{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T19:41:31Z","timestamp":1789674091603,"version":"4.0.1"},"reference-count":25,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Diabetes"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec sec-type=\"background\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Effective diabetes management requires precise glycemic control to prevent both hypoglycemia and hyperglycemia, yet existing machine learning (ML) and reinforcement learning (RL) approaches often fail to balance competing objectives. Traditional RL-based glucose regulation systems primarily focus on single-objective optimization, overlooking factors such as minimizing insulin overuse, reducing glycemic variability, and ensuring patient safety. Furthermore, these approaches typically rely on centralized data processing, which raises privacy concerns due to the sensitive nature of health care data. There is a critical need for a decentralized, privacy-preserving framework that can personalize blood glucose regulation while addressing the multiobjective nature of diabetes management.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"objective\">\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>This study aimed to develop and validate PRIMO-FRL (Privacy-Preserving Reinforcement Learning for Individualized Multi-Objective Glycemic Management Using Federated Reinforcement Learning), a novel framework that optimizes clinical objectives\u2014maximizing time in range (TIR), reducing hypoglycemia and hyperglycemia, and minimizing glycemic risk\u2014while preserving patient privacy.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"methods\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We developed PRIMO-FRL, integrating multiobjective reward shaping to dynamically balance glucose stability, insulin efficiency, and risk reduction. The model was trained and tested using simulated data from 30 simulated patients (10 children, 10 adolescents, and 10 adults) generated with the Food and Drug Administration (FDA)\u2013approved UVA\/Padova simulator. A comparative analysis was conducted against state-of-the-art RL and ML models, evaluating performance using metrics such as TIR, hypoglycemia (&lt;70 mg\/dL), hyperglycemia (&gt;180 mg\/dL), and glycemic risk scores.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"results\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The PRIMO-FRL model achieved a robust overall TIR of 76.54%, with adults demonstrating the highest TIR at 81.48%, followed by children at 77.78% and adolescents at 70.37%. Importantly, the approach eliminated hypoglycemia, with 0.0% spent below 70\u202fmg\/dL across all cohorts, significantly outperforming existing methods. Mild hyperglycemia (180-250\u202fmg\/dL) was observed in adolescents (29.63%), children (22.22%), and adults (18.52%), with adults exhibiting the best control. Furthermore, the PRIMO-FRL approach consistently reduced glycemic risk scores, demonstrating improved safety and long-term stability in glucose regulation..<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"conclusions\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Our findings highlight the potential of PRIMO-FRL as a transformative, privacy-preserving approach to personalized glycemic management. By integrating federated RL, this framework eliminates hypoglycemia, improves TIR, and preserves data privacy by decentralizing model training. Unlike traditional centralized approaches that require sharing sensitive health data, PRIMO-FRL leverages federated learning to keep patient data local, significantly reducing privacy risks while enabling adaptive and personalized glucose control. This multiobjective optimization strategy offers a scalable, secure, and clinically viable solution for real-world diabetes care. The ability to train personalized models across diverse populations without exposing raw data makes PRIMO-FRL well-suited for deployment in privacy-sensitive health care environments. These results pave the way for future clinical adoption, demonstrating the potential of privacy-preserving artificial intelligence in optimizing glycemic regulation while maintaining security, adaptability, and personalization.