Acoustic Analysis in Respiratory Medicine

Terence Taylor, Martin Holmes, Tom McCartan, Dr Imran Sulaiman, Dr Isabelle Killane, Prof. Richard Costello, Prof. Richard Reilly



Asthma and chronic obstructive pulmonary disease are two common types of chronic respiratory disease, characterized by airways obstruction. Prevalence of these two respiratory diseases has risen sharply over recent decades, creating a large economic burden due to treatment costs (Mannino 2002, Braman 2006). Obstructive airway diseases are primarily treated by delivering medication to the airways using inhalational devices.Medication efficacy is significantly dependent on the amount of drug reaching the small airways of the lungs. The medication in the inhalers is delivered to the respiratory tract via an inhalational maneuver. The clinical effect of inhaled therapy depends on the lung dose, its particle size distribution, the patient’s inspiratory flow rate, inhaled volume and degree of airways obstruction (Sumby et al 1992). Several studies have highlighted that errors in inhaler technique may be as detrimental as the lack of temporal adherence (Crompton et al 2006, Nikander et al 2011).

The two main types of inhalers used are metered dose inhalers (MDIs) and dry powder inhalers (DPIs). DPIs are considered advantageous over MDIs since they avoid the use of propellants, and are instead actuated during the inhalation (Malmberg et al 2010). The elimination of propellants allows patient coordination issues to be overcome. However, a disadvantage of DPIs is that the particle size distribution of the aerosol generated depends on the inspiratory flow rate through the device (Nielsen et al 1997). This flow rate can vary depending on the resistance of the DPI being used and on a patient’s inspiratory effort and inspiratory capacity (IC). The inspiratory flow rate influences the efficiency of removing particles of the drug formulation from the inhaler and it also affects the efficiency of deaggregation of the particles and the fine particle dose that gets into the respiratory tract (Vidgren et al 1988). This correlates to an absolute peak inspiratory flow rate (PIFR) above 30 L min−1 and ideally above 60 L min−1 (Pauwels et al 1997).

Previous studies have demonstrated that the primary cause of poor clinical outcomes in inhaler therapy is due to the inability of patients to generate sufficiently high PIFRs through their DPI (Jarvis et al 2007, Wieshammer and Dreyhaupt 2008). Currently, the only methods available to clinicians for assessing inhaler technique are subjective checklist methods and the Clement Clarke In-Check DialTM. Subjective checklist methods give no indication of whether the patient had sufficient inhalational flow effort or IC to achieve adequate drug delivery. The In-Check Dial simulates the resistance of the main types of inhalers to give an estimation of the patient’s PIFR. This method is quite effort dependent and may not correlate well with the in vivo PIFRs developed by the patient while using the actual inhaler device. There is a definite need for a device, which gives real-time feedback of inhalational technique and effort over a prolonged period of time.

This study aims to investigate if a relationship exists between inhaler inhalation sounds and airflow. Inhaler inhalation sounds are a mixture of both respiratory sounds and sounds created from turbulence in the inhaler. It is well known that a relationship exists between airflow and respiratory sounds. Most of the previous research in this area has examined the relationship between respiratory sounds generated at the trachea and on the chest wall to airflow. A relationship between airflow and sound amplitude was established by Shykoff et al (1988). 

Charbonneau et al (1987) also derived a formula for this relationship using mean amplitude and frequency to predict flow rates from respiratory sounds. Hossain and Moussavi (2004) again established a relationship between amplitude and airflow, but found that the average power in the frequency band 150–450 Hz provided a slightly stronger relationship with flow rate compared to mean amplitude in healthy subjects. Together these studies indicate that a definite relationship exists between airflow and respiratory sounds.

The main objectives of this research is to investigate the relationship between temporal and spectral features of the inhalation signal and inspiratory flow rate and volume measurements in healthy subjects. It is hypothesized that features obtained from the inhalation signal estimate PIFR and IC and that by using the inhalation signal one can predict such inspiratory values , and thusprovide clinicians with objective measurements on patient inhaler use.



