90% accurate diagnosis
Cyprus International University (CIU) Faculty of Engineering Vice Dean Assoc. Prof. Dr. Hüseyin Öztoprak and Biotechnology Research Center Coordinator Assoc. Prof. Dr. Nahit Rızaner, in their study on bipolar disorders, determined based on the region of the brain that was measured which frequencies were effective in diagnosing the disease.
Assoc. Prof. Dr. Öztoprak stated that Bipolar disorders are one of the most important diseases of our day and age that negatively affect the quality of life of individuals, and are also known as a bipolar mood disorder or manic depression; and said, "these are characterized by depressive and manic periods lasting from days to years, and the individual may also end up being in a completely healthy mood. It is a mood disorder an individual may suffer from.”
Öztoprak pointed out that bipolar patients need treatment and regular control both in the depressive period and in the manic period when they can engage in risky behaviors, and also noted that this disorder, which is often diagnosed as major depression, negatively affects the lives of the patients and their families.
Oztoprak noted that in the study he carried out with Assoc. Prof. Dr. Rızaner, aimed to differentiate between bipolar and healthy people using the deep learning method; he underlined that using the artificial intelligence method, they reached a 90% accurate diagnosis among 58 participants.
Underlining that disease detection guided by artificial intelligence is an effective tool to identify undiagnosed patients in many ailments, Öztoprak said, “These algorithms were developed with the awareness that inadequate diagnosis causes a heavy burden for patients and healthcare professionals. These algorithms also help identify the biophysical causes of diseases.”
Öztoprak stated that the study used functional near-infrared spectroscopy (fNIRS), also used near-infrared spectroscopy for functional neuroimaging which is an optical brain monitoring technique, and that fNIRS data obtained from participants was used in the study adding the scientists at Başkent University and Ankara University in Turkey also contributed to this study.
Assoc. Prof. Dr. Öztoprak, in his statement, underlined that the final decision on the diagnosis of the disease should be made by experts in the field of Medicine, and emphasized that the developed method could be a useful instrument that Medical Doctors can use if it is supported by other studies.
Öztoprak and Rızaner's study titled "Deep neural network: Multi-channel near infrared spectroscopy to differentiate brain activity between patients with euthymic bipolar disorder and healthy controls during verbal fluency performance" was published in the journal Psychiatry Research: Neuroimaging.