New AI speech patch could give voice to voiceless

Measuring at just one square inch, the adhesive neck patch represents a novel approach in aiding those unable to speak due to vocal cord challenges, marking a new era where technology meets healthcare to create tangible solutions for disabilities.

The device is a thin and flexible and attaches seamlessly to the neck. It works by detecting the muscle movements in the larynx and, through advanced machine learning algorithms, and then translates these movements for them to be broadcast as audible speech. The innovation is notable for its self-powered mechanism, making it a non-invasive option for individuals facing difficulties from vocal cord conditions or those recovering from related surgeries.

At the heart of this breakthrough is the dedication of Jun Chen, an Assistant Professor of bioengineering at the UCLA Samueli School of Engineering, and his team. Their device is composed of two primary components: a sensing element that captures signals from muscle movements and an actuation component that converts these signals into spoken words. These components are crafted from layers of PDMS (polydimethylsiloxane), a biocompatible silicone, and copper induction coils, creating a magnetic field that responds to the subtle movements of the larynx muscles.

Technological advances such as this are due to benefit greatly from the recent boom in AI chips, which are becoming more affordable and powerful than ever before, making sophisticated machine learning applications more accessible. The application of such AI technology in medical devices underscores a growing trend where AI and IoT converge to offer solutions that were once considered futuristic.

The device's testing phase revealed its remarkable accuracy, with nearly 95% of muscle movements correctly translated into speech. This was demonstrated through a series of experiments involving healthy adults, where the device accurately captured and vocalised sentences with nuanced precision.

Jun Chen's vision extends beyond this device, aiming to harness technology to bridge gaps in human communication and assist those with disabilities. His previous work includes a glove that translates American Sign Language into spoken English, underscoring his commitment to inclusive technology.

As the team looks to the future, they plan to expand the device's vocabulary through machine learning and begin testing on individuals with speech disorders. This ongoing development represents a hopeful horizon for many, promising a world where technological innovation can restore the fundamental human right to communicate.

This work, supported by a coalition of funding bodies, including the National Institutes of Health and the U.S. Office of Naval Research, highlight the collaborative effort behind the transformative technology and the perceived applications it could create.