Related Experiment Video
Updated: Mar 26, 2026
Asthma Detection Research Based on Voice Signal Processing and Machine Learning
Published on: July 22, 2025
Monitoring Inhaler Inhalations Using an Acoustic Sensor Proximal to Inhaler Devices
Terence E Taylor1,2, Martin S Holmes1,2, Imran Sulaiman3
11 Trinity Centre for Bioengineering, Trinity College Dublin , Dublin, Ireland .
Background:
The efficacy of drug delivery from inhalers is very much dependent on the user's peak inspiratory flow rate (PIFR). Current methods to measure PIFR in inhalers are based on subjective checklists. There is a lack of methods currently available to objectively remotely monitor PIFR in pressurized metered dose inhalers (pMDIs) and dry powder inhalers (DPIs). In this study, for the first time, non-contact acoustic methods were employed to estimate PIFR through three commonly used inhalers (Diskus™ DPI, Turbuhaler™ DPI, and Evohaler™ pMDI) with the aim of applying these methods to remotely monitor inhaler inhalation technique in future clinical applications.
Methods:
Each inhaler was placed inside an airtight container connected to a spirometer to measure PIFR. A high quality microphone was placed 5 cm from the mouthpiece of the inhalers to record inhalation sounds. Over 2000 inhaler inhalation sounds were recorded from 11 healthy participants. A range of temporal and spectral acoustic features from the inhalation sounds were correlated with PIFR. The variation of acoustic features and the repeatability of the inhalation acoustic spectral profile were investigated to further characterize inhaler inhalation sounds and to determine the reliability of acoustics to estimate PIFR.
Results:
All acoustic features were significantly correlated with PIFR (p < 0.001). The mean power of the inhalation sound generated the most consistent correlation across all inhalers [R2 = 0.77 (Diskus™), R2 = 0.7 (Turbuhaler™), R2 = 0.75 (Evohaler™)]. Acoustic features generated low variation and the spectral profile of inhalation sounds was repeatable regardless of flow rate, suggesting that acoustic methods are a reliable method of estimating PIFR.
Conclusions:
The methods presented in this study may be employed in a wearable monitoring device in future applications to measure inhaler PIFR. Objective monitoring of PIFR in inhalers may help patients improve their inhaler inhalation technique and therefore may be of significant clinical benefit to both patients and clinicians.
