The battle against COVID-19 has greatly accelerated the development of many fields in medicine. Among them is the emerging field of voice-based diagnostics. Several teams around the world are racing to develop a tool that can use voice samples to detect if someone is infected with the virus. These teams were previously developing tools to detect different types of pulmonary diseases by detecting telltale signs in the voice and breath sounds of patients.
While the use of voice biomarkers is early in its development, scientists are betting that vocal analysis can help identify a broad range of disorders given how complex is human speech. According to an article published in Nature,
Speaking requires the coordination of numerous anatomical structures and systems. The lungs send air through the vocal cords, which produce sounds that are shaped by the tongue, lips and nasal cavities, among other structures. The brain, along with other parts of the nervous system, helps to regulate all these processes and determine the words someone is saying. A disease that affects any one of these systems might leave diagnostic clues in a person’s speech.
One of the companies trying to develop voice-based COVID-19 diagnostic is Vocalis Health. They are a voice-analysis company with home bases in the US and Israel. When the COVID virus started to spread, they decided to apply their knowhow in identifying voice biomarkers, to see if they could develop a diagnostic test for the virus. They are using voice samples and COVID-19 test results submitted by volunteers in order to train their artificial intelligence models. (You can volunteer your data to the voice study here).
The Vocalis Research app is being tested around the world and will soon be able to share their findings. While this tool is not intended as a final diagnosis, it could potentially help prioritize potential cases and identify people in need of testing or care. If it proves effective, it could also be used as a screening mechanism to access indoor spaces. This could potentially be much more accurate than temperature checks.
Neurodegenerative diseases and mental health
Another big area of focus of voice biomarker research has been around age-related neurodegenerative diseases. These include both Parkinson’s and Alzheimer’s diseases for which no early diagnostic is available today. Researchers have been able to identify certain speech patterns as well as vocal changes that occur in the early stages of these types of diseases.
The potential of voice biomarkers also extends to mental health. Our voice carries a lot of information about our emotions. Features like the rate, rhythm, volume, pitch, intonation and stress in our speech can signal if we are feeling depressed, anxious or manic. Other teams are working on detecting symptoms of PTSD and psychological distress
Amazon takes the lead on the consumer side
Amazon might be the first company to launch a consumer product that uses voice analysis in a mental-health context with the release of their new wearable, the Amazon Halo. This health-focused wearable not only uses sensors to measure your overall activity level, sleep quality, and heart rate, but also has the option to analyze your voice and speech pattern to interpret your mood.
The voice analysis tool embedded in Halo is called "Tone." Users can set up a quick voice profile by reading some short snippets of text. The AI then interprets qualities of the user’s voice—such as pitch, intensity, tempo, and rhythm—to predict how others would perceive and describe the tone of their voice.
Tone analyzes the positivity and energy of your voice, for instance. Positivity is measured by how happy or sad you sound, and energy is how excited or tired you sound. The Amazon Halo app then turns that analysis into simple, daily summaries—for example you might see that in the morning you sounded calm, delighted, and warm.
Using Tone could potentially help you identify patterns in your mood and behavior so you can be more self-aware and consciously make improvements. For example, Tone results may reveal that a difficult work call led to less positivity in family discussions, an indication of the impact of stress on social well-being. Insights like this will in turn allow you to implement changes that benefit long-term mental health.
It is too early to know how effective Tone will be. It is important to remember that the use of voice biomarkers is still at an early stage and the platforms have some way to go to get more accurate. However, as the systems process more and more information, it can be expected that their predictive ability will continuously improve.
What you can do
If you are thinking about getting a health tracker, in addition to the Oura Ringand the Whoop, you should consider the Halo. It is the only one with the capability of analyzing yout voice. It could potentially be a very powerful tool to help you monitor your emotional well-being. If you do try it, let us know what you think!