{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,14]],"date-time":"2026-08-14T05:15:10Z","timestamp":1786684510246,"version":"build-2736575974"},"reference-count":45,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J Med Internet Res"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Traumatic brain injury (TBI) is a critically ill disease with a high mortality rate, and clinical treatment is committed to continuously optimizing treatment strategies to improve survival rates.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aims to establish a reinforcement learning algorithm (RL) to optimize the survival prognosis decision-making scheme for patients with TBI in the intensive care unit<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n We included a total of 2745 patients from the Medical Information Mart for Intensive Care (MIMIC)\u2013IV database and randomly divided them into a training set and an internal validation set at 8:2. We extracted 34 features for analysis and modeling using a 2-hour time compensation, 2 action features (mean arterial pressure and temperature), and 1 outcome feature (survival status at 28 d). We used an RL algorithm called weighted dueling double deep Q-network with embedded human expertise to maximize cumulative returns and evaluated the model using a doubly robust off-policy evaluation method. Finally, we collected 2463 patients with TBI from MIMIC III as an external validation set to test the model.<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n The action features are divided into 6 intervals, and the expected benefits are estimated using a doubly robust off-policy evaluation method. The results indicate that the survival rate of artificial intelligence (AI) strategies is higher than that of clinical doctors (88.016%, 95% CI 85.191%\u201090.840% vs 81.094%, 95% CI 80.422%\u201081.765%), with an expected return of 28.978 (95% CI 28.797\u201029.160) versus 27.092 (95% CI 24.584\u201029.600). Compared with clinical doctors, AI algorithms select normal temperatures more frequently (36.56 \u00b0C to 36.83 \u2103) and recommend mean arterial pressure levels of 87.5\u201095.0 mm Hg. In external validation, the AI strategy still has a high survival rate of 87.565%, with an expected return of 27.517.<\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n This RL algorithm for patients with TBI indicates that a more personalized and targeted optimization of the vital signs is possible. This algorithm will assist clinicians in making decisions on an individualized patient-by-patient basis.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/63847","type":"journal-article","created":{"date-parts":[[2025,7,3]],"date-time":"2025-07-03T12:11:37Z","timestamp":1751544697000},"page":"e63847-e63847","source":"Crossref","is-referenced-by-count":4,"title":["Optimizing Vital Signs in Patients With Traumatic Brain Injury: Reinforcement Learning Algorithm Development and Validation"],"prefix":"10.2196","volume":"27","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-3793-9968","authenticated-orcid":false,"given":"Hongwei","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7989-5067","authenticated-orcid":false,"given":"Mengyuan","family":"Diao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-8930-3726","authenticated-orcid":false,"given":"Sheng","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2963-2648","authenticated-orcid":false,"given":"Peifeng","family":"Ni","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-1093-5817","authenticated-orcid":false,"given":"Weidong","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-4845-2652","authenticated-orcid":false,"given":"Chenxi","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4360-1653","authenticated-orcid":false,"given":"Ying","family":"Zhu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7113-9040","authenticated-orcid":false,"given":"Wei","family":"Hu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2025,7,3]]},"reference":[{"issue":"2","key":"R1","doi-asserted-by":"publisher","first-page":"213","DOI":"10.1016\/j.mcna.2019.11.001","article-title":"Traumatic brain injury","volume":"104","author":"Capizzi","journal-title":"Med Clin North Am"},{"issue":"3","key":"R2","doi-asserted-by":"publisher","first-page":"286","DOI":"10.1016\/S1474-4422(18)30469-1","article-title":"Traumatic