<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE root>
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ali="http://www.niso.org/schemas/ali/1.0/" article-type="research-article" dtd-version="1.2" xml:lang="en"><front><journal-meta><journal-id journal-id-type="publisher-id">Discrete and Continuous Models and Applied Computational Science</journal-id><journal-title-group><journal-title xml:lang="en">Discrete and Continuous Models and Applied Computational Science</journal-title><trans-title-group xml:lang="ru"><trans-title>Discrete and Continuous Models and Applied Computational Science</trans-title></trans-title-group></journal-title-group><issn publication-format="print">2658-4670</issn><issn publication-format="electronic">2658-7149</issn><publisher><publisher-name xml:lang="en">Peoples' Friendship University of Russia named after Patrice Lumumba (RUDN University)</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="publisher-id">46741</article-id><article-id pub-id-type="doi">10.22363/2658-4670-2025-33-3-327-344</article-id><article-id pub-id-type="edn">HJAJCB</article-id><article-categories><subj-group subj-group-type="toc-heading" xml:lang="en"><subject>Letters to the Editor</subject></subj-group><subj-group subj-group-type="toc-heading" xml:lang="ru"><subject>Письма в редакцию</subject></subj-group><subj-group subj-group-type="article-type"><subject>Research Article</subject></subj-group></article-categories><title-group><article-title xml:lang="en">Methods for developing and implementing large language models in healthcare: challenges and prospects in Russia</article-title><trans-title-group xml:lang="ru"><trans-title>Методы разработки и внедрения больших языковых моделей в здравоохранении: проблемы и перспективы в России</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author"><contrib-id contrib-id-type="orcid">https://orcid.org/0000-0003-3651-7629</contrib-id><contrib-id contrib-id-type="scopus">16408533100</contrib-id><contrib-id contrib-id-type="researcherid">O-8287-2017</contrib-id><name-alternatives><name xml:lang="en"><surname>Shchetinin</surname><given-names>Eugeny Yu.</given-names></name><name xml:lang="ru"><surname>Щетинин</surname><given-names>Е. Ю.</given-names></name></name-alternatives><bio xml:lang="en"><p>Doctor of Physical and Mathematical Sciences, Professor at the Department of Information Technology and Systems</p></bio><email>riviera-molto@mail.ru</email><xref ref-type="aff" rid="aff1"/></contrib><contrib contrib-type="author"><contrib-id contrib-id-type="orcid">https://orcid.org/0000-0003-4466-8531</contrib-id><name-alternatives><name xml:lang="en"><surname>Velieva</surname><given-names>Tatyana R.</given-names></name><name xml:lang="ru"><surname>Велиева</surname><given-names>Т. Р.</given-names></name></name-alternatives><bio xml:lang="en"><p>Candidate of Physical and Mathematical Sciences, Assistent Professor of Department of Probability Theory and Cyber Security</p></bio><email>velieva-tr@rudn.ru</email><xref ref-type="aff" rid="aff2"/></contrib><contrib contrib-type="author"><contrib-id contrib-id-type="orcid">https://orcid.org/0009-0004-4661-5059</contrib-id><name-alternatives><name xml:lang="en"><surname>Yurgina</surname><given-names>Lyubov A.</given-names></name><name xml:lang="ru"><surname>Юргина</surname><given-names>Л. А.</given-names></name></name-alternatives><bio xml:lang="en"><p>Ph.D. of Pedagogical Sciences, Head of the Department of Mathematics and Information Technology of the Sochi branch</p></bio><email>yurgina_la@pfur.ru</email><xref ref-type="aff" rid="aff2"/></contrib><contrib contrib-type="author"><contrib-id contrib-id-type="orcid">https://orcid.org/0000-0003-1000-9650</contrib-id><name-alternatives><name xml:lang="en"><surname>Demidova</surname><given-names>Anastasia V.</given-names></name><name xml:lang="ru"><surname>Демидова</surname><given-names>А. В.</given-names></name></name-alternatives><bio xml:lang="en"><p>Candidate of Physical and Mathematical Sciences, Associate Professor of Department of Probability Theory and Cyber Security</p></bio><email>demidova-av@rudn.ru</email><xref ref-type="aff" rid="aff2"/></contrib><contrib contrib-type="author"><contrib-id contrib-id-type="orcid">https://orcid.org/0000-0002-1856-4643</contrib-id><name-alternatives><name xml:lang="en"><surname>Sevastianov</surname><given-names>Leonid A.