Development of a Miniaturized Mechanoacoustic Sensor for Cough Detection

This technical report discusses the development and evaluation of a miniaturized mechanoacoustic sensor (MAS) designed for continuous and objective cough detection in children with cystic fibrosis (CF). The study involved 36 participants and aimed to assess the usability, acceptability, and performance of the MAS in both clinical and home settings. The report outlines the challenges of existing cough assessment methods, which are primarily subjective and often unreliable in noisy environments. The MAS was trained and validated using custom deep learning algorithms, achieving high accuracy in cough detection and classification. Feedback from children and parents led to design modifications for improved usability. The report details the sensor's ability to measure cough and other physiological signals, demonstrating its potential for longitudinal evaluation of cough in pediatric patients with CF. The findings suggest that the MAS could enhance the monitoring of cough frequency and characteristics, which are critical for managing respiratory health in children with CF.