Researchers at the University College London Institute of Neurology have created novel AI language model-based tools that can detect subtle speech patterns in schizophrenic patients. They published this study in PNAS in order to better understand how automated language analysis could help doctors and scientists diagnose and evaluate psychiatric conditions.

Currently, psychiatric diagnosis is based primarily on patient interviews and input from those close to them, with little reliance on tests such as blood tests and brain scans.

Nonetheless, this ambiguity impedes a deeper understanding of the causes of mental illnesses as well as the tracking of treatment efficacy.

To address this, the researchers tasked 26 participants with schizophrenia and 26 control participants with completing two verbal fluency tasks. In these tasks, they had to name as many words as possible within five minutes, either from the category "animals" or starting with the letter "p."

AI Language Models

To analyze participants' responses, the team employed an AI language model trained on extensive internet text data, which represented word meanings similarly to humans. They assessed whether the AI model could predict the words spontaneously recalled by people and if this predictability decreased in patients with schizophrenia.

The researchers believe that this distinction may relate to how the brain forms connections between memories and concepts, storing them in what they refer to as 'cognitive maps.' They find support for this theory in a second part of the same study, where they used brain scans to measure activity in brain regions involved in creating and storing these 'cognitive maps.'

"Until very recently, the automatic analysis of language was out of reach of doctors and scientists," said lead author Dr. Matthew Nour (UCL Queen Square Institute of Neurology and University of Oxford). This situation is changing, however, with the introduction of artificial intelligence (AI) language models such as ChatGPT.

"This work shows the potential of applying AI language models to psychiatry - a medical field intimately related to language and meaning."

Latest Information About Schizophrenia

Schizophrenia, a common and incapacitating psychiatric disorder, affects approximately 24 million people worldwide, with over 685,000 affected in the United Kingdom. According to the NHS, symptoms of this condition can include hallucinations, delusions, muddled thoughts, and behavioral changes.

The UCL and Oxford research team intends to apply this technology in a broader patient sample, encompassing a wider range of speech environments, to assess its potential clinical utility.

Dr. Nour said:

"We are entering a very exciting time in neuroscience and mental health research. By combining state-of-the-art AI language models and brain scanning technology, we are beginning to uncover how meaning is constructed in the brain, and how this might go awry in psychiatric disorders. There is enormous interest in using AI language models in medicine. If these tools prove safe and robust, I expect they will begin to be deployed in the clinic within the next decade.