brain injury in China","volume":"18","author":"Jiang","journal-title":"Lancet Neurol"},{"issue":"5","key":"R3","doi-asserted-by":"publisher","first-page":"507","DOI":"10.1080\/10903127.2023.2187905","article-title":"Prehospital guidelines for the management of traumatic brain injury - 3rd edition","volume":"27","author":"Lulla","journal-title":"Prehosp Emerg Care"},{"issue":"5","key":"R4","doi-asserted-by":"publisher","first-page":"1015","DOI":"10.1016\/j.suc.2017.06.003","article-title":"Acute management of traumatic brain injury","volume":"97","author":"Vella","journal-title":"Surg Clin North Am"},{"key":"R5","doi-asserted-by":"publisher","DOI":"10.3389\/fmed.2022.766447","article-title":"Establishment and Implementation of potential fluid therapy balance strategies for ICU sepsis patients based on reinforcement learning","volume":"9","author":"Su","journal-title":"Front Med (Lausanne)"},{"issue":"11","key":"R6","doi-asserted-by":"publisher","first-page":"1716","DOI":"10.1038\/s41591-018-0213-5","article-title":"The artificial intelligence clinician learns optimal treatment strategies for sepsis in intensive care","volume":"24","author":"Komorowski","journal-title":"Nat Med"},{"issue":"5","key":"R7","doi-asserted-by":"publisher","DOI":"10.3390\/jcm13051505","article-title":"Using artificial intelligence to predict mechanical ventilation weaning success in patients with respiratory failure, including those with acute respiratory distress syndrome","volume":"13","author":"Stivi","journal-title":"J Clin Med"},{"issue":"1","key":"R8","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-021-00388-6","article-title":"Development and validation of a reinforcement learning algorithm to dynamically optimize mechanical ventilation in critical care","volume":"4","author":"Peine","journal-title":"NPJ Digit Med"},{"key":"R9","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2023.102742","article-title":"Guideline-informed reinforcement learning for mechanical ventilation in critical care","volume":"147","author":"den Hengst","journal-title":"Artif Intell Med"},{"key":"R10","doi-asserted-by":"publisher","DOI":"10.1016\/j.compchemeng.2020.106886","article-title":"A review on reinforcement learning: introduction and applications in industrial process control","volume":"139","author":"Nian","journal-title":"Comput Chem Eng"},{"issue":"7","key":"R11","doi-asserted-by":"publisher","DOI":"10.2196\/18477","article-title":"Reinforcement learning for clinical decision support in critical care: comprehensive review","volume":"22","author":"Liu","journal-title":"J Med Internet Res"},{"issue":"1","key":"R12","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-023-00755-5","article-title":"A value-based deep reinforcement learning model with human expertise in optimal treatment of sepsis","volume":"6","author":"Wu","journal-title":"NPJ Digit Med"},{"key":"R13","doi-asserted-by":"crossref","unstructured":"Puterman ML . Markov decision processes. In: Handbooks in Operations Research and Management Science. Elsevier; 1990:331-434. [doi: 10.1016\/S0927-0507(05)80172-0] ISBN: 978-0-444-87473-3","DOI":"10.1016\/S0927-0507(05)80172-0"},{"issue":"4","key":"R14","doi-asserted-by":"publisher","first-page":"474","DOI":"10.1177\/0272989X09353194","article-title":"Markov decision processes: a tool for sequential decision making under uncertainty","volume":"30","author":"Alagoz","journal-title":"Med Decis Making"},{"key":"R15","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2020.101964","article-title":"Reinforcement learning for intelligent healthcare applications: a survey","volume":"109","author":"Coronato","journal-title":"Artif Intell Med"},{"issue":"1","key":"R16","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1177\/08850666231188051","article-title":"Predictive performance for hospital mortality of SAPS 3, SOFA, ISS, and new ISS in critically Ill trauma patients: a validation cohort study","volume":"39","author":"Roepke","journal-title":"J Intensive Care Med"},{"issue":"14","key":"R17","doi-asserted-by":"publisher","first-page":"1754","DOI":"10.1001\/jama.286.14.1754","article-title":"Serial evaluation of the SOFA score to predict outcome in critically ill