</given-names></name><name xml:lang="ru"><surname>Севастьянов</surname><given-names>Л. А.</given-names></name></name-alternatives><bio xml:lang="en"><p>Professor, Doctor of Sciences in Physics and Mathematics, Professor at the Department of Computational Mathematics and Artificial Intelligence of RUDN University, Leading Researcher of Bogoliubov Laboratory of Theoretical Physics, Joint Institute for Nuclear Research</p></bio><email>sevastianov-la@rudn.ru</email><xref ref-type="aff" rid="aff2"/><xref ref-type="aff" rid="aff3"/></contrib></contrib-group><aff-alternatives id="aff1"><aff><institution xml:lang="en">Sevastopol State University</institution></aff><aff><institution xml:lang="ru">Севастопольский государственный университет</institution></aff></aff-alternatives><aff-alternatives id="aff2"><aff><institution xml:lang="en">RUDN University</institution></aff><aff><institution xml:lang="ru">Российский университет дружбы народов</institution></aff></aff-alternatives><aff-alternatives id="aff3"><aff><institution xml:lang="en">Joint Institute for Nuclear Research</institution></aff><aff><institution xml:lang="ru">Объединённый институт ядерных исследований</institution></aff></aff-alternatives><pub-date date-type="pub" iso-8601-date="2025-10-15" publication-format="electronic"><day>15</day><month>10</month><year>2025</year></pub-date><volume>33</volume><issue>3</issue><issue-title xml:lang="en">VOL 33, NO3 (2025)</issue-title><issue-title xml:lang="ru">ТОМ 33, №3 (2025)</issue-title><fpage>327</fpage><lpage>344</lpage><history><date date-type="received" iso-8601-date="2025-10-28"><day>28</day><month>10</month><year>2025</year></date></history><permissions><copyright-statement xml:lang="en">Copyright ©; 2025, Shchetinin E.Y., Velieva T.R., Yurgina L.A., Demidova A.V., Sevastianov L.A.</copyright-statement><copyright-statement xml:lang="ru">Copyright ©; 2025, Щетинин Е.Ю., Велиева Т.Р., Юргина Л.А., Демидова А.В., Севастьянов Л.А.</copyright-statement><copyright-year>2025</copyright-year><copyright-holder xml:lang="en">Shchetinin E.Y., Velieva T.R., Yurgina L.A., Demidova A.V., Sevastianov L.A.</copyright-holder><copyright-holder xml:lang="ru">Щетинин Е.Ю., Велиева Т.Р., Юргина Л.А., Демидова А.В., Севастьянов Л.А.</copyright-holder><ali:free_to_read xmlns:ali="http://www.niso.org/schemas/ali/1.0/"/><license><ali:license_ref xmlns:ali="http://www.niso.org/schemas/ali/1.0/">https://creativecommons.org/licenses/by-nc/4.0</ali:license_ref></license></permissions><self-uri xlink:href="https://journals.rudn.ru/miph/article/view/46741">https://journals.rudn.ru/miph/article/view/46741</self-uri><abstract xml:lang="en"><p>Large language models (LLMs) are transforming healthcare by enabling the analysis of clinical texts, supporting diagnostics, and facilitating decision-making. This systematic review examines the evolution of LLMs from recurrent neural networks (RNNs) to transformer-based and multimodal architectures (e.g., BioBERT, MedPaLM), with a focus on their application in medical practice and challenges in Russia. Based on 40 peer-reviewed articles from Scopus, PubMed, and other reliable sources (2019-2025), LLMs demonstrate high performance (e.g., Med-PaLM: F1-score 0.88 for binary pneumonia classification on MIMIC-CXR; Flamingo-CXR: 77.7% preference for in/outpatient X-ray re-ports). However, limitations include data scarcity, interpretability challenges, and privacy concerns. An adaptation of the Mixture of Experts (MoE) architecture for rare disease diagnostics and automated radiology report generation achieved promising results on synthetic datasets. Challenges in Russia include limited annotated data and compliance with Federal Law No. 152-FZ. LLMs enhance clinical workflows by automating routine tasks, such as report generation and patient triage, with advanced models like KARGEN improving radiology report quality. Russia’s focus on AI-driven healthcare aligns with global trends, yet linguistic and infrastructural barriers necessitate tailored solutions. Developing robust validation frameworks for LLMs will ensure their reliability in diverse clinical scenarios. Collaborative efforts with international AI research communities could accelerate Russia’s adoption of advanced medical AI technologies, particularly in radiology automation. Prospects involve integrating LLMs with healthcare systems and developing specialized models for Russian medical contexts. This study provides a foundation for advancing AI-driven healthcare in Russia.