- Australian researchers are developing an AI system to decode brainwaves into text.
- The AI model uses deep learning to translate EEG signals into specific words and phrases.
- Currently, the system is trained on a limited set of words and sentences for recognition.
Australian scientists are developing an artificial intelligence (AI) system that decodes brainwaves into text, or simply put, thoughts from brainwaves. While doctors use electroencephalogram (EEG) to diagnose brain conditions, researchers at Sydney's University of Technology (UTS) are using it to read thoughts.
An AI Model, developed by PhD students Charles (Jinzhao) Zhou and his supervisors Chin-Teng Lin and Dr Leong, uses deep learning to translate the brain signals from EEG into specific words.
"I am jumping happily, it's just me," the AI model produced the result when Dr Leong wore the 128-electrode EEG cap and did not utter a single word.
Currently, the AI model has been trained on a small set of words and sentences to simplify the process of recognising each word, according to a report in ABC News.
The AI is used to filter out noise and clarify the brain signals because signals from different brain sources overlap on the skull's surface, he added.
While Elon Musk's Neuralink is known for producing something similar, the research by the Sydney scientists is non-invasive in nature.
"We can't get very precise because with non-invasive, you can't actually put it into that part of the brain that decodes words," said Mr Lin.
The technology has immense implications for stroke rehabilitation, speech therapy in autism, and restoring communication for paralysis patients.
Also Read | WhatsApp Is Officially Getting Ads. Here's What Changes
AI and human brain
Across the globe, scientists have been combining EEG and AI to produce impressive results. In April, researchers at Mass General Brigham came up with an AI tool capable of predicting brain decline in patients, years in advance.
The AI tool analyses subtle changes in brain activity during sleep using EEG to make the prediction. During one of the studies, it correctly flagged 85 per cent of individuals who eventually experienced cognitive decline, with an overall accuracy of 77 per cent.
