ERS: Analysis of speech spots COPD flare-ups

News
Businessmen gesticulating and discussing contents on a smartphone
Luis Villasmil

Researchers have found that changes in speech recorded using a smartphone can be used to predict when a patient with chronic obstructive pulmonary disease (COPD) is at risk of an exacerbation.

COPD is the third leading cause of death worldwide, killing more than three million people a year, according to World Health Organization (WHO) estimates, with a devastating impact on patients' lives as they battle breathlessness and terrifying exacerbations (attacks) that signal the progression of the disease.

At the European Respiratory Society (ERS) congress in Vienna, Loes van Bemmel of Maastricht University Medical Center in the Netherlands, presented the results of a pilot study (PDF) – called TACTICAS – that could form the basis of a mobile app for people living with COPD.

In the 28-subject study, patients were asked to use a smartphone app to record themselves saying "aah" for as long as they could manage with one breath and then read a short text or answer a question, every day for 12 weeks. They also completed a daily questionnaire to record their COPD symptoms.

The research team that patients' voices became higher pitched when an exacerbation was imminent, while there were signs of 'jitter' – a breathy or hoarse voice – when an exacerbation was beginning.

"After an exacerbation, patients report a lower quality of life, their lung function can decline, and patients are at an increased and prolonged risk for cardiovascular events such as heart attack and stroke," said van Bemmel.

"It's vital to detect exacerbations as early as possible so appropriate treatment can be given," she told the ERS. "Unfortunately, it has proved difficult to detect exacerbations at their onset, since symptoms usually start when patients are at home."

The results are preliminary, based on 16 flare-up incidents over the course of the study, and will need to be confirmed in a larger trial, according to van Bemmel.

Nevertheless, the team is joining forces with colleagues at Radboud University Medical Center in the Netherlands to co-create a mobile app, SPEAK, that could be used by patients at home and provide an early warning of exacerbations. At the same time, they are exploring ways to responsibly collect, store, and analyse speech data so that the privacy of the speaker is maintained.

"If it can be validated, this could lead to a quick and efficient system to alert a patient and their doctor that treatment is needed," commented Professor Frits Franssen, secretary of the ERS assembly on respiratory clinical care and physiology, and head of the department of respiratory medicine at Maastricht University Medical Centre.

"Because it works via a smartphone, voice analysis could be used by anyone, at any time and anywhere," he said. "This could ultimately save money, time and patients' lives."

The pilot ties in with the larger trend of using vocal biomarkers to detect a range of health conditions, from respiratory diseases like COPD and asthma to cancer, neurological disorders, and mental illness.

One company specialising in the field, Sonde Health, has also developed software to detect and monitor for symptoms of COVID-19, asthma, COPD, and other pulmonary conditions based on six-second recordings taken on a smartphone.

Photo by Luis Villasmil on Unsplash