Artificially intelligent chatbots like ChatGPT could be medically engineered and could prove critical within the early detection of Alzheimer’s disease, recent research from Drexel University’s School of Biomedical Engineering, Science and Health Systems suggests.
“Our proof-of-concept shows that it will possibly be an easy, accessible and sensitive tool for community-based testing,” Professor Hualou Liang, Ph.D. from a faculty in Philadelphia and co-author of the study.
“This may be very useful for early screening and risk assessment before making a clinical diagnosis.”
The weekly bot was able to detect signals from an individual’s spontaneous speech that was 80% accurate in predicting the early stages of dementia, reported Science Each day.
![Artificial intelligence could become crucial in detecting diseases such as Alzheimer's.](https://nypost.com/wp-content/uploads/sites/2/2023/01/Alzheimer-74.jpg?w=1024)
Language impairment – including hesitation in speech, grammar and pronunciation errors, and forgetting the meaning of words – is an early red flag for neurodegenerative disease in up to 80% of cases.
“From ongoing research, we all know that the cognitive effects of Alzheimer’s disease may manifest in language production,” Liang added.
“Probably the most commonly used tests for early detection of Alzheimer’s disease, as well as to cognitive tests, are for acoustic characteristics similar to pauses, articulation and voice quality. Nevertheless, we imagine that improving natural language processing programs will provide one other avenue to support the early identification of Alzheimer’s disease.”
![ChatGPT could play a key role in detecting Alzheimer's disease.](https://nypost.com/wp-content/uploads/sites/2/2023/01/Alzheimers-51.jpg?w=1024)
According to the study’s lead writer Felix Agbavor, the evolving and adaptive nature of ChatGPT, also often called GPT3, could make this system a useful gizmo in spotting red flags.
“GPT3’s systemic approach to language evaluation and creation makes it a promising candidate for identifying subtle speech features that may predict the onset of dementia,” said Agbavor. “Training the GPT-3 with an enormous set of interview data – some of that are from Alzheimer’s patients – would offer it with the data it needs to extract speech patterns, which could then be used to discover markers in future patients.”
Working with the National Institutes of Health, researchers trained the AI using dataset transcripts and speech recordings to test its ability to detect dementia alerts. The GPT was then retrained to grow to be an Alzheimer’s detection device, proving to be simpler than the 2 best language processing programs.
![The medical community may have a lot of use for ChatGPT in the future.](https://nypost.com/wp-content/uploads/sites/2/2023/01/ChatGPT-17.jpg?w=1024)
“Our results reveal that text embedding generated by GPT-3 can be reliably used not only to detect Alzheimer’s amongst healthy controls, but in addition to infer a subject’s cognitive test scores, each based solely on speech data,” the authors said. research wrote.
“As well as, we show that text embedding outperforms conventional acoustic feature-based approaches and even competes with finely tuned models. All these results suggest that GPT-3 based text embedding is a promising approach [Alzheimer’s Disease] assessment and should improve the early diagnosis of dementia.”