The neurotechnology environment makes some major advancements. The researchers based at the University of California, Davis have come up with a device is known as a Brain-computer interface capable of translating brain neural signals into speech in a short span of time. The details of the study were published in the New England Journal of Medicine on August 14 2024, and provides further upside for people suffering from very severe speech.
The Challenge
For so many years neuroscientists have been developing BCIs that would allow aphasic patients who are bedridden because of paralysis to speak using the BCIs reading the patient’s brain and spelling the things they want to say. They have, however, faced some issues concerning previous interfaces;
They moved too 1 to two quarter of decoders, with about 75% of words moving correctly decoded which was their single point declarability
They required long periods to be configured before being used
All those challenges made it hard for both the patients as well as caregivers encouraging routine use in the home setting.
The Breakthrough
While doing so Drs. Nicholas Card, Sergey Stavisky and David Brandman of UC Davis, has pushed and smashed all those constraints with BCI system. Borisridden spoke fluently but transformed with amyotrophic lateral sclerosis and lost his ability to speak normally containing 69-year old who had cerebral stroke and progressing aphasia of bulbar
How It Works
The participant was implanted with four plates of 64 electrodes each in the brain, in particular the left precentral gyrus in the region preparing for movement related to the speech muscles.
When the participant makes an attempt to talk, neural activity is captured by electrodes.
In real time , these signals are interpreted by a deep learning method called an artificial neural network.
The subjects see printed text generated from the internalized statements and hear a voice identical to the pre-ALS voice of the participant to which the system translates the subjects’ statements using a text-to-speech program.
Amazing Results
Day 1 (25 days after implantation):
- Just 30 minutes of calibration
- Vocabulary contains 50 words
- Accuracy of 99 % of the words used
Day 2:
- 1.5 more hours of further calibration.
- Vocabulary training consisted of 125 000 practised words
- Accuracy greater than 90% of the words used
Long-term performance:
- After sixteen hours of use: about 97.5% accuracy of the words used.
- This accuracy was maintained over 8 months
- 248 hours of assisted device usage were reported during the study by the participant
- Residual speaking time rate of about 32 wpm was witnessed.
What This Means
This technology represents a new era in the potential of BCIs:
- Consistency of High Accuracy: Always above 90 %, order regardless of volume of vocabulary present.
- Super Quick Up: From few hours or few days down to just a few minutes.
- A Very Large Vocabulary: Natural communication that flows and makes sense is further made possible by the number 125,000.
- Short Terms Operability: Sustained high level operational ability for months.
One of the weMmost active members of this study, Dr. David Brandman, also highlights the human side of the issue: “This technology is radical because it gives a chance to people who want to speak but cannot do so. I wish that such devices like this speech brain-computer interface will allow patients in the future to speak to members of their family or friends.”
Preview
Despite the fact that this paper has described the treatment with one ALS patient only, an impact should be expected. This technology ought to help those who have lost their ability to talk due to a variety of illnesses, such as strokes and traumatic brain injuries.
As the work proceeds, perhaps better results would be achieved in terms of higher precision, faster, simpler, and easier operations. The dream of turning inner sense into spoken words will be closer than ever before leading to a great number of people being able to speak again and bringing them back to the normal world.
We look forward to new news in this promising direction of neurotechnology!
