{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T06:40:03Z","timestamp":1782456003788,"version":"3.54.5"},"reference-count":26,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,4,17]],"date-time":"2025-04-17T00:00:00Z","timestamp":1744848000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Med."],"abstract":"<jats:p>Pulmonary diseases, such as pneumonia and lung abscess, can trigger sepsis, while sepsis-induced immune dysfunction exacerbates Pulmonary tissue damage, creating a vicious cycle. Therefore, designing a safe and effective clinical treatment planning method for sepsis is critically significant. In recent years, deep reinforcement learning (DRL), as one of the artificial intelligence technologies, has achieved remarkable results in the field of sepsis treatment. However, DRL models may be attacked due to their sensitive training data and their high commercial value, especially with the increasing number of DRL models being released on the Internet. Consequently, protecting the \u201cprivacy\u201d of DRL models and training data has become an urgent problem. To address this issue, we propose a differential privacy-based DRL model for sepsis treatment. Furthermore, we investigate the impact of differential privacy mechanisms on the performance of the DRL model. Experimental results demonstrate that integrating differential privacy into DRL models enables clinicians to design sepsis treatment plans while protecting patient privacy, thereby mitigating lung tissue damage and dysfunction caused by sepsis.<\/jats:p>","DOI":"10.3389\/fmed.2025.1590824","type":"journal-article","created":{"date-parts":[[2025,4,17]],"date-time":"2025-04-17T04:10:51Z","timestamp":1744863051000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Integrating differential privacy in deep reinforcement learning for sepsis treatment with pulmonary implications"],"prefix":"10.3389","volume":"12","author":[{"given":"Shuling","family":"Wang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Feng","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Suixue","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rongdao","family":"Sun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,4,17]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"2133","DOI":"10.1056\/NEJMra2403213","article-title":"Sepsis and septic shock","volume":"391","author":"Meyer","year":"2024","journal-title":"N Engl J Med"},{"key":"B2","doi-asserted-by":"publisher","first-page":"409","DOI":"10.1211\/pj.2024.1.323850","article-title":"Sepsis: diagnosis and management","volume":"101","author":"Gauer","year":"2020","journal-title":"Am Fam Physician"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1312.5602","article-title":"Playing atari with deep reinforcement learning","author":"Mnih","year":"2013","journal-title":"Comput Sci"},{"key":"B4","article-title":"A markov decision process to suggest optimal treatment of severe infections in intensive care","volume-title":"Neural Information Processing Systems Workshop on Machine Learning for Health","author":"Komorowski","year":"2016"},{"key":"B5","doi-asserted-by":"crossref","first-page":"2621","DOI":"10.1109\/BIBM62325.2024.10821720","article-title":"ContraMAE: contrastive alignment masked autoencoder framework for cancer survival prediction","volume-title":"2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","author":"Wang","year":"2024"},{"key":"B6","doi-asserted-by":"publisher","first-page":"5251","DOI":"10.1109\/TCE.2024.3413732","article-title":"A cloud-edge collaboration framework for cancer survival prediction to develop medical consumer electronic devices","volume":"70","author":"Wang","year":"2024","journal-title":"IEEE Trans Consum Electron"},{"key":"B7","doi-asserted-by":"crossref","first-page":"642","DOI":"10.1109\/BIBM58861.2023.10385635","article-title":"HC-MAE: hierarchical cross-attention masked autoencoder integrating histopathological images and multi-omics for cancer survival prediction","volume-title":"2023 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","author":"Wang","year":"2023"},{"key":"B8","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1145\/3446776","article-title":"Understanding deep learning (still) requires rethinking generalization","volume":"64","author":"Zhang","year":"2021","journal-title":"Commun ACM"},{"key":"B9","doi-asserted-by":"crossref","DOI":"10.1109\/SP.2017.41","article-title":"Membership inference attacks against machine learning models","volume-title":"IEEE Symposium on Security and Privacy (SP)","author":"Shokri","year":"2017"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1904.11082","article-title":"How