精神醫療再進化：結合人工智慧與精神醫學

Abstract

精神疾病常被認為僅是心理問題，而忽略了大腦病理機轉的重要影響，造成精神醫學的診斷和評估仍停滯於主觀的症狀學診斷，而未能如內科疾病引入客觀的儀器診斷。近年來，越來越多的證據顯示精神疾病是大腦的問題。對於結構性和功能性磁振造影的解析，使得精神醫學有機會邁入生物標記診斷評估精神疾病的新階段。而機器學習方法的興起，對於大腦結構和功能數據的分析，更是帶來重要突破和深遠的影響。運用人工智慧與腦影像大數據，我們將有機會建立客觀評估精神疾病診斷與症狀嚴重程度的科學方法，除了能有效協助臨床醫師輔助評估精神疾病，提升精神醫療的照護品質，更重要的是能進一步解析大腦結構和功能與精神症狀的病理機轉，促進我們對於大腦此一人體最複雜的器官的理解，加速現代精神醫學的再進化。  Mental illness is occasionally dismissed as merely a psychological problem, with the underlying cause of brain pathology often ignored. This lack of knowledge has hindered the development of psychiatric diagnosis and related pathophysiology research. Psychiatric medicine has also failed to adopt objective diagnostic criteria such as those employed in internal medicine. Accumulating evidence reveals that mental illnesses may have underlying neurological causes. Structural and functional magnetic resonance imaging can enable the investigation of novel biomarkers for assessing psychiatric disorders. Additionally, the emergence of machine learning techniques has resulted in considerable breakthroughs and far-reaching consequences in the analysis of brain structure and functional data. Artificial intelligence and brain imaging big data may also provide an opportunity for the establishment of a scientific method for objectively assessing the diagnosis and severity of mental illness symptoms. This can effectively help clinicians in assessing mental illnesses and improve the quality of mental health care. In particular, brain-based psychiatric diagnosis may further delineate the pathological mechanisms of psychiatric symptoms, promote the understanding of the human brain (the most complex organ of the human body), and accelerate the re-evolution of modern psychiatry