An AI-powered chatbot that helps detect early stages of Alzheimer’s
ChatGPT, the AI-powered chatbot that drove everyone crazy with its ability to generate human-like responses, has just made an interesting medical breakthrough.
According to a new study from the School of Biomedical Engineering, Science, and Health Systems at Drexel University in Philadelphia, researchers have found that medical retooling of AI-enabled chatbots like ChatGPT could help doctors detect the early stages of Alzheimer’s disease.
Professor Hualuo Liang, co-author of the study at Drexel University, said: “It could be very useful for early screening and risk assessment before a clinical diagnosis is made.”
The Science Daily reported that the robot was able to detect cues from a person’s spontaneous speech, predicting the early stages of dementia with an accuracy of 80 percent.
According to the study, speech disorders, including hesitation in speaking, grammatical errors and pronunciation, along with forgetting the meaning of words, are an early signal of a neurodegenerative disease in 80% of cases.
Liang added: “We know from ongoing research that the cognitive effects of Alzheimer’s disease can be manifested in speech production. The most common tests for early detection of manifestations of Alzheimer’s disease are vocal characteristics such as pauses, articulation and voice quality, as well as cognition tests. “But we think better programs can help.” Natural language processing provides another way to support early recognition of Alzheimer’s disease.”
According to study lead author Felix Agpafor, the evolving and adaptive nature of ChatGPT could make the software a useful tool for spotting warning signs in the future.
“System analysis of language and production with GPT3 makes it a promising candidate for identifying subtle speech characteristics that can predict the onset of dementia,” Agpafor said. Training GPT-3 on a huge set of interview data, including with Alzheimer’s patients, will give it the information it needs to extract speech patterns, which can then be applied to identify markers in patients in the future.”
Working with the National Institutes of Health, the researchers trained the AI on texts from the dataset, as well as speech recordings, to test its ability to detect dementia warnings.
GPT was then retrained to become an Alzheimer’s detector that proved to be more effective than the two best speech processing programs.
According to the study authors: “Our results show that GPT-3-generated text embedding can be reliably used not only to identify people with Alzheimer’s disease among healthy individuals, but also to infer the results of a subject’s cognitive test, both of which are solely based on speech data.
They continued: “We also show that text embedding is superior to the traditional acoustic feature-based approach and even performs competitively with fine-tuned models. Collectively, these results show that GPT-3-based text embedding is a promising approach for assessing Alzheimer’s disease.” “disease. It may improve the early diagnosis of dementia.”
Source: New York Post.