you act tells a lot: privacy-leakage attack on deep reinforcement learning","author":"Pan","year":"2019","journal-title":"ArXiv Preprint ArXiv:190411082"},{"key":"B11","first-page":"601","article-title":"Stealing machine learning models via prediction APIs","volume-title":"25th USENIX Security Symposium (USENIX Security 16)","author":"Tram\u00e8r","year":"2016"},{"key":"B12","doi-asserted-by":"publisher","first-page":"8952219","DOI":"10.1155\/2021\/8952219","article-title":"A privacy-preserving reinforcement learning approach for dynamic treatment regimes on health data","volume":"2021","author":"Sun","year":"2021","journal-title":"Wirel Commun Mob Comput"},{"key":"B13","doi-asserted-by":"crossref","first-page":"456","DOI":"10.1109\/TETC.2019.2896325","article-title":"Privacy-preserving reinforcement learning design for patient-centric dynamic treatment regimes","volume":"9","author":"Liu","year":"2019","journal-title":"IEEE Trans Emerg Top Comput."},{"key":"B14","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/JIOT.2021.3057653","article-title":"A resource-constrained and privacy-preserving edge computing enabled clinical decision system: a federated reinforcement learning approach","volume":"99","author":"Xue","year":"2021","journal-title":"IEEE Internet Things J"},{"key":"B15","article-title":"Private q-learning with functional noise in continuous spaces","author":"Wang","year":"2019","journal-title":"ArXiv Preprint ArXiv:190110634"},{"key":"B16","first-page":"387","article-title":"Corrupt bandits for preserving local privacy","volume-title":"Algorithmic Learning Theory","author":"Gajane","year":"2018"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1905.12298","article-title":"Differential privacy for multi-armed bandits: what is it and what is its cost?","author":"Basu","year":"2019","journal-title":"ArXiv Preprint ArXiv:190512298"},{"key":"B18","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1901.10634","article-title":"Locally private distributed reinforcement learning","author":"Ono","year":"2020","journal-title":"ArXiv Preprint ArXiv:200111718"},{"key":"B19","doi-asserted-by":"publisher","first-page":"338","DOI":"10.1007\/978-1-4419-5906-5_752","article-title":"Differential privacy","author":"Dwork","year":"2011","journal-title":"Encyclopedia of Cryptography and Security"},{"key":"B20","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1561\/0400000042","article-title":"The algorithmic foundations of differential privacy","volume":"9","author":"Dwork","year":"2014","journal-title":"Found Trends Theor Comput Sci"},{"key":"B21","doi-asserted-by":"publisher","first-page":"265","DOI":"10.1007\/11681878_14","article-title":"Calibrating noise to sensitivity in private data analysis","author":"Dwork","year":"2006","journal-title":"Theory of Cryptography Conference"},{"key":"B22","doi-asserted-by":"crossref","first-page":"308","DOI":"10.1145\/2976749.2978318","article-title":"Deep learning with differential privacy","volume-title":"Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security","author":"Abadi","year":"2016"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1711.09602","article-title":"Deep reinforcement learning for sepsis treatment","author":"Raghu","year":"2017","journal-title":"ArXiv Preprint ArXiv:171109602"},{"key":"B24","doi-asserted-by":"publisher","first-page":"801","DOI":"10.1001\/jama.2016.0287","article-title":"The third international consensus definitions for sepsis and septic shock (Sepsis-3)","volume":"315","author":"Singer","year":"2016","journal-title":"JAMA"},{"key":"B25","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/sdata.2016.35","article-title":"MIMIC-III, a freely accessible critical care database","volume":"3","author":"Johnson","year":"2016","journal-title":"Sci Data"},{"key":"B26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1511.06807","article-title":"Adding gradient noise improves learning for very deep networks","author":"Neelakantan","year":"2015","journal-title":"ArXiv Preprint ArXiv:151106807"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2025.1590824\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,17]],"date-time":"2025-04-17T04:10:57Z","timestamp":1744863057000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2025.1590824\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,17]]},"references-count":26,"alternative-id":["10.3389\/fmed.2025.1590824"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2025.1590824","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,17]]},"article-number":"1590824"}}