Shockingly, almost one third of people will experience a vocal disorder at some point in their life; Equally disheartening is the current landscape of treatment being uncomfortable, invasive, or requiring long recovery periods. With this in mind, researchers at UCLA set out to create a wearable, user-friendly device that would improve treatment options for vocal disorders.

Vocal disorders often stem from the improper vibration of vocal cords, but individuals can still form the muscle movements in the face and neck that would otherwise correlate with speech. Therefore, solving the problem was a matter of developing a system that could sense and translate these movements into the speech even without healthy vocal cords. The device that Che et al. produced consists of two layers of metal coils to conduct the signals to an external device and a conditionally conductive layer triggered when stretched, all sandwiched between two outer layers of PVDF, a material flexible and comfortable enough to make contact with the skin. The pattern of signals produced is transmitted to an AI model, which correctly predicts sentences about 95% of the time. 

The system has many key features that are designed for user-friendliness and technical effectiveness. For example, the AI model predicts sentences rather than individual words so that context clues can guide differentiation between similar words like “Mike” and “make”. The entire device can be worn bi-directionally and detection is self-powered so that no power source is needed. The device was also shown to be effective while patients were standing, walking, running, and jumping, accommodating for active lifestyles. 

While the authors note that there is still much optimization and manufacturing work to be done before the device can be used in the medical community, it certainly broadens the horizon of  treatment for  vocal disorders. Perhaps one day we will all be able to sing the praises of these researchers. 

This study was led by Ziyuan Che and Xiao Wan with corresponding author Jun Chen at the University of California Los Angeles.

Managing Correspondent: Olivia Lavidor

Press Article: A Simple Sticker Could Allow People With Voice Disorders to Speak Again (Science Alert)

Original Journal Article: Speaking without vocal folds using a machine-learning-assisted wearable sensing-actuation system (Nature Communications)

Image Credit: Pixabay