Using Binaural Audio to Diagnose COVID-19 from Sound Analysis
What does COVID-19 Sound Like?
As COVID-19, the international pandemic threatening millions of human lives, continues climbing to an indiscernible apex of new cases, researchers are now collecting audio recordings of respiratory sounds and analyzing the data in an attempt for mass diagnoses. A University of Cambridge research team made up of computer scientists, data analyzers, mobile application engineers, biology professors and physicians may be very close to collecting enough data in order to produce a mobile app that could screen an individual’s speech, breathing and coughing sounds to tell if they have been infected with the novel coronavirus. The COVID-19 Sounds App is now available as a web app for Chrome and Firefox browsers. Versions for Android and iOS will be available soon.
Using Sound To Hear Coughing
A study was approved to collect the data needed to study respiratory sounds that may indicate and diagnose an infected COVID-19 individual from a healthy one. Thousands of audio samples are being collected and recorded from both healthy and COVID-19 infected subjects online here. Study subjects are asked to speak, breathe and cough as they are recorded from their computers – to get a “picture” of what healthy and COVID-19 infected lungs sound like. The study was approved by the Ethics Committee of the Department of Computer Science and Technology, and is partly funded by the European Research Council through Project EAR.
We’re Only Hearing Half The Picture
Unfortunately, once an official mobile app is produced and distributed to diagnose COVID-19 from recording mono audio files with a mobile device, such as an iPhone or Android phone, to BILLIONS of people around the globe, there could be a LOT of room for detection error. Researchers from the Carnegie Mellon team’s Covid Voice Detector (built from the foundations of earlier voice-profiling work done at the Pittsburgh-based university) came to a conclusion that their data collection needed “a rethink” as well. “It doesn’t matter how many disclaimers you put up there – how clearly you tell people that this has not been medically validated – some people will take the machine as the word of God,” explains Dr Rita Singh. “If a system tells a person who has contracted Covid-19 that they don’t have it, it may kill that person. And if it tells a healthy person they have it, and they go off to be tested, they may use up precious resources that are limited. So, we have very little room for error either way, and are deliberating on how to present the results so that these risks vanish.”
How could we navigate the data collection to be more accurate before these apps are distributed worldwide, preventing billions of misdiagnoses that could result in erroneous deaths? By obtaining more accurate audio samples that would indicate spatial nuances of COVID-19 sounds.
Hear What The Patient Hears, From a Safe Distance
Is a mono recording of respiratory sounds an accurate or precise representation of how a three-dimensional, healthy or infected pair of lungs’ sound from a sonic perspective? No. So, how could the collective data be more precise? The data would be more telling if the respiratory sounds were recorded binaurally, to get a 3D snapshot of the waveforms’ true, 3D spectral analysis. Nuances and indicators of COVID-19 respiratory sounds – such as specific spatial locations for sounds emanating from the lungs and throat – could be recorded binaurally – making a huge difference in obtaining a more accurate snapshot of what COVID-19 really sounds like, and could be far more accurate in detection. For example, the depth of a COVID-19 subject’s cough versus the depth of an acute COVID-19’s subject’s cough (with pneumonia) could be analyzed and detected faster, resulting in acute case diagnosis that may require hospitalization, possibly saving lives more expediently.
Anytime we build a system in this 3D world, we need to build it with the intention of actually using it in the 3D world. When you’re dealing with sound, it’s important to remember that sound travels in a 360-degree sphere, and can be easily recorded that way. We would not design and build a breathalyzer on a 2D model and expect it to collect accurate data, nor should we design a mobile detection system for one of the greatest pandemics we may ever experience in our lifetimes on a 2D model if we can avoid it! Collecting the sound data of COVID-19 in a binaural capacity could possibly eliminate the massive headroom for error before a faulty, one-dimensional detection system is distributed across our 3D planet.
Agree or disagree? Please leave your comments and suggestions for this serious issue below. Billions of lives may depend on the sound research collected, and the accuracy of the detection apps that will ultimately be distributed as a result.