Искусственный интеллект в психиатрии: достижения, перспективы, проблемы

Mental disorders are a complex health problem that requires significant resources and highly qualified specialists. Artificial Intelligence (AI) offers innovative solutions to transform psychiatric care, encompassing prevention, diagnosis, therapy and research. Currently, AI algorithms demonstrate high accuracy in the diagnosis of various disorders, including schizophrenia, depression and autism, using data from electronic medical records, neuroimaging and "digital phenotypes". AI helps predict the course of diseases, the response to treatment, and risks such as suicide or aggressive behavior. Virtual assistants, chatbots and virtual reality technologies support patients by providing psychoeducation, cognitive behavioral therapy and condition monitoring. AI automates systematic literature reviews, analyzes large amounts of data, and builds clinical and psychological models, for example, for the treatment of addictive disorders. AI helps doctors in obtaining information, preparing for exams and making recommendations for patients. Promising technologies: psychovisualization (combining neuroimaging, biometric data and AI to visualize thoughts, perceptions and emotions). There are a number of problems with the introduction of AI into psychiatry: lack of high-quality data, opacity of AI models, difficulties of validation and regulation, lack of knowledge about AI among clinicians, the need to change workflows, risks of error automation, questions about data confidentiality, responsibility for decisions, algorithm bias, balance between efficiency and safety. Interdisciplinary cooperation, increasing confidence in AI systems through understanding the logic of decision-making, and training specialists to work with new technologies will become ways to overcome problems. AI has great potential to transform psychiatric care, but it requires a responsible approach and solutions to existing problems.

Keywords: "artificial intelligence in psychiatry", "neural networks in psychiatry", "computer vision in psychiatry", "psychovisualization in psychiatry using machine learning", "emotion recognition using artificial intelligence", "diagnosis of mental disorders using machine learning", "personalized psychiatry using artificial intelligence".

Information about the author:

Bogdanov Yaroslav Vyacheslavovich – e-mail: yarik@yabogdanov.ru ; https://orcid.org/0009-0002-3880-7152

To cite thise article: Bogdanov Ya.V., Artificial intelligence in psychiatry: achievements, expectations, prospects, problems

Автор, ответственный за переписку: Богданов Ярослав Вячеславович – e-mail: yarik@yabogdanov.ru

Correstonding author: Bogdanov Yaroslav Vyacheslavovich – e-mail: yarik@yabogdanov.ru

Введение

Актуальность

Наблюдается устойчивый рост научных публикаций, посвященных применению искусственного интеллекта в психиатрии. Значительное количество работ в этой области были опубликованы в 2021-2023 годах, что демонстрирует высокий интерес к данной теме. Поэтому необходим систематический анализ научных публикаций для выявления основных тенденций и перспектив развития этого направления.

Цель исследования: проанализировать текущее состояние и перспективы применения искусственного интеллекта в психиатрии, охватывая различные области: диагностику, терапию, направление исследований, выявить основные вызовы, связанные с внедрением искусственного интеллекта (ИИ) в психиатрическую практику.