patients","volume":"286","author":"Ferreira","journal-title":"JAMA"},{"issue":"1","key":"R18","doi-asserted-by":"publisher","first-page":"25","DOI":"10.1007\/s00068-012-0212-z","article-title":"Serial lactate and admission SOFA scores in trauma: an analysis of predictive value in 724 patients with and without traumatic brain injury","volume":"39","author":"D\u00fcbendorfer","journal-title":"Eur J Trauma Emerg Surg"},{"issue":"3","key":"R19","doi-asserted-by":"publisher","first-page":"1017","DOI":"10.1007\/s12028-020-01121-z","article-title":"Mean arterial pressure and discharge outcomes in severe pediatric traumatic brain injury","volume":"34","author":"Erickson","journal-title":"Neurocrit Care"},{"issue":"12","key":"R20","doi-asserted-by":"publisher","first-page":"e512","DOI":"10.1161\/STROKEAHA.122.041274","article-title":"Interaction of mean arterial pressure and blood pressure variability in critically Ill brain injured patients","volume":"53","author":"de Havenon","journal-title":"Stroke"},{"issue":"2","key":"R21","doi-asserted-by":"publisher","first-page":"329","DOI":"10.1089\/neu.2006.0035","article-title":"Multivariable prognostic analysis in traumatic brain injury: results from the IMPACT study","volume":"24","author":"Murray","journal-title":"J Neurotrauma"},{"issue":"11","key":"R22","doi-asserted-by":"publisher","first-page":"1833","DOI":"10.1016\/j.injury.2011.08.014","article-title":"Redefining hypotension in traumatic brain injury","volume":"43","author":"Berry","journal-title":"Injury"},{"issue":"5","key":"R23","doi-asserted-by":"publisher","first-page":"1135","DOI":"10.1097\/TA.0b013e31824af90b","article-title":"Traditional systolic blood pressure targets underestimate hypotension-induced secondary brain injury","volume":"72","author":"Brenner","journal-title":"J Trauma Acute Care Surg"},{"issue":"14","key":"R24","doi-asserted-by":"publisher","first-page":"1995","DOI":"10.1089\/neu.2020.7289","article-title":"Escalating mean arterial pressure in severe traumatic brain injury: a prospective, observational study","volume":"38","author":"Kow","journal-title":"J Neurotrauma"},{"issue":"7","key":"R25","doi-asserted-by":"publisher","first-page":"1616","DOI":"10.1097\/00003246-200207000-00036","article-title":"Changes in intracranial pressure and cerebral autoregulation in patients with severe traumatic brain injury","volume":"30","author":"Ter Minassian","journal-title":"Crit Care Med"},{"issue":"8","key":"R26","doi-asserted-by":"publisher","first-page":"1224","DOI":"10.2174\/1570159X15666170613083606","article-title":"Recent advances in pathophysiology of traumatic brain injury","volume":"16","author":"Kaur","journal-title":"Curr Neuropharmacol"},{"issue":"2","key":"R27","doi-asserted-by":"publisher","first-page":"420","DOI":"10.1016\/j.wneu.2015.03.039","article-title":"Attenuation of cell death in injured cortex after post-traumatic brain injury moderate hypothermia: possible involvement of autophagy pathway","volume":"84","author":"Jin","journal-title":"World Neurosurg"},{"issue":"3","key":"R28","doi-asserted-by":"publisher","first-page":"250","DOI":"10.1007\/BF00296740","article-title":"Post-traumatic brain hypothermia reduces histopathological damage following concussive brain injury in the rat","volume":"87","author":"Dietrich","journal-title":"Acta Neuropathol"},{"issue":"2","key":"R29","doi-asserted-by":"publisher","first-page":"370","DOI":"10.3171\/2012.10.JNS12725","article-title":"Neuroprotective effect of preoperatively induced mild hypothermia as determined by biomarkers and histopathological estimation in a rat subdural hematoma decompression model","volume":"118","author":"Yokobori","journal-title":"J Neurosurg"},{"issue":"22","key":"R30","doi-asserted-by":"publisher","first-page":"2992","DOI":"10.1001\/jama.289.22.2992","article-title":"Prolonged therapeutic hypothermia after traumatic brain injury in adults: a systematic review","volume":"289","author":"McIntyre","journal-title":"JAMA"},{"issue":"23","key":"R31","doi-asserted-by":"publisher","first-page":"2447","DOI":"10.1056\/NEJMoa0706930","article-title":"Hypothermia therapy after traumatic brain injury in children","volume":"358","author":"Hutchison","journal-title":"N Engl J Med"},{"issue":"2","key":"R32","doi-asserted-by":"publisher","first-page":"131","DOI":"10.1016\/S1474-4422(10)70300-8","article-title":"Very early hypothermia