INFORMATION TECHNOLOGY

Tsinghua’s new achievement allows the voiceless to “wear” the throat on the body


A graphene sheet the size of a coin can help people with speech impairments regain a new “voice”. A few days ago, the reporter learned from Tsinghua University that the team of Ren Tianling, a professor at the School of Integrated Circuits, for the first time, converted graphene into a wearable intelligent artificial laryngeal device with “integrated transmission and reception”. The results were published in Nature Machine Intelligence, an artificial intelligence sub-journal of the journal Nature.

In our country, more than 300,000 people lose their voice every year due to accidental or diseased surgery. Although there are currently traditional artificial laryngies that can be used, in addition to high cost and high consumption, there are also problems such as troublesome use, poor experience, and ambiguous pronunciation for patients. In particular, the sound of the traditional artificial larynx is the sound of an electronic machine, which inevitably appears abrupt and cold in daily communication.

Therefore, if high-precision non-invasive portable voice interaction can be realized, it will undoubtedly light up a light for many patients with anechoia.

As early as 2017, Ren Tianling’s team innovatively proposed an integrated acoustic device based on graphene-using piezoresistive effect to receive signals and emit sound based on thermoacoustic effects, so as to realize the sound transmission and reception of a single device.

“Although this is a small mechanical sensor, it can ‘work wonders’.” Team members said that on the one hand, graphene intelligent artificial throat has high thermal conductivity and low heat capacity, and can emit wide-spectrum sound from 100Hz-40kHz through thermoacoustic effects; On the other hand, it can distinguish between low groans, screams, coughing, swallowing, nodding and other actions, and convert this “meaningless sound” into a sound with controllable frequency and intensity, which undoubtedly brings a “new revolution” in acoustic devices.

If at the beginning, the “graphene artificial laryngeal” can only realize simple acquisition and sound functions, after two iterations, the graphene intelligent artificial larynx (WAGT) has made major breakthroughs in device flexible attachability, sound transceiver system integration, motion monitoring system, light wearable, etc., realizing the closed loop of sound input to output, and can observe laryngeal movement in real time through the oscilloscope.

Design and performance characterization of a new generation of graphene artificial throat devices Courtesy of Tsinghua University

In the third iteration, the team developed an integrated “artificial throat” that is smaller and more functional. The diaphragm is connected via wires to a microcontroller measuring only 2 square centimeters in size, which can be carried in a pocket and comes with a button-sized battery, which is simple and convenient.

When the wearer whispers, the graphene artificial larynx senses the vibration state of the throat, extracts key acoustic features, and performs audio signal processing through artificial intelligence models to synthesize smooth and natural speech, realizing the action sound system.

It is understood that the new generation of graphene artificial larynx can basically restore the patient’s language communication ability, and identify the vaguely spoken daily words of a laryngectomy patient, with an accuracy rate of more than 90%. Even whispers, screams and coughs of different frequencies can be recorded, encoded, and classified by the device to “decode” different kinds of deaf “language”, and after training, it is expected to realize the dream of “chanting and singing”. (Source: Chen Bin, China Science News)

Related paper information:https://doi.org/10.1038/s42256-023-00616-6



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button