Sounding healthy: from treating anxiety with a music app to listening in for signs of disease

We’ve all experienced the power of music to influence our mood. It can relax us, motivate us, console us, or validate our feelings about anything from a bad breakup to a social injustice. But now, says Man Hei Law, of Hong Kong University of Science and Technology, China, there’s a new app that can monitor our emotions and feed us the right tunes to modulate our emotions. “Think of it as a Spotify therapist,” she says.

Not that it’s something you can currently buy. It’s a research tool, one of several presented by physicists at the 183^rd Meeting of the Acoustical Society of America, 5-9 December in Nashville, Tennessee, US.

The idea, Law says, came from realizing that in the wake of COVID-19, there is a growing need to find ways to treat anxiety and depression, especially among people who might otherwise resist mental health treatment.

In the wake of COVID-19, there is a growing need to find ways to treat anxiety and depression, especially among people who might otherwise resist mental health treatment.

“So, what if we put away the ordinary face-to-face treatment and deliver a music treatment?” she asks.

To find out, her group recruited 55 stressed-out undergraduate students and assessed their moods with a standard psychological test.

Each student was then given 15 minutes of music, selected from 4,000 tunes previously categorized as happy, sad, relaxing, etc. Some received a random selection. Others got relaxing songs, while a third group received consoling music, designed to match their current mood. “Psychologically,” Law says, “showing empathy to someone will make them feel better, so if they feel sad, pick [sad] songs.”

A fourth group got songs designed to reverse their present moods. “The idea is to replace the negative mood with a positive one,” she says.

Students were allowed to do the test multiple times, so long as they were at least 24 hours apart, so in total, she got 111 trials. After each test, the students’ moods were reassessed.

So much for all the crying-in-my-beer, my-baby-done-me-wrong songs in country music. They might validate your feelings, but they won’t change your mood.

The results were intriguing. Not surprisingly, the random selection of songs had little effect. But neither did the consoling songs. So much for all the crying-in-my-beer, my-baby-done-me-wrong songs in country music. They might validate your feelings, but they won’t change your mood. If you want musical therapy, Law says, go for uplifting or relaxing. “[Those] did a great job on changing emotion.”

Other acoustics engineers are looking at how to use our bodies’ own sounds to monitor our health, without need for a doctor’s visit, using sounds that range from the resonances of our coughs to the sometimes more dramatic noises we make each morning on the toilet.

The cough scientist is Jin Yong Jeon of Hanyang University, Seoul, South Korea, who is examining how to use machine learning to determine when the sound of your cough might suggest pneumonia, Covid-19, COPD (chronic obstructive pulmonary disease), asthma, or some other serious disease, rather than just allergies or a bit of otherwise normal congestion.

Acoustics engineers are looking at how to use our bodies’ own sounds to monitor our health, without need for a doctor’s visit.

To test this, Yeon compared cough sounds from 15 pneumonia patients and 15 hospital patients with other respiratory diseases, and fed them into a computer algorithm designed not only to differentiate them, but to do so in rooms with differing acoustics (a factor that can easily make them harder to distinguish). The math was complex, but the results not so much so: a pneumonia-induced cough has a distinctive signature that can be teased out of whatever room acoustics you happen to be in when you cough with an accuracy of 97.5 percent, which for a passive, remote-monitoring diagnostic is very good.  

Yeon’s test, of course, focused solely on pneumonia, and compared it only to other lung diseases severe enough to warrant hospitalization. (Not to mention that it was tested only in a hospital environment.) But that’s just the start of what might be possible.

Just to start with, it should be possible to adapt it for use in an ambulance or an emergency room to help doctors, nurses, and first responders know what type of breathing problem they are dealing with, without waiting for other test results to come back.

It might also be possible to develop a home test that could use a smart phone app to tell you if that nagging cough is something to be worried about, or if it is showing subtle changes that might raise concerns, helping you determine when to seek help, before an ambulance ride or emergency room visit become necessary.

It might also be possible to develop a home test that could use a smart phone app to tell you if that nagging cough is something to be worried about.

That, Yean says, could be combined with records of your everyday activities, “so that we can detect if there is any subtle difference.” I.e., cough each day into your phone app, and if it says you’re ok, you probably are.

Which, ultimately, leads us back to the toilet, because after all, who doesn’t like a good story about farts?

The researcher who will win kudos on this topic from poo-obsessed kids of all ages is Maia Gatlin of Georgia Institute of Technology, Atlanta. But odd as her research—which involves instrumenting toilets with microphones—sounds (forgive the pun), it has very serious intent: serious enough that it could save hundreds of thousands of lives. 

That’s because cholera and related diseases are bigtime killers, worldwide, producing about 150,000 deaths a year, according to the World Health Organization. They are also easily treatable, if public health authorities can detect outbreaks quickly enough to react.

Gatlin’s solution: equip public toilets in cholera-prone regions with microphones connected to computers trained to distinguish the sounds of potentially deadly cholera from those of ordinary flatulence.

When it comes to detecting cholera outbreaks, the goal is simply to know where they might be happening.

To do this, she says, nobody needs a personal record of their bare butt doing its thing. A mere 10 seconds of anonymous sound is enough. When it comes to detecting cholera outbreaks, the goal is simply to know where they might be happening.

At the moment, Gatlin says, the goal is simply to wire public toilets in at-risk regions for cholera with enough sensors to be able to detect outbreaks and save lives more quickly than is possible under current, more expensive, monitoring programs. “The point of this is to be able to track a community, not an individual,” Gatlin says.

But that’s “now.” What about the future? For that, she says, it might be possible to equip individual, private toilets with the same type of sensors, designed to detect changes in how you sound when, as your youngest kids will say, you poo. Did you eat a huge batch of beans last night? That’s one thing. Could colorectal cancer be changing the geometry of your rectum, and the sounds of your poo? That would be something else. “You might not understand that the sounds of your farts or poop is changing,” Gatlin says, “but if we could track that in terms of rectal cancer [it might make a difference].”