Àá½Ã¸¸ ±â´Ù·Á ÁÖ¼¼¿ä. ·ÎµùÁßÀÔ´Ï´Ù.
KMID : 0605720240300010001
Journal of the Korean Society of Biological Therapies in Psychiatry
2024 Volume.30 No. 1 p.1 ~ p.8
Use of Artificial Intelligence in Diagnosis and Treatment of Psychiatric Disorders
Baek Min-Ji

Roh Kwon-Chan
Bai Dai-Seg
Kim Hee-Jin
Seo Wan-Seok
Abstract
Artificial intelligence (AI) is widely used as an auxiliary device to diagnosis and treat mental illness, and the scope of its use is gradually expanding. It is widely used as a tool to assist diagnosis in various fields such as radiology and ophthalmology. Meanwhile, in psychiatry, the use of AI has been limited so far due to the lack of a specific objective test for diagnosis and the reliability of medical records for sensitive records such as suicide attempts. However, AI can detect people¡¯s behavior and facial expressions changes more sensitively and more accurately find the correlation between quantitative neurophysiological test results and behavior and emotions. Nowadays, AI is useful as an auxiliary tool in diagnosing not only pediatric and adolescent mental disorders such as attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD), but also affective disorders and psychotic disorders. AI is also actively used to treat insomnia, ASD, ADHD, cognitive dysfunction, depression, anxiety disorders, substance use disorders, and schizophrenia. Many psychiatrists are still cautious about the use of AI. There are concerns that AI cannot replace the doctor-patient relationship, which is the most important element of traditional psychotherapy. Nevertheless, AI can diagnose psychiatric disorders more accurately and can also be useful in improving patients¡¯ symptoms and quality of life through AI or digital treatments. I hope that psychiatrists to contribute to the treatment of human mental illness by becoming users, supervisors, and monitors of AI.
KEYWORD
Artificial intelligence, Digital therapeutics, Psychiatric disorders, Neural networks, Accuracy
FullTexts / Linksout information
Listed journal information