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/72874","type":"journal-article","created":{"date-parts":[[2025,7,4]],"date-time":"2025-07-04T11:54:01Z","timestamp":1751630041000},"page":"e72874-e72874","source":"Crossref","is-referenced-by-count":4,"title":["Privacy-Preserving Glycemic Management in Type 1 Diabetes: Development and Validation of a Multiobjective Federated Reinforcement Learning Framework"],"prefix":"10.2196","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7088-2561","authenticated-orcid":false,"given":"Fatemeh","family":"Sarani Rad","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7668-5996","authenticated-orcid":false,"given":"Juan","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2025,7,4]]},"reference":[{"issue":"1","key":"R1","doi-asserted-by":"publisher","first-page":"13","DOI":"10.2337\/db20211-13","article-title":"The interplay between diabetes, cardiovascular disease, and kidney disease","volume":"2021","author":"Usman","journal-title":"ADA Clinical Compendia"},{"issue":"4","key":"R2","doi-asserted-by":"publisher","first-page":"399","DOI":"10.2337\/ds21-0014","article-title":"Literature review of type 2 diabetes management and health literacy","volume":"34","author":"Alsaedi","journal-title":"Diabetes Spectr"},{"issue":"5","key":"R3","doi-asserted-by":"publisher","first-page":"1191","DOI":"10.2337\/dc13-2108","article-title":"Closed-loop artificial pancreas systems: engineering the algorithms","volume":"37","author":"Doyle","journal-title":"Diabetes Care"},{"key":"R4","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2020.101836","article-title":"Reinforcement learning application in diabetes blood glucose control: a systematic review","volume":"104","author":"Tejedor","journal-title":"Artif Intell Med"},{"issue":"9","key":"R5","doi-asserted-by":"publisher","DOI":"10.3390\/biomedicines12092143","article-title":"Reinforcement learning: a paradigm shift in personalized blood glucose management for diabetes","volume":"12","author":"D\u00e9nes-Fazakas","journal-title":"Biomedicines"},{"issue":"1","key":"R6","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-50764-5","article-title":"Personalized insulin dosing using reinforcement learning for high-fat meals and aerobic exercises in type 1 diabetes: a proof-of-concept trial","volume":"15","author":"Jafar","journal-title":"Nat Commun"},{"issue":"9","key":"R7","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0274608","article-title":"Evaluation of blood glucose level control in type 1 diabetic patients using deep reinforcement learning","volume":"17","author":"Viroonluecha","journal-title":"PLoS ONE"},{"key":"R8","unstructured":"Fox I Lee J Pop-Busui R Wiens J . Deep reinforcement learning for closed-loop blood glucose control. In: Proceedings of the 5th Machine Learning for Healthcare Conference. PMLR; 2020:508-536. URL: https:\/\/proceedings.mlr.press\/v126\/fox20a.html [Accessed 10-06-2025]"},{"issue":"14","key":"R9","doi-asserted-by":"publisher","first-page":"11475","DOI":"10.1007\/s00521-020-05519-w","article-title":"Blockchain for healthcare data management: opportunities, challenges, and future recommendations","volume":"34","author":"Yaqoob","journal-title":"Neural Comput & Applic"},{"key":"R10","doi-asserted-by":"crossref","unstructured":"Yang X Li J . Edge AI empowered personalized privacy-preserving glucose prediction with federated deep learning. Presented at: 2023 IEEE International Conference on E-health Networking, Application & Services (Healthcom); Dec 15-17, 2023; Chongqing, China. [doi: 10.1109\/Healthcom56612.2023.10472368]","DOI":"10.1109\/Healthcom56612.2023.10472368"},{"key":"R11","doi-asserted-by":"crossref","unstructured":"Wang S Gu W . An improved strategy for blood glucose control using multi-step deep reinforcement learning. Presented at: ICBBT \u201924: Proceedings of the 2024 16th International Conference on Bioinformatics and Biomedical Technology; May 24-26, 2024; Chongqing, China. [doi: 10.1145\/3674658.3674689]","DOI":"10.1145\/3674658.3674689"},{"key":"R12","doi-asserted-by":"publisher","DOI":"10.2196\/60834","article-title":"Effectiveness of a digital health intervention leveraging