AI algorithm that diagnoses depression from MRI scans gains key regulatory approval
An artificial intelligence algorithm said to be capable of identifying signs of depression on brain images has just received a key regulatory approval in Japan.
Standard diagnoses of mental health conditions like depression are subjective, typically based on interviews with providers and patients’ self-reported symptoms. This can make it difficult to find the appropriate treatment, especially since treatment effectiveness varies widely based on a number of factors. Now, experts have developed an algorithm they believe can help bring more objectivity to the diagnostic process by giving providers visual evidence of alterations in brain activity and connectivity in individuals who report experiencing depressive symptoms.
The algorithm is the result of extensive research conducted by a team of experts at the Advanced Telecommunications Research Institute International in Japan. The algorithm analyzes data from MRI scans to quantify patterns visualized during a 10-minute exam. Certain patterns of brain activity are indicative of depression and may not be obvious to human readers.
Researchers trained the AI on resting state MRI data from 700 individuals, some of whom were diagnosed with depression and others who were considered healthy controls. In clinical testing, researchers determined the algorithm accurately identifies up to 70% of individuals with depression based on their imaging alone.
The regulatory approval represents the first of a two-step process. The approval paves the way for the device to be used in controlled clinical settings while additional research on its safety and effectiveness is conducted. Researchers intend to apply for the next stage of approval in the spring of 2026, with hopes to earn insurance coverage through Japan’s public healthcare system by fiscal year 2027.
The team is hopeful the algorithm will reach widespread clinical use within the next three years. In the meantime, they are continuing to collect data, in addition to expanding their research into its utility for identifying other psychiatric conditions.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.