</p></abstract><trans-abstract xml:lang="ru"><p>Большие языковые модели (LLM) трансформируют здравоохранение, позволяя анализировать клинические тексты, поддерживать диагностику и упрощать принятие решений. В этом систематическом обзоре рассматривается эволюция LLM от рекуррентных нейронных сетей (RNN) до основанных на трансформаторах и многомодальных архитектур (например, BioBERT, Med-PaLM), с акцентом на их применение в медицинской практике и проблемы, с которыми они сталкиваются в России. Согласно 40 рецензируемым статьям из Scopus, PubMed и других надёжных источников (2019-2025 гг.), LLM демонстрируют высокую производительность (например, Med-PaLM: F1-критерий 0,88 для бинарной классификации пневмонии на MIMIC-CXR; Flamingo-CXR: предпочтение 77,7% для стационарных/амбулаторных рентгенологических заключений). Однако к ограничениям относятся дефицит данных, трудности с интерпретацией и вопросы конфиденциальности. Адаптация архитектуры «Смесь экспертов» (MoE) для диагностики редких заболеваний и автоматизированного создания отчётов по радиологии дала многообещающие результаты на синтетических наборах данных. В России существуют такие проблемы, как ограниченный объём аннотированных данных и соблюдение Федерального закона № 152-ФЗ. LLM улучшают клинические рабочие процессы, автоматизируя рутинные задачи, такие как создание отчётов и сортировка пациентов, благодаря передовым моделям, таким как KARGEN, повышающим качество отчётов по радиологии. Ориентация России на здравоохранение на основе ИИ соответствует мировым тенденциям, однако лингвистические и инфраструктурные барьеры требуют разработки индивидуальных решений. Разработка надёжных фреймворков валидации для LLM обеспечит их надёжность в различных клинических сценариях. Совместные усилия с международными исследовательскими сообществами в области ИИ могут ускорить внедрение в России передовых медицинских технологий ИИ, особенно в области автоматизации радиологии. Перспективы включают интеграцию LLM с системами здравоохранения и разработку специализированных моделей для российского медицинского контекста. Данное исследование закладывает основу для развития здравоохранения на основе ИИ в России.</p></trans-abstract><kwd-group xml:lang="en"><kwd>large language models</kwd><kwd>healthcare</kwd><kwd>deep learning</kwd><kwd>clinical text analysis</kwd><kwd>radiology report generation</kwd><kwd>interpretability</kwd><kwd>Russian healthcare</kwd></kwd-group><kwd-group xml:lang="ru"><kwd>большие языковые модели</kwd><kwd>здравоохранение</kwd><kwd>глубокое обучение</kwd><kwd>анализ клинических текстов</kwd><kwd>создание отчётов по рентгенологии</kwd><kwd>интерпретируемость</kwd><kwd>российское здравоохранение</kwd></kwd-group><funding-group><award-group><funding-source><institution-wrap><institution xml:lang="en">The research was carried out with the financial support of Sevastopol State University, project No. 42-01-09/319/2025-1.</institution></institution-wrap><institution-wrap><institution xml:lang="ru">The research was carried out with the financial support of Sevastopol State University, project No. 42-01-09/319/2025-1.</institution></institution-wrap></funding-source></award-group></funding-group></article-meta><fn-group/></front><body></body><back><ref-list><ref id="B1"><label>1.</label><mixed-citation>Tu, T., Azizi, S., Singhal, K., et al. Med-PaLM M: A multimodal generative foundation model for health 2024.</mixed-citation></ref><ref id="B2"><label>2.</label><mixed-citation>Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., et al. Large language models in medicine: Opportunities and challenges. Nature Medicine 29, 1930-1940. doi:10.1038/s41591-023-02448-8 (2023).</mixed-citation></ref><ref id="B3"><label>3.</label><mixed-citation>Sultan, I. Revolutionizing precision oncology: The role of artificial intelligence in personalized pediatric cancer care. Frontiers in Medicine 12, 1555893. doi:10.3389/fmed.2025.1555893 (2025).</mixed-citation></ref><ref id="B4"><label>4.