Unlocking Speech Through Brain Implant & Machine Learning

Click For Summary
SUMMARY

Researchers have successfully developed a brain implant that decodes thoughts into text for individuals with anarthria, utilizing a network of electrodes and machine learning algorithms. The study achieved a median decoding rate of 15.2 words per minute and a word error rate of 25.6%. Over an 81-week period, the system demonstrated a 98% detection rate for attempts to produce individual words, with a classification accuracy of 47.1%. This groundbreaking work highlights the potential of combining neuroscience and artificial intelligence to restore communication abilities.

PREREQUISITES
  • Understanding of brain-computer interface (BCI) technology
  • Familiarity with machine learning algorithms for signal processing
  • Knowledge of neural signal recording techniques
  • Basic principles of anarthria and its impact on communication
NEXT STEPS
  • Research advanced machine learning techniques for real-time signal decoding
  • Explore the latest developments in brain-computer interface technology
  • Investigate neural signal recording methods used in clinical settings
  • Study the implications of BCI technology on speech rehabilitation
USEFUL FOR

Neuroscientists, machine learning engineers, speech-language pathologists, and researchers interested in the intersection of neuroscience and artificial intelligence.

gleem
Science Advisor
Education Advisor
Messages
2,722
Reaction score
2,209
TL;DR
An experimental system connecting electrodes implanted in a brain can display words that a person thinks of saying.
In a collaboration of several institutions researchers have been able to record the thoughts of a person with a brain stem stroke resulting in anarthria (an inability to articulate speech) by mplanting network of electrodes in the brain and using a machine learning algorithm to decode the thoughts of words and display them on a computer screen.
https://www.nytimes.com/2021/07/14/...3473&user_id=799d953295eabb464706b7f2130bc45f

 
  • Like
Likes   Reactions: 256bits, berkeman, pbuk and 1 other person
Computer science news on Phys.org
I see the article as being behind a paywall
 
phinds said:
I see the article as being behind a paywall
I'm able to access the NYT article in the OP link.
 

RESULTS​

We decoded sentences from the participant’s cortical activity in real time at a median rate of 15.2 words per minute, with a median word error rate of 25.6%. In post hoc analyses, we detected 98% of the attempts by the participant to produce individual words, and we classified words with 47.1% accuracy using cortical signals that were stable throughout the 81-week study period.That's about the state of my typing skills.
 
  • Like
Likes   Reactions: Keith_McClary

Similar threads

  • · Replies 2 ·
Replies
2
Views
1K
Replies
3
Views
1K
Replies
2
Views
3K
  • · Replies 2 ·
Replies
2
Views
1K
Replies
4
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
Replies
4
Views
1K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K