Wide-bandwidth Sponge-structured Nanocomposite Smart Sensors for Human Physiological Monitoring

Abstract

Smart wearables have rapidly developed during the last decade, with representative products such as Apple Watches, Fitbit, Google Glasses, Xiaomi Band, and Huawei Watch. These successful commercial Internet of Things (IoT)-based wearable devices show great potential in healthcare, sports, education, communication, etc., by monitoring human body signals in a non-invasive and continuous way. The current generation of smart wearable devices is based on rigid sensors and electronics and is dominated by inertial sensors represented by accelerometers. The bulky and rigid sensors would cause poor sensing accuracy and uncomfortable wearing due to the mismatch between the soft human body and the rigid sensor. To overcome these challenges, the next generation of smart wearable devices should be soft and conformable to the complex human body surface. The key engineering goal is to develop skin-attachable, flexible, stretchable sensors with high-sensitivity and stable sensing performance. To achieve this goal, the development of nano nanomaterials and composite structures will help allow sensors to utilize sensing materials that are beyond conventional rigid materials and structures.

This dissertation focuses on three wearable devices based on flexible pressure sensors. The sensing layer was made from the carbon nanotube/polydimethylsiloxane nanocomposite and shaped as a porous, thin sponge structure (~400 μm) with a modified imprint technique. Owing to the excellent properties of the applied nanomaterials and designed structures, the developed piezoresistive flexible pressure sensing layer has a high sensitivity in a wide bandwidth that can be used to monitor multiple human physiological signals with various assembly methods. This work demonstrated its application in a smart mask, flexible throat microphone, and muscle activity detector.

Wearing masks has been a recommended protective measure due to the risks of COVID-19, even in its coming endemic phase. Therefore, deploying a “smart mask” to monitor human physiological signals is highly beneficial for personal and public health. A smart mask integrating the freestanding ultrathin nanocomposite sponge structured sensor was proposed, which allows high sensitivity in a wide-bandwidth dynamic pressure range, i.e., capable of detecting human breathing, speaking, and coughing. Thirty-one subjects tested the smart mask in recording their respiratory activities. Machine learning methods, i.e., support vector machine and convolutional neural networks, were used to recognize these activities, which showed average macro-recalls of ~95% in both individual and generalized models. With rich high-frequency (~4000 Hz) information recorded, the two-/three-phase coughs can be mapped while speaking words could be identified, demonstrating that the smart mask can be applied as a daily wearable IoT device for respiratory disease identification, voice interaction tool, etc., in the future.

Then, a skin-attached flexible throat microphone based on the ultrathin nanocomposite sponge structured sensor with parafilm as a substrate was developed. Voice assessment is a non-invasive method that can help voice disorder-related disease diagnosis and can be conducted remotely, where the fundamental frequency is one of the essential acoustic parameters. For voice recording, the throat microphone has the advantage of being robust to environmental noise so that it can be applied outdoors daily and obtain highly reliable sounds compared to the air microphone. After recording thirty words from the human subject using the flexible sponge-structured throat microphone, the quantitative comparison with the results from an air microphone showed that it detected the fundamental frequency in human speech correctly. The ultra-flexible throat microphone has the potential to be developed as a health-monitoring smart wearable by playing a role in voice assessment.

Finally, a flexible 16-channel sensor array was fabricated with the nanocomposite sponge structured sensing layer and polyimide substrate. Muscle activity is an effective indicator for human motion monitoring and recognition, which play essential roles in various applications. The standard muscle activity measurement technique detects the weak electrical signal, i.e., electromyography (EMG). Another method that involves detecting mechanical signals of muscle, which has significant advantages compared to EMG, is mechanomyography (i.e., the lateral oscillation of the muscle). A wearable muscle activity detector was proposed based on the flexible sensor array that can measure the dynamic muscle-shape change (i.e., “muscle deformation”) and vibrational mechanomyography signals with a frequency range of approximately 0–60 Hz. In addition, the proposed device could detect at least ten lower limb motions with ~99.4% accuracy. Overall, this work presents a wearable device for monitoring mechanical signals of muscle activities with excellent lower limb motion recognition performance and demonstrates the device’s potential for next-generation human–computer interface applications.

In summary, this dissertation shows the work that bridges the technological gap between ultra-lightweight, flexible, high-sensitivity, and wide-frequency response sensor material fabrication, signal processing, and machine learning to demonstrate various wearable devices for potential applications with continual human physiological signals monitoring in daily life. It makes contributions to the next generation of soft smart wearable devices development.

Date of Award	28 Jul 2023
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Wen Jung LI (Supervisor)