Designing for reliable textile neonatal ECG monitoring using multi-sensor recordings

When designing an ECG monitoring system embedded with textile electrodes for comfort, it is challenging to ensure reliable monitoring, because textile electrodes suffer from motion artifacts and incidental poor signal quality. For the design of a comfortable monitoring system for prematurely born babies in the Neonatal Intensive Care Unit (NICU), we propose the concepts of `diversity measurement' and `context awareness' to improve reliability. Clinical multi-modal sensor data was collected in the NICU with the Smart Jacket connected to a state-of-the-art amplifier. We found that the ECG signals quality varied among sensors and varied over time, and found correlations between ECG signal, acceleration data, and context, which supports the feasibility of the concepts. Our explorative system level approach has lead to design parameters and meta-insights into the role of clinical validation in the design process.