</label><mixed-citation>Lee, J., Yoon, W., Kim, S., Kim, D., Kim, S., So, C. H. &amp; Kang, J. BioBERT: A pre-trained biomedical language representation model for biomedical text mining. Bioinformatics 36, 1234-1240. doi:10.1093/bioinformatics/btz682 (2020).</mixed-citation></ref><ref id="B5"><label>5.</label><mixed-citation>Huang, K., Altosaar, J. &amp; Ranganath, R. ClinicalBERT: Modeling clinical notes and predicting hospital readmission 2019.</mixed-citation></ref><ref id="B6"><label>6.</label><mixed-citation>Li, B., Zhang, Y., Chen, L., Wang, J., Yang, J. &amp; Liu, Z. BLIP-2: Bootstrapping language-image pre-training with frozen image encoders and large language models in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023), 197-208. doi:10.1109/CVPR52729. 2023.00025.</mixed-citation></ref><ref id="B7"><label>7.</label><mixed-citation>Lin, T. Y., Zhang, Y. &amp; Chen, X. Mixture of experts for medical imaging and text. Medical Physics 51, 1234-1245. doi:10.1002/mp.16890 (2024).</mixed-citation></ref><ref id="B8"><label>8.</label><mixed-citation>Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Hassabis, D., et al. AlphaFold for drug discovery: Protein-ligand interaction prediction. Nature 614, 709-716. doi:10. 1038/s41586-023-05788-0 (2023).</mixed-citation></ref><ref id="B9"><label>9.</label><mixed-citation>Thériault-Lauzier, P. et al. Temporal learning for longitudinal imaging-based recurrence prediction in pediatric gliomas. NEJM AI 2. doi:10.1056/AIra2400123 (2025).</mixed-citation></ref><ref id="B10"><label>10.</label><mixed-citation>Li, Y., Wang, Z., Liu, Y., Zhou, L., et al. KARGEN: Knowledge-enhanced automated radiology report generation using large language models 2024.</mixed-citation></ref><ref id="B11"><label>11.</label><mixed-citation>Liu, Z., Zhang, Y. &amp; Chen, X. BioBART for accelerated biomedical literature review. Bioinformatics doi:10.1093/bioinformatics/btad456 (2023).</mixed-citation></ref><ref id="B12"><label>12.</label><mixed-citation>Grisoni, F. ChemBERTa: A chemical language model for drug discovery. Journal of Chemical Information and Modeling 63, 1345-1353. doi:10.1021/acs.jcim.2c01567 (2023).</mixed-citation></ref><ref id="B13"><label>13.</label><mixed-citation>Van Veen, D. et al. Collaboration between clinicians and vision-language models in radiology report generation. Nature Medicine 30, 3056-3064. doi:10.1038/s41591-024-03208-y (2024).</mixed-citation></ref><ref id="B14"><label>14.</label><mixed-citation>Bannur, S. et al. RaDialog: A large vision-language model for radiology report generation and conversational assistance 2025.</mixed-citation></ref><ref id="B15"><label>15.</label><mixed-citation>Kasakewitch, J. P. G., Lima, D. L., Balthazar, C. A., et al. The Role of Artificial Intelligence Large Language Models in Literature Search Assistance to Evaluate Inguinal Hernia Repair Approaches. Journal of Laparoendoscopic &amp; Advanced Surgical Techniques 35, 437-444. doi:10.1089/lap.2024.0277 (2025).</mixed-citation></ref><ref id="B16"><label>16.</label><mixed-citation>Wu, X., Zhang, Y. &amp; Chen, L. Visual ChatGPT: Multimodal dialogue for medical applications in Medical Image Computing and Computer Assisted Intervention 14221 (2023), 345-354. doi:10.1007/ 978-3-031-43901-8_33.</mixed-citation></ref><ref id="B17"><label>17.</label><mixed-citation>Delgado, D. Artificial Intelligence-Enabled Analysis of Thermography to Diagnose Acute Decompensated Heart Failure. JACC: Advances 4, 101888. doi:10.1016/j.jacadv.2025.101888 (2025).</mixed-citation></ref><ref id="B18"><label>18.</label><mixed-citation>Arora, A. &amp; Arora, A. The promise of large language models in health care. The Lancet 401, 641. doi:10.1016/S0140-6736(23)00217-6 (2023).</mixed-citation></ref><ref id="B19"><label>19.</label><mixed-citation>Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F. &amp; Ting, D. S. W. Federated learning for medical AI: A practical approach. Nature Machine Intelligence 5, 389-398. doi:10.1038/s42256-023-00645-8 (2023).</mixed-citation></ref><ref id="B20"><label>20.</label><mixed-citation>Namiri, N. K., Puglisi, C. E. &amp; Lipsky, P. E. Machine learning in rheumatic autoimmune inflammatory diseases. Nature Reviews Rheumatology 17, 669-680. doi:10.1038/s41584-02100692-7 (2021).</mixed-citation></ref><ref id="B21"><label>21.