Hey Siri, are my reports positive?

The scientists slowly started to get a clear picture of COVID-19 by March 2020, when it hit hard everywhere. Mass donations of masks, personal protective equipment kits to hospitals began. People who had recovered were donating their plasma. In this time, the defense ministry of Israel and their company, Vocalis. Health asked people to donate voices. This company made a smartphone app that detected the differences in voices for signs like shortness of breath, fatigue etc in case of chronic pulmonary disorders. Two tests, including speaking on the phone to describe an image and counting from 50–70 were done. These were recorded and compared with a database of 1500 samples. This method identifies potential cases that may arrive. They may also be used as confirmatory material. They are making an attempt to find vocal biomarkers specific to certain diseases.

This could one day allow epidemiologists to track diseases through the medicos at home service via their phones. In future, phones would easily say that you have this disease and help you out with a possible cure for it. Machine learning has helped here by detecting thousands of aberrations at a time and identify the condition.

Now, this method has been applied to study neurological health, which impacts the speech the most. Again by collecting and analysing voice samples of several individuals, researchers understood that neurodegenerative diseases whose detection through pathological tests is not possible or enough, they can be done through speech. Most of these diseases carry symptoms like tremors, changes in speech, disturbance in coordination and balance. This has shown an accuracy of 99% in cases of Parkinson’s disease. In Alzheimer’s, it has shown an accuracy of approximately 93%. These positive results indicate that this method is here to stay and is the best detected method for understanding neurological health of a person.

However, it comes with its own pros and cons. This method is still not a cent percent reliable source of information as speech cannot be the only method to confirm any disorders. This also leads us to another fact that proper privacy and maintenance of information is necessary for the safety of the vocalists and the diagnosed equally. They may also be markers of characters such as gender, age, body size etc. Our voice is something that dynamically changes with time. Hence we must learn to keep up with it at all times and so it is imperative that this method is not a confirmatory in all cases. However it can take diagnosis to the next level with various improvisations over time.