Azernews.Az

by Alimat Aliyeva

Scientists in Australia have developed an “explainable” artificial intelligence (AI) tool that could help doctors diagnose schizophrenia by analyzing brainwave patterns, AzerNEWS reports, citing foreign media.

Researchers from James Cook University (JCU) in Australia found that machine learning models can distinguish between healthy individuals and patients with schizophrenia—even under conditions of acute stress—according to a university statement released on Tuesday.

The team tested new machine learning algorithms on electroencephalography (EEG) data collected from healthy participants, stressed individuals, and patients diagnosed with schizophrenia. The results showed that the brain responds differently in people with schizophrenia when exposed to stress compared to those without the condition.

Schizophrenia affects around 1% of the global population and is associated with increased mortality risk, making early detection and accurate diagnosis especially important for effective treatment and long-term management, according to the study published in the journal Biomedical Signal Processing and Control.

The researchers used open-access EEG datasets and developed algorithms capable of accounting for how stress influences brainwave activity, producing results that align with established neurological findings.

Importantly, the system is described as “explainable AI,” meaning it does not just provide a diagnosis but also shows the reasoning behind its conclusions—an increasingly important feature in medical applications where transparency is critical.

The researchers emphasized that AI is intended to support clinicians rather than replace them. In particular, explainable AI systems could help improve access to mental health diagnostics in remote or underserved regions, where specialist psychiatrists are often unavailable.

An interesting aspect of this development is that it highlights a broader shift in medical AI: the focus is moving away from purely high-accuracy “black box” systems toward models that can justify their decisions in human-understandable terms. If successfully integrated into clinical practice, such tools could not only speed up diagnosis but also help doctors better understand the neurological effects of stress-related mental disorders.