</label><mixed-citation>Touvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Scialom, T., et al. Llama 2: Open foundation and fine-tuned chat models 2023.</mixed-citation></ref><ref id="B22"><label>22.</label><mixed-citation>Meskó, B. &amp; Görög, M. A short guide for medical professionals in the era of artificial intelligence. NPJ Digital Medicine 3, 126. doi:10.1038/s41746-020-00333-z (2020).</mixed-citation></ref><ref id="B23"><label>23.</label><mixed-citation>Beltagy, I., Lo, K. &amp; Cohan, A. SciBERT: A pretrained language model for scientific text in Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (2023), 1234-1245. doi:10.18653/v1/2023.emnlp-main.76.</mixed-citation></ref><ref id="B24"><label>24.</label><mixed-citation>Wang, J., Zhang,Y. &amp; Li, X. Multimodal LLMs for scientific trend prediction. JournalofInformetrics doi:10.1016/j.joi.2024.101345 (2024).</mixed-citation></ref><ref id="B25"><label>25.</label><mixed-citation>Chen, Y., Zhang, L. &amp; Wang, J. GPT-3 for drug repurposing in infectious diseases. Journal of Medical Chemistry 66, 2345-2353. doi:10.1021/acs.jmedchem.2c01567 (2023).</mixed-citation></ref><ref id="B26"><label>26.</label><mixed-citation>Khader, F., Müller-Franzes, G. &amp; Wang, S. Synthetic data generation for medical imaging using GANs. Medical Image Analysis 78. doi:10.1016/j.media.2022.102399 (2022).</mixed-citation></ref><ref id="B27"><label>27.</label><mixed-citation>Rajpurkar, P., Chen, E., Banerjee, O. &amp; Topol, E. J. AI in health and medicine. Nature Medicine 28, 31-38. doi:10.1038/s41591-021-01614-0 (2022).</mixed-citation></ref><ref id="B28"><label>28.</label><mixed-citation>Loaiza-Bonilla, A. &amp; Penberthy, S. Challenges in integrating artificial intelligence into health care: Bias, privacy, and validation. NEJM AI 2. doi:10.1056/AIp2400789 (2025).</mixed-citation></ref><ref id="B29"><label>29.</label><mixed-citation>Che, J., Zhang, X. &amp; Liu, Y. Ensemble methods for improving LLM reliability. Journal of Artificial Intelligence Research 68, 567-580. doi:10.1613/jair.1.13845 (2023).</mixed-citation></ref><ref id="B30"><label>30.</label><mixed-citation>Chowdhery, A., Narang, S., Devlin, J., Bosma, M., Mishra, G., Roberts, A., Fiedel, N., et al. PaLM: Scaling language modeling with pathways 2022.</mixed-citation></ref><ref id="B31"><label>31.</label><mixed-citation>Radanliev, P. &amp; De Roure, D. The ethics of shared Covid-19 risks: An epistemological framework for ethical health technology assessment of risk in vaccine supply chain infrastructures. Health and Technology 11, 1083-1091. doi:10.1007/s12553-021-00587-3 (2021).</mixed-citation></ref><ref id="B32"><label>32.</label><mixed-citation>Pang, T., Li, P. &amp; Zhao, L. A survey on automatic generation of medical imaging reports based on deep learning. BioMedical Engineering Online 22, 48. doi:10.1186/s12938-023-01113-y (2023).</mixed-citation></ref><ref id="B33"><label>33.</label><mixed-citation>Kim, S. et al. Large language models: A guide for radiologists. Korean Journal of Radiology 25, 126-133. doi:10.3348/kjr.2023.0997 (2024).</mixed-citation></ref><ref id="B34"><label>34.</label><mixed-citation>Fink, M. A. et al. Automatic structuring of radiology reports with on-premise open-source large language models. European Radiology 34, 6285-6294. doi:10.1007/s00330-024-09876-5 (2024).</mixed-citation></ref><ref id="B35"><label>35.</label><mixed-citation>López-Úbeda, P., Martín-Noguerol, T., Díaz-Angulo, C. &amp; Luna, A. Evaluation of large language models performance against humans for summarizing MRI knee radiology reports: A feasibility study. International Journal of Medical Informatics 187, 105443. doi:10.1016/j.ijmedinf.2024.105443 (2024).</mixed-citation></ref><ref id="B36"><label>36.</label><mixed-citation>Jorg, T. et al. Automated integration of AI results into radiology reports using common data elements. Journal of Imaging Informatics in Medicine 38, 45-53. doi:10.1007/s10278-024-01023-4 (2025).</mixed-citation></ref><ref id="B37"><label>37.</label><mixed-citation>Gertz, R. J., Bunck, A. C., Lennartz, S., et al. GPT-4 for automated determination of radiological study and protocol based on radiology request forms: A feasibility study. Radiology 307, e230877. doi:10.1148/radiol.230877 (2023).</mixed-citation></ref></ref-list></back></article>
