A Highly Sensitive Coaxial Nanofiber Mask for Respiratory Monitoring Assisted with Machine Learning

Abstract

Respiration is a critical physiological process of the body and plays an essential role in maintaining human health. Wearable piezoelectric nanofiber-based respiratory monitoring has attracted much attention due to its self-power, high linearity, noninvasiveness, and convenience. However, the limited sensitivity of conventional piezoelectric nanofibers makes it difficult to meet medical and daily respiratory monitoring requirements due to their low electromechanical conversion efficiency. Here, we present a universally applicable, highly sensitive piezoelectric nanofiber characterized by a coaxial composite structure of polyvinylidene fluoride (PVDF) and carbon nanotube (CNT), which is denoted as PS-CC. Based on elucidating the enhancement mechanism from the percolation effect, PS-CC exhibits excellent sensing performance with a high sensitivity of 3.7 V/N and a fast response time of 20 ms for electromechanical conversion. As a proof-of-concept, the nanofiber membrane is seamlessly integrated into a facial mask, facilitating accurate recognition of respiratory states. With the assistance of a one-dimensional convolutional neural network (CNN), a PS-CC-based smart mask can recognize respiratory tracts and multiple breathing patterns with a classification accuracy of up to 97.8%. Notably, this work provides an effective strategy for monitoring respiratory diseases and offers widespread utility for daily health monitoring and clinical applications.

Graphical abstract