by Researchers at the UCL Institute for Neurology have developed innovative tools that utilize AI language models to identify subtle speech signatures in patients diagnosed with schizophrenia. Published in PNAS, the study aims to explore how automated language analysis could assist doctors and scientists in diagnosing and evaluating psychiatric conditions.
Currently, psychiatric diagnoses heavily rely on patient interviews and information from close acquaintances, with minimal contribution from medical tests such as blood tests or brain scans. However, this lack of precision impedes a comprehensive understanding of the root causes of mental illnesses as well as the monitoring of treatment effectiveness.
The study involved 26 participants diagnosed with schizophrenia and 26 control participants who completed two verbal fluency tasks. In these tasks, participants were given five minutes to list as many words as they could that either belonged to the category of animals or started with the letter “p.” The researchers then used an AI language model, which had been trained on extensive internet text, to analyze the participants’ responses and predict the words they would spontaneously recall. They also assessed whether this predictability decreased in schizophrenia patients.
The findings revealed that control participants’ answers were more predictable using the AI model compared to those of individuals with schizophrenia. The difference was most significant among patients with more severe symptoms. The researchers hypothesized that this disparity may involve the brain’s ability to form connections between memories and concepts and store this information in cognitive maps. To support this theory, the authors conducted a second part of the study using brain scans to measure activity in areas of the brain associated with cognitive mapping and memory storage.
Schizophrenia is a prevalent and debilitating psychiatric disorder that affects approximately 24 million people worldwide, with over 685,000 individuals diagnosed in the UK alone. Symptoms of the condition may include hallucinations, delusions, confused thoughts, and changes in behavior, according to the NHS.
The UCL and Oxford research team plans to expand their study by incorporating a larger sample of patients from various speech settings to determine the practicality of this technology in clinical settings.
Dr. Nour, one of the researchers, expressed excitement about the advancements in neuroscience and mental health research. He emphasized the significance of combining AI language models and brain scanning technologies in understanding how the brain constructs meaning and how this process may be disrupted in psychiatric disorders. Dr. Nour also anticipated the potential integration of AI language models in healthcare within the next decade, citing the widespread interest in their medical applications.
In conclusion, the utilization of AI language models has shown promise in identifying subtle speech signatures in patients with schizophrenia. This novel approach has the potential to revolutionize psychiatric diagnosis and treatment evaluation, offering a deeper understanding of mental illnesses and improving patient care.
*Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