induction in patients with severe brain injury (the national acute brain injury study: hypothermia II): a randomised trial","volume":"10","author":"Clifton","journal-title":"Lancet Neurol"},{"issue":"1","key":"R33","first-page":"27","volume":"8","author":"Qiu","journal-title":"Chin J Traumatol"},{"issue":"8","key":"R34","doi-asserted-by":"publisher","first-page":"556","DOI":"10.1056\/NEJM200102223440803","article-title":"Lack of effect of induction of hypothermia after acute brain injury","volume":"344","author":"Clifton","journal-title":"N Engl J Med"},{"issue":"9","key":"R35","doi-asserted-by":"publisher","DOI":"10.1002\/14651858.CD001048.pub5","article-title":"Hypothermia for traumatic brain injury","volume":"9","author":"Lewis","journal-title":"Cochrane Database Syst Rev"},{"issue":"1","key":"R36","doi-asserted-by":"publisher","first-page":"6","DOI":"10.1227\/NEU.0000000000001432","article-title":"Guidelines for the management of severe traumatic brain injury, fourth edition","volume":"80","author":"Carney","journal-title":"Neurosurgery"},{"issue":"7","key":"R37","doi-asserted-by":"publisher","first-page":"629","DOI":"10.1016\/S0025-6196(11)64885-4","article-title":"Occurrence of potentially detrimental temperature alterations in hospitalized patients at risk for brain injury","volume":"73","author":"Albrecht","journal-title":"Mayo Clin Proc"},{"issue":"3","key":"R38","doi-asserted-by":"publisher","first-page":"651","DOI":"10.1007\/s12028-021-01294-1","article-title":"Brain temperature influences intracranial pressure and cerebral perfusion pressure after traumatic brain injury: a CENTER-TBI study","volume":"35","author":"Birg","journal-title":"Neurocrit Care"},{"issue":"1","key":"R39","doi-asserted-by":"publisher","first-page":"82","DOI":"10.1007\/s12028-009-9213-0","article-title":"Induced normothermia attenuates intracranial hypertension and reduces fever burden after severe traumatic brain injury","volume":"11","author":"Puccio","journal-title":"Neurocrit Care"},{"issue":"5","key":"R40","doi-asserted-by":"publisher","first-page":"919","DOI":"10.1007\/s00134-019-05900-x","article-title":"A management algorithm for adult patients with both brain oxygen and intracranial pressure monitoring: the Seattle International Severe Traumatic Brain Injury Consensus Conference (SIBICC)","volume":"46","author":"Chesnut","journal-title":"Intensive Care Med"},{"issue":"1","key":"R41","doi-asserted-by":"publisher","DOI":"10.1186\/s13054-024-04951-x","article-title":"Targeted temperature control following traumatic brain injury: ESICM\/NACCS best practice consensus recommendations","volume":"28","author":"Lavinio","journal-title":"Crit Care"},{"issue":"11","key":"R42","doi-asserted-by":"publisher","first-page":"1004","DOI":"10.1016\/S1474-4422(22)00309-X","article-title":"Traumatic brain injury: progress and challenges in prevention, clinical care, and research","volume":"21","author":"Maas","journal-title":"Lancet Neurol"},{"issue":"6","key":"R43","doi-asserted-by":"publisher","first-page":"1148","DOI":"10.1093\/neuros\/nyab357","article-title":"Evolution and impact of the brain trauma foundation guidelines","volume":"89","author":"Hawryluk","journal-title":"Neurosurg"},{"issue":"22","key":"R44","doi-asserted-by":"publisher","first-page":"3183","DOI":"10.1089\/neu.2019.6474","article-title":"Evolution of evidence and guideline recommendations for the medical management of severe traumatic brain injury","volume":"36","author":"Volovici","journal-title":"J Neurotrauma"},{"key":"R45","doi-asserted-by":"publisher","DOI":"10.3389\/fphar.2022.713100","article-title":"A precision medicine agenda in traumatic brain injury","volume":"13","author":"Cruz Navarro","journal-title":"Front Pharmacol"}],"container-title":["Journal of Medical Internet Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.jmir.org\/2025\/1\/e63847","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:30:47Z","timestamp":1785537047000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.jmir.org\/2025\/1\/e63847"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,3]]},"references-count":45,"URL":"https:\/\/doi.org\/10.2196\/63847","relation":{},"ISSN":["1438-8871"],"issn-type":[{"value":"1438-8871","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7,3]]},"article-number":"v27i2e63847"}}