ChatGPT AI can help screen for Alzheimer’s disease
The artificial intelligence chatbot program ChatGPT, known for its ability to produce human-like responses when prompted, could help doctors detect the early stages of Alzheimer’s disease, researchers say.
A study from the Philadelphia-based Drexel University School of Biomedical Engineering, Science, and Health Systems found that Open AI’s ChatGPT was 80% accurate at predicting early signs of dementia. memory.
The study’s authors write in their peer-reviewed paper, published December 22 in the journal PLOS Digital Health, that the program offers a “promising method” for assessment. Alzheimer.
Hualou Liang, Drexel professor and co-author of the study, said: “Our proof of concept suggests that this could be a simple, accessible and sensitive enough tool for testing based on community”.
“This can be very useful for early screening and risk assessment prior to clinical diagnosis.”
OpenAI is a nonprofit research organization founded by Silicon Valley investor Sam Altman and billionaire Elon Musk, whose backers include venture capitalist Peter Thiel.
ChatGPT gained widespread attention after OpenAI launched a public test of its AI-powered chatbot, surprising users with its ability to provide answers to a wide range of questions.
The researchers behind the ChatGPT study say that because language impairment affects between 60 and 80 percent of dementia patients, the programs are able to detect “subtle clues,” such as grammar and Pronunciation errors, hesitations, and forgetfulness of words, can help determine whether a patient needs a thorough examination.
Although there is no cure for Alzheimer’s, researchers say treatment options exist if the disease is caught early.
“The most commonly used tests for early detection of Alzheimer’s disease look at acoustic characteristics, such as pauses, pronunciation and vocal quality, in addition to tests,” says Liang. about awareness. “But we believe that improving natural language processing programs will provide another avenue to aid in the early identification of Alzheimer’s disease.”
With the help of the US National Institutes of Health, the researchers say they trained the program using voice recordings, collected specifically to test its ability to predict dementia. memory of natural language processing programs.
They say the program can generate a profile of Alzheimer’s speech, also known as “embedding,” which the researchers then used to retrain the program to pick out transcriptions created by someone developing Alzheimer’s disease.
The program also performed better at predicting the severity of a person’s dementia than a “acoustic” analysis that looked at characteristics like pauses, voice pitch, and slurred speech.
Felix Agbavor, a doctoral researcher and lead author of the paper, says training ChatGPT with a “huge interview dataset”, including several interviews with Alzheimer’s patients, will provide information needed to detect speech patterns that can then identify patients in the future.
In the future, the researchers plan to develop a web application that can be used as a pre-screening tool for Alzheimer’s at home or in a doctor’s office.
With files from CNN and Reuters