reinforcement learning: results from the Diabetes and Mental Health Adaptive Notification Tracking and Evaluation (DIAMANTE) randomized clinical trial","volume":"26","author":"Aguilera","journal-title":"J Med Internet Res"},{"issue":"7","key":"R13","doi-asserted-by":"publisher","DOI":"10.2196\/27858","article-title":"Effective treatment recommendations for type 2 diabetes management using reinforcement learning: treatment recommendation model development and validation","volume":"23","author":"Sun","journal-title":"J Med Internet Res"},{"key":"R14","unstructured":"Sarani Rad F Li J . Optimizing blood glucose control through reward shaping in reinforcement learning. Presented at: Proceedings of the IEEE International Conference on E-health Networking, Application & Services (HealthCom); Dec 15-17, 2023; Chongqing, China. URL: https:\/\/ieeexplore.ieee.org\/abstract\/document\/10472341 [Accessed 07-06-2025]"},{"issue":"1","key":"R15","doi-asserted-by":"publisher","first-page":"230","DOI":"10.3390\/engproc2023059230","article-title":"Federated learning for healthcare: a comprehensive review","volume":"59","author":"Dhade","journal-title":"Eng Proc"},{"issue":"1","key":"R16","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-021-00489-2","article-title":"Privacy-first health research with federated learning","volume":"4","author":"Sadilek","journal-title":"NPJ Digit Med"},{"key":"R17","doi-asserted-by":"publisher","DOI":"10.20517\/ir.2021.02","article-title":"Federated reinforcement learning: techniques, applications, and open challenges","author":"Qi","journal-title":"IR"},{"issue":"9","key":"R18","doi-asserted-by":"publisher","first-page":"3113","DOI":"10.1007\/s00607-024-01317-7","article-title":"Federated learning for digital healthcare: concepts, applications, frameworks, and challenges","volume":"106","author":"Sachin","journal-title":"Computing"},{"issue":"11","key":"R19","doi-asserted-by":"publisher","first-page":"6497","DOI":"10.3390\/app13116497","article-title":"Federated reinforcement learning in IoT: applications, opportunities and open challenges","volume":"13","author":"Pinto Neto","journal-title":"Appl Sci"},{"key":"R20","doi-asserted-by":"publisher","DOI":"10.2196\/58637","article-title":"Decades in the making: the evolution of digital health research infrastructure through synthetic data, common data models, and federated learning","volume":"26","author":"Austin","journal-title":"J Med Internet Res"},{"key":"R21","doi-asserted-by":"publisher","DOI":"10.2196\/55496","article-title":"Accessible ecosystem for clinical research (federated learning for everyone): development and usability study","volume":"8","author":"Pirmani","journal-title":"JMIR Form Res"},{"key":"R22","unstructured":"Haarnoja T Zhou A Abbeel P et al. Soft actor-critic: off-policy maximum entropy deep reinforcement learning with a stochastic actor. arXiv. Preprint posted online on Jan 4, 2018. URL: https:\/\/arxiv.org\/abs\/1801.01290"},{"issue":"1","key":"R23","doi-asserted-by":"publisher","first-page":"26","DOI":"10.1177\/1932296813514502","article-title":"The UVA\/PADOVA type 1 diabetes simulator: new features","volume":"8","author":"Man","journal-title":"J Diabetes Sci Technol"},{"issue":"1","key":"R24","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1177\/193229680900300106","article-title":"In silico preclinical trials: a proof of concept in closed-loop control of type 1 diabetes","volume":"3","author":"Kovatchev","journal-title":"J Diabetes Sci Technol"},{"key":"R25","unstructured":"Xie J . Simglucose. GitHub. 2018. URL: https:\/\/github.com\/jxx123\/simglucose [Accessed 20-05-2023]"}],"container-title":["JMIR Diabetes"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/diabetes.jmir.org\/2025\/1\/e72874","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T18:03:19Z","timestamp":1782842599000},"score":1,"resource":{"primary":{"URL":"https:\/\/diabetes.jmir.org\/2025\/1\/e72874"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,4]]},"references-count":25,"URL":"https:\/\/doi.org\/10.2196\/72874","relation":{},"ISSN":["2371-4379"],"issn-type":[{"value":"2371-4379","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7,4]]},"article-number":"v10i1e72874"}}