advertisement

Follow Mint Lounge

Latest Issue

Home > Smart Living> Innovation > New AI-assisted device can turn muscle movements into speech

New AI-assisted device can turn muscle movements into speech

The new AI-assisted wearable device adheres to the user’s neck and translates the muscle movements of the larynx into audible speech

This innovative device, which measures 1.2 square inches on each side, has shown 95% accuracy in translating muscle movements into speech
This innovative device, which measures 1.2 square inches on each side, has shown 95% accuracy in translating muscle movements into speech (Courtesy: Jun Chen lab/UCLA)

Often, people with voice disorders or who are recovering from laryngeal cancer surgeries can find it difficult or impossible to speak. Aiming to change that, researchers have invented a thin, stretchy device that can help people with dysfunctional vocal cords regain their voice function.

The device, created by bioengineers at the University of California, Los Angeles (UCLA), adheres to the user’s neck and translates the muscle movements of the larynx into audible speech. The self-powered AI device has been trained using machine learning to recognise which muscle movements correspond with which words, the university’s statement explained.

Also read: New AI model can use human perception to tune out noisy audio

This innovative device, which measures 1.2 square inches on each side, has shown 95% accuracy in translating muscle movements into speech. It has two components, each containing two layers: one of biocompatible silicone compound polydimethylsiloxane (PDMS), with elastic properties, and a magnetic induction layer made of copper induction coils. The fifth layer is between the two and consists of PDMS mixed with micromagnets, which triggers a magnetic field, the statement elaborated. 

According to the researchers, the existing solutions such as handheld electro-larynx devices and tracheoesophageal- puncture procedures can be inconvenient, and invasive. “This new device presents a wearable, non-invasive option capable of assisting patients in communicating during the period before treatment and during the post-treatment recovery period for voice disorders," lead author Jun Chen said in the statement.

The team aims to continue expanding the vocabulary of the device through machine learning and testing it in people with speech disorders. The findings were published earlier this month in Nature Communications.

Previous studies have also looked into helping people with speech disorders. In 2021, researchers from the Indian Institute of Technology (IIT), Jodhpur and the All India Institute Of Medical Science (AIIMS), Jodhpur developed low-cost “talking gloves” for people. According to Press Trust of India, the gloves were priced at less than 5,000 and used principles of artificial intelligence and machine learning to automatically generate speech. In the press statement, the team explained that the device can help individuals convert hand gestures into text or pre-recorded voices.

In another study, published in the Nature journal, researchers helped a woman, who suffered a stroke and was paralyzed, speak again. They connected sensors and wires to her head, which could read her brain waves and convert them into speech.

Also read: These talking gloves can help people with speech disability

Next Story