Breakthrough Brain Computer Interface Decodes Self Talk

Brain Computer Interface (BCIS) is Cutting– Supporting technology that provides hope to people with disabilities who have lost the ability to talk or move due to a variety of causes, such as neurodegenerative diseases, neurological disorders, or neurological disorders. Traumatic brain injury. New Landmark BCI study A led by neuroscientists at Stanford Medicine Brain Computer Interface It can decode command-indicated internal speeches with up to 74% accuracy.

“We discovered that internal speech was robustly expressed and demonstrated a proof-of-concept real-time content BCI that allows self-paced imaginary text to be deciphered from a large vocabulary (125,000 words),” said senior research author Dr. Frank Willett, Ph.D., co-director of translation and translation workers in neural prostheses, who writes a Ph.D. Neurosurgery at Stanford University, Indiana. collaboration With a team of over 20 scientists from Stanford Medicine, Massachusetts General Hospital, Harvard Medical School, Emory University, Georgia Tech, University of California, Davis and Brown University.

In addition to Willtt, Erin Kunz, Benyamin Abramovich Krasa, Foram Kamdar, Donald Avansino, Nickhun Yon, Akansha Singh, Samuel Nason-Tomaszewski, Nicholas Card, Justin Jude, Brandon Jacques, Payton Jacques, Payton Jacques, Payton Jacques, Payton Jacques, Brandon Hochberg, Daniel Rubin, Ziv Williams, David Brandman, Sergey Stavisky, Nicholas Auyong, Chethan Pandarnath, Shaul Druckmann, Jaimie Henderson.

Researchers report that the word error rate for BCI’s ability to decode internal speech in real time using a vocabulary of 125,000 words is as low as 26%.

Brain-Computer Interfaces allow people to control external devices using their thinking to perform tasks to improve the quality of daily life, such as operating-type operations such as wheelchairs, robotic limbs, computers, smartphones and other devices.

Many existing BCI systems rely on brain records of neural activity in patients attempting to speak. Researchers in this study sought to decode internal speech, internal monologues, self-talk, silent speech, internal speech, and imaginary speech, also known as audio images. Inner voiceinternal monologues, verbal thinking, secret self-talk, and internal dialogue.

Four study participants from Tetraplegia, who were part of the feasibility study of the Braingate2 clinical trial, placed Braingate neural interface system sensors located in speech-related areas of the motor cortex to record brain activity.

Participants included two males and one woman diagnosed with amyotrophic lateral sclerosis (ALS), and female stroke survivors diagnosed with tetraplegia and enterocytosis. The analog signal was digitized via the Neuroplex E system by BlackRock Microsystems as participants were asked to try a speech or imagine an internal audio.

Researchers found that there is a high correlation between neural representations of imaginary speech and trial speech, and that they can be distinguished via neural dimensions that represent motor intentions.

“We investigated the possibility of deciphering private internal speeches and found that several aspects of free-form internal speech can be deciphered during sequence recall and counting tasks,” the researchers reported.

An interesting finding is that by requesting users to consider keywords to unlock the brain computer interface, unintentional decoding of imaginary voices can be prevented. In one participant, the team found that keyword strategies work with up to 98.75% accuracy in real-time experiments.

Using the Brain Computer Interface artificial intelligence (AI) Identify noisy, recorded patterns of brain activity to predict intended speech. The scientists in this study chose to use a five-layer recurrent neural network (RNN) architecture to convert brain activity from imaginary speech into a time series of probabilities of speech. RNN is a type of AI deep learning model that can process and output sequential predictions from sequential inputs, and is often used for speech recognition, natural language processing (NLP), image captioning, and sentiment analysis.

With this new groundbreaking discovery, scientists have demonstrated that the interface of brain computers can decode speeches directed by large imaginary vocabulary.

Copyright©2025 Cami Rosso All Rights Reserved.