AI-Powered Patch Offers New Chance for Voice Restoration

A new artificial intelligence (AI)-powered adhesive patch that is designed to assist individuals with voice disorders, has been unveiled by researchers in the United States, with the AI being used to decipher throat muscle movements and convert them into audible speech.

Details of the research were published in the journal Nature Communications. The compact patch, powered only with the user’s muscle movements, is applied directly to the skin. It will then capture the motions of the larynx muscles, translating them into electrical signals.

A machine learning algorithm will then process the converted electrical signals to generate speech.

While currently limited to vocalizing pre-programmed sentences through a speaker, the researchers emphasize the patch’s potential for those with vocal cord damage.

The published study also noted that around a third of the population will have a voice disorder once in their lifetime, which the AI powered device will aim to address.

“Specifically, 29.9% of the general population had at least one voice disorder during their lifetime, 7% are currently undergoing voice problems, and 7.2% of employed participants reported missing work days due to voice disorder,” the study said, quoting another study titled “Voice disorders in the general population.”

The waterproof patch takes the size of a large coin, and is touted to be both lightweight and flexible, as the patch only weighs seven grams and has the texture of a rubber glove.

Trials involving eight participants without voice disorders demonstrated effectiveness across various activities. The patch successfully interpreted spoken and whispered sentences with an accuracy rate of around 95%.

However, the research also identified limitations, as the patch occasionally struggled with words or names that caused similar throat muscle movements. (GFB)