[1] Taylor T E, Holmes M S, Sulaiman I, Costello R W and Reilly, R B 2016 Monitoring inhaler inhalations using an acoustic sensor proximal to inhaler devices Journal of aerosol medicine and pulmonary drug delivery. Paper is available here

[2] Killane I, Sulaiman I, Mac Hale E, Breathnach A, Taylor T, Holmes M, Reilly RB, Costello RW. Predicting asthma exacerbations employing remotely monitored adherence. Healthc Technol Lett. 2016 Mar 23;3(1):51-5. doi: 10.1049/htl.2015.0058. Paper is available here

[3] Sulaiman I, Mac Hale E, Holmes M, Hughes C, D’Arcy S, Taylor T, Rapcan V, Doyle F, Breathnach A and Seheult J 2016 A protocol for a randomised clinical trial of the effect of providing feedback on inhaler technique and adherence from an electronic device in patients with poorly controlled severe asthma BMJ open 6 e009350
Paper is available here

[4] Sulaiman I, Seheult J, Killane I , MacHale E, Reilly RB, Costello RW. A new clinically relevant method of calculating adherence. Eur Respir J. 2015;46. Paper is available here

[5] Martin S. Holmes, Jansen Seheult, Peter O'Connell, Shona D'Arcy, Carsten Ehrhardt, Ann Marie Healy, Richard W. Costello, Richard B. Reilly. An Acoustic-Based Method to Detect and Quantify the Effect of Exhalation into a Dry Powder Inhaler. , Journal of Aerosol Medicine & Pulmonary Drug Delivery. 28 (2015) 247-253. (

[6] Martin S. Holmes, Shona D'Arcy, Richard W. Costello, Richard B. Reilly. Acoustic analysis of inhaler sounds from community dwelling asthmatic patients for automatic assessment of adherence. , IEEE Journal of Translational Engineering in Medicine and Health. 2 (2014) 1-10. (

[7] Martin S. Holmes, Jansen Seheult, Colm Geraghty, Shona D'Arcy, Ultan O'Brien, Gloria Crispino O'Connell, Richard W. Costello, Richard B. Reilly. A method of estimating inspiratory flow rate and volume from an inhaler using acoustic measurements., Physiological Measurement. 34 (2013) 903-914. (

[8] Shona D'Arcy, Elaine MacHale, Jansen Seheult, Martin S. Holmes, Cian Hughes, Imran Sulaiman, Deirdre Hyland, Conor O'Reilly, Senan Glynn, Thekra Al-Zaabi, John McCourt, Terence Taylor, Frank Keane, Isabelle Killane, Richard B. Reilly, Richard W. CostelloA Method to Assess Adherence in Inhaler Use through Analysis of Acoustic Recordings of Inhaler Events., PLOS ONE (2014).(

[9] Jansen N. Seheult, Peter O’Connell, Kee Chun Tee, Tariq Bholah, Hasan Al Bannai, Imran Sulaiman, Elaine MacHale, Shona D’Arcy, Martin S. Holmes, David Bergin, Emer Reeves, Richard B. Reilly, Gloria Crispino-O’Connell, Carsten Ehrhardt, Anne Marie Healy, Richard W. Costello, The Acoustic Features of Inhalation can be Used to Quantify Aerosol Delivery from a Diskus™ Dry Powder Inhaler. , Pharmaceutical Research (2014).(


Proceedings Papers for Conferences

[1] Taylor T E, Holmes M S, Sulaiman I, Costello R W and Reilly R B 2015 Influences of gender and anthropometric features on inspiratory inhaler acoustics and peak inspiratory flow rate. In: Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE, pp 2227-30
Paper is available here

[2] Taylor T E, Holmes M S, Sulaiman I, D’Arcy S, Costello R W and Reilly R B 2014 An acoustic method to automatically detect pressurized metered dose inhaler actuations. In: Engineering in Medicine and Biology Society (EMBC), 2014 36th Annual International Conference of the IEEE, pp 4611-4
Paper is available here