Researchers from Lublin University of Technology, the Medical University of Lublin, the John Paul II Catholic University of Lublin, and the University of Rochester (USA) are conducting studies on using artificial intelligence (AI) to diagnose schizophrenia based on retinal scans.
“The retina is closely linked to the central nervous system; in a way, it serves as a simplified model of what happens in the brain,” said Dr. Paweł Karczmarek, head of the Department of Computational Intelligence at Lublin University of Technology, who is leading the study. He emphasized that retinal database could serve as a valuable resource for distinguishing between healthy individuals and those with schizophrenia.
Previous research has confirmed that the retinal layers in individuals with schizophrenia tend to be thinner. However, this characteristic can also be associated with other medical conditions, necessitating a highly precise analysis.
The study employs optical coherence tomography (OCT), a device commonly used in cardiology, dermatology, and ophthalmology. This imaging technique allows for layer-by-layer examination of the retina, producing images with histological-level resolution.
“Such an examination can be completed in about a minute, significantly reducing the time and cost compared to MRI scans,” noted Dr. Karczmarek.
His team analyzed retinal scans from 59 schizophrenia patients and 61 healthy individuals. “We interpreted the OCT images using advanced AI algorithms, applying various analytical models to enhance the accuracy of distinguishing between healthy individuals and those with schizophrenia,” he explained.
The study found that combining retinal imaging with deep learning techniques based on neural networks can significantly improve schizophrenia diagnosis. The research findings were published in December 2024 in the journal Scientific Reports.
According to the authors, this discovery proves that thinner retinal layers can serve as biomarkers for schizophrenia. They believe this could lead to faster and more accurate diagnosis, enabling earlier treatment and improving patients’ quality of life.
“Modern imaging technologies supported by AI could revolutionize diagnostics not only in ophthalmology but also in neuropsychiatry, facilitating more precise monitoring of conditions such as schizophrenia,” said Dr. Karczmarek.
Schizophrenia is a mental illness classified as a psychotic disorder, characterized by distorted perception of reality, thought disturbances, and emotional dysregulation. It is most commonly diagnosed in adolescence or early adulthood, affecting approximately 400,000 people in Poland.
Currently, no objective laboratory or imaging tests exist for diagnosing schizophrenia. Diagnosis is primarily based on clinical interviews and observations. This is why researchers have high hopes for their study, as it could pave the way for a groundbreaking diagnostic tool.
(PAP)
Source:
