagnostic and monitoring yield in the detection of asymptomatic cardiac events and reduced ECG artifact to improve arrhythmia detection. The major novelties of the novel Zebra II ECG system lie in the prolonged measurement for multi-step experiments (up to 1 hour), high throughput screening with multiple fish, controlled environment with minimal side effects, and automated cloud-based analytics, among other controlled features. Second, non-contact biopotential sensing provides an enhanced user experience by eliminating the conventional requirement for adhesives, gels, and direct skin contact, making it well-suited for prolonged, mobile use in home settings. While previous initiatives have shown the feasibility of non-contact sensors using readily available components for ECG and other biopotential signals, main drawbacks from this type of sensors are not fully solvable, such as signal saturation and motion artifacts. As part of this work, multiple approaches were investigated to overcome such drawbacks from non-contact sensors.