Q&A: Smart mask designed to uncover respiratory diseases

ByIsabella Hornick

ByWen Jung Li, BS, MS, PhD

Fact checked byKristen Dowd

A newly created smart mask can detect and decipher the difference between breathing, coughing and speaking, which researchers say can help in identifying several respiratory diseases.

Due to the prevalence of COVID-19 infections, face masks have become a widespread tool that help prevent illnesses. This consequence from the pandemic led a team of researchers to develop the “Internet of Things Smart Mask,” which contains an extremely thin soundwave sensor that could play an important role both in the health of individuals and the general public by tracking their daily respiratory statistics and finding possible symptoms of respiratory diseases.

Infographic explaining the features of the "Internet of Things Smart Mask" — *Source: Healio Interviews*

Wen Jung Li

Healio spoke with Wen Jung Li, BS, MS, PhD, member of the research team and chair professor of biomedical engineering in the department of mechanical engineering at City University of Hong Kong, to learn more about the Internet of Things (IoT) Smart Mask, its design features and future developments.

Healio: What was the rationale behind developing/creating the IoT Smart Mask?

Li: Smart watches, such as the Apple Watch or Fitbit, can monitor physiological signals such as heart rate, ECG, blood oxygen level, etc, which are important and potential indicators that may be used to diagnose heart-related diseases. However, they cannot monitor respiratory activities, such as how many times a person has coughed throughout a day and the patterns of the coughs, which are important indicators of respiratory illnesses such as COVID, pertussis, pneumoconiosis and asthma. Hence, we wanted to create an “IoT Smart Mask” that can be worn by humans throughout their daily activities to track and identify respiratory disease symptoms and progression.

Healio: What are key design features of the sensor/mask?

Li: We used an ultrathin sponge-like flexible sensor (fabricated by our novel sacrificial-release process) to detect the dynamic pressure of the fundamental voice frequency of 100 Hz to 800 Hz; the sensor showed excellent performance in frequency-detection accuracy and could detect acoustic harmonics up to 4,000 Hz. That is, this flexible sensor is able to sense air movements, including air directional flow and air vibrations. For example, air directional flow caused relatively irregular signals, with energy mainly focused in the low frequency range, whereas air vibration caused a periodical signal, with energy focused in the fundamental voice frequency and the corresponding harmonics.

CityU’s smart mask will help improve personal and public health.

Source: City University of Hong Kong

Healio: How does the smart mask measure respiratory sounds?

Li: The smart mask uses an ultrathin (400 µm) sponge structure-based soundwave sensor made of a carbon nanotube/polydimethylsiloxane nanocomposite that enables high sensitivity in both static and dynamic pressure measurement ranges for tracking, classifying and recognizing different respiratory activities, including breathing, speaking and coughing. The smart mask can record a wide bandwidth of air vibration frequencies ranging from 1 Hz to approximately 4,000 Hz. Using machine learning algorithms, the two-/tri-phase coughs can be mapped while speaking words can be identified.

Healio: What respiratory diseases could potentially be identified through this mask?

Li: Essentially, any respiratory disease that causes changes of voice, breathing pattern or increased frequency of cough (including different coughing patterns) of a subject can be detected by this smart mask.

Healio: You conducted a study to test out the smart mask. How did the mask perform and what were the main takeaways?

Li: We recruited 31 human subjects to collect their respiratory activities data while wearing the smart mask. These data were further processed and classified by support vector machine and convolutional neural networks. All 31 individuals had macro recalls greater than 90% (with a maximum as high as 100%), and the average reached 95.23% for these three different types of respiratory sounds (ie, breathing, speaking and coughing). The macro-recall reached approximately 95.88% for the three respiratory sounds among all 31 individuals.

Healio: Who is the target population for wearing the mask in a real-world setting?

Li: As potentially a daily smart wearable device, the smart mask will help personal and public health management in respiratory disease diagnosis/screening, especially for cities with dense populations, such as Hong Kong, Beijing, Tokyo, New York, London, etc. The ultra-thin and lightweight sensor developed in this work can be integrated and work effectively with both rigid masks and deformable non-woven fabric masks. Advanced artificial intelligence technologies enabled the integrated mask to become “smart” to recognize different coughing and breathing patterns automatically and potentially be used to diagnose respiratory-related diseases in the future.

Healio: What are your plans for future studies/developments for the smart mask?

Li: Our future studies will focus on using the smart masks to directly obtain clinical parameters from patients with chronic obstructive pulmonary disease, including the diagnosis of pneumoconiosis and COVID-19-related respiratory functions. The quantitative information obtained will provide supportive evidence for justifying the severity of pneumoconiosis and other chronic respiratory diseases.