COMPUTATIONAL PSYCHIATRY: A BRIDGE BETWEEN TRANSLATION AND PRECISION

Abstract Background Current classifications of neuropsychiatric disorders are primarily based on qualitative groupings of well-defined symptoms, whereas a change in the diagnostic framework is needed. The goal of precision psychiatry is to provide a personalized and tailored approach to prevention, diagnosis, and treatment for better individual outcomes. It is based on multiple data domains such as unique symptom expression, genetics, cognition, neuroimaging, and psychosocial factors to identify different clinical phenotypes and individual biotypes among patients. It also requires a translational approach to the underlying neurobiological mechanisms and the identification of reliable biomarkers. Computational psychiatry seems to be an essential tool to connect these two fields. Aims & Objectives The study aims to highlight the place of computational psychiatry in modern mental health care and the challenges associated with its implementation. Research objectives are as follows: 1) summarize types of computational approaches to multi-level complex data used in computational psychiatry; 2) discuss utilization areas of computational modelling in psychiatry; 3) present general limitations and challenges in implementation. Method Focusing on the basic theoretical assumptions of computational psychiatry and its applications in mental health care, a narrative review of the literature published in English in the PubMed and EMBASE databases until January 10, 2024 was conducted. Results The leading areas of medicine currently exploiting the opportunities offered by new technologies to achieve contextualized precision diagnosis and treatment are radiology, oncology, neurology, and cardiology. While computational modelling of behavior has been used in neuroscience, direct translation of the results into the context of both diagnosis and psychiatric treatment appears to be much more difficult due to the interaction of genetic, physiological, comorbidity and environmental factors on mental status. Several key methods from computational psychiatry can improve precision psychiatry. First, biophysically realistic neural network (BRNN) models allow the simulation of brain functions to understand cognitive patterns in mental disorders. Second, algorithmic reinforcement learning (ARL) models are proposed for psychiatric analysis. Finally, probabilistic approaches, such as Bayesian models (BM), can be used to predict mental states and behaviors, taking into account individual variability. These techniques facilitate a personalized approach to psychiatry, enabling tailored insights and treatments for individual patients. In addition to the methods known in other leading areas of medicine, there is an increasing interest in natural language processing (NLP) to search for the traits of the changes in mental status. In this area especially, combination of the probabilistic methods and large language models (LLM’ s) based on transformer architecture are the prospective solutions for the psychiatric treatment. Discussion & Conclusions Precision psychiatry can enhance its approach by integrating big data and machine learning techniques from computational psychiatry. However, addressing the challenges requires a multi-faceted approach, including more ecologically valid models, better integration of computational methods into clinical practice, and further research into the reliability and validity of these techniques. References [1]Fernandes, B. S., Williams, L. M., Steiner, J., Leboyer, M., Carvalho, A. F., &Berk, M. (2017). The new field of ‘precision psychiatry’. BMC medicine, 15(1), 1-7. [2]Friston, K. J. (2017). Precision psychiatry. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 2(8), 640-643. [3]Zanardi, R., Prestifilippo, D., Fabbri, C., Colombo, C., Maron, E., &Serretti, A. (2021). Precision psychiatry in clinical practice. International Journal of Psychiatry in Clinical Practice, 25(1), 19-27 [4]Anticevic, A., &Murray, J. D. (Eds.). (2017). Computational psychiatry: Mathematical modeling of mental illness. Academic Press. [5]Koutsouleris, N., Hauser, T. U., Skvortsova, V., &De Choudhury, M. (2022). From promise to practice: towards the realisation of AI-informed mental health care. The Lancet Digital Health, 4(11), e829-e840. [6]Hauser, T. U., Skvortsova, V., De Choudhury, M., &Koutsouleris, N. (2022). The promise of a model- based psychiatry: building computational models of mental ill health. The Lancet Digital Health, 4(11), e816-e828. [5]Castro Martí nez, J. C., &Santamarí a-Garcí a, H. (2023). Understanding mental health through computers: An introduction to computational psychiatry. Frontiers in Psychiatry, 14, 1092471. [6]Mujica-Parodi, L. R., &Strey, H. H. (2020). Making sense of computational psychiatry. International Journal of Neuropsychopharmacology, 23(5), 339-347. [7]Bzdok, D., &Meyer-Lindenberg, A. (2018). Machine learning for precision psychiatry: opportunities and challenges. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 3(3), 223-230. [8]Jeon, E., Yoon, N., &Sohn, S. Y. (2023). Exploring new digital therapeutics technologies for psychiatric disorders using BERTopic and PatentSBERTa. Technological Forecasting and Social Change, 186, 122130. [9]Ray, A., Bhardwaj, A., Malik, Y. K., Singh, S., &Gupta, R. (2022). Artificial intelligence and Psychiatry: An overview. Asian Journal of Psychiatry, 70, 103021. [10]Sun, J., Dong, Q. X., Wang, S. W., Zheng, Y. B., Liu, X. X., Lu, T. S.,... &Han, Y. (2023). Artificial intelligence in psychiatry research, diagnosis, and therapy. Asian Journal of Psychiatry, 103705. [11]Rumshisky, A., Ghassemi, M., Naumann, T., Szolovits, P., Castro, V. M., McCoy, T. H., &Perlis, R. H. (2016). Predicting early psychiatric readmission with natural language processing of narrative discharge summaries. Translational psychiatry, 6(10), e921-e921. [12]Ló pez-Ojeda, W., &Hurley, R. A. (2023). Medical Metaverse, Part 2: Artificial Intelligence Algorithms and Large Language Models in Psychiatry and Clinical Neurosciences. The Journal of Neuropsychiatry and Clinical Neurosciences, 35(4), 316-320.