{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T18:10:56Z","timestamp":1784139056211,"version":"3.55.0"},"reference-count":40,"publisher":"Public Library of Science (PLoS)","issue":"7","license":[{"start":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T00:00:00Z","timestamp":1784073600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Margaret Hackett Family Program"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Stroke, a leading cause of global mortality and disability, requires accurate prediction of discharge outcomes to support early care planning. We developed an explainable artificial intelligence (AI) framework to predict four discharge categories (home, specialized care, home with help, expired) and identify key predictors.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>This single-center retrospective study included 1,731 patients with ischemic stroke, hemorrhagic stroke, or transient ischemic attack (TIA). Twenty routinely available electronic health record variables were used. Ten classifiers were compared using stratified 5-fold cross-validation, and the final model was calibrated with training-set out-of-fold predictions and interpreted using SHapley Additive exPlanations (SHAP).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The multilayer perceptron (MLP) achieved the highest mean cross-validated macro-F1 score and was selected as the best-performing model. On the independent hold-out test set, the MLP achieved an accuracy of 0.646, macro-specificity of 0.873, macro-precision of 0.559, macro-sensitivity of 0.557, and macro-F1 of 0.548. Class-wise area under the receiver operating characteristic curve (AUC) values were 0.901 for home, 0.874 for specialized care, 0.674 for home with help, and 0.889 for expired. SHAP analysis identified admission National Institutes of Health Stroke Scale (NIHSS), length of stay, age, and primary diagnosis as shared predictors across all discharge categories. The SHAP age-threshold analysis identified 72.0 years as a clinically relevant threshold associated with a lower likelihood of home discharge and higher likelihoods of specialized care, home with help, and expired discharge status. The model also highlighted clinically actionable or addressable domains, including blood glucose, depression, insurance type, last known well time, anticoagulant use, and treatment-related variables.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec004\">\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>This interpretable AI-based framework identified clinically relevant predictors of stroke discharge disposition within this single-center retrospective dataset. These findings may inform future decision-support development; however, clinical implementation, resource optimization, and health-system impact require prospective multicenter validation.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pone.0353683","type":"journal-article","created":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T17:36:37Z","timestamp":1784136997000},"page":"e0353683","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["An explainable artificial intelligence framework for clinical decision support in stroke discharge planning"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4924-4239","authenticated-orcid":true,"given":"Seifollah","family":"Gholampour","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arshia","family":"Dehghan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Evelyn B.","family":"Voura","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jesse","family":"Huang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mohammad","family":"Jamshidi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"George W.","family":"Koutsouras","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,7,15]]},"reference":[{"key":"pone.0353683.ref001","article-title":"Heart Disease and Stroke Statistics: A Report of US and Global Data From the American Heart Association","volume":"151","author":"SS Martin","year":"2025","journal-title":"Circulation"},{"issue":"16","key":"pone.0353683.ref002","doi-asserted-by":"crossref","first-page":"9476","DOI":"10.3390\/ijms23169476","article-title":"Molecular Mechanisms and Pathophysiology of Acute Stroke: Emphasis on Biomarkers in the Different Stroke Subtypes","volume":"23","author":"T Gasull","year":"2022","journal-title":"Int J Mol Sci"},{"issue":"3","key":"pone.0353683.ref003","doi-asserted-by":"crossref","first-page":"1703","DOI":"10.1007\/s10143-020-01367-3","article-title":"Effect of bifurcation in the hemodynamic changes and rupture risk of small intracranial aneurysm","volume":"44","author":"S Gholampour","year":"2021","journal-title":"Neurosurg Rev"},{"issue":"6","key":"pone.0353683.ref004","doi-asserted-by":"crossref","DOI":"10.1161\/STR.0000000000000098","article-title":"Guidelines for Adult Stroke Rehabilitation and Recovery: A Guideline for Healthcare Professionals From the American Heart Association\/American Stroke Association","volume":"47","author":"CJ Winstein","year":"2016","journal-title":"Stroke"},{"issue":"1","key":"pone.0353683.ref005","doi-asserted-by":"crossref","first-page":"449","DOI":"10.1186\/s12913-024-10820-8","article-title":"Transitioning to home and beyond following stroke: a prospective cohort study of outcomes and needs","volume":"24","author":"G O\u2019Callaghan","year":"2024","journal-title":"BMC Health Serv Res"},{"issue":"8","key":"pone.0353683.ref006","doi-asserted-by":"crossref","first-page":"917","DOI":"10.1177\/17474930211016963","article-title":"Longer term patient management following stroke: A systematic review","volume":"16","author":"C Boehme","year":"2021","journal-title":"Int J Stroke"},{"issue":"12","key":"pone.0353683.ref007","doi-asserted-by":"crossref","first-page":"2608","DOI":"10.1080\/09638288.2020.1852616","article-title":"Out of sight, out of mind: long-term outcomes for people discharged home, to inpatient rehabilitation and to residential aged care after stroke","volume":"44","author":"EA Lynch","year":"2022","journal-title":"Disabil Rehabil"},{"issue":"11","key":"pone.0353683.ref008","doi-asserted-by":"crossref","first-page":"107984","DOI":"10.1016\/j.jstrokecerebrovasdis.2024.107984","article-title":"Characteristics associated with 30-day post-stroke readmission within an academic urban hospital network","volume":"33","author":"KM Spiegler","year":"2024","journal-title":"J Stroke Cerebrovasc Dis"},{"issue":"1","key":"pone.0353683.ref009","doi-asserted-by":"crossref","first-page":"287","DOI":"10.1007\/s10143-025-03438-9","article-title":"Readmission patterns and 6-month outcomes in patients with spontaneous intracerebral hematoma: a single centre retrospective analysis in Ghana","volume":"48","author":"PO Tenkorang","year":"2025","journal-title":"Neurosurg Rev"},{"key":"pone.0353683.ref010","doi-asserted-by":"crossref","first-page":"106871","DOI":"10.1016\/j.clineuro.2021.106871","article-title":"Admission glycemic gap in the assessment of patients with intracerebral hemorrhage","volume":"208","author":"AY Dorn","year":"2021","journal-title":"Clin Neurol Neurosurg"},{"issue":"1","key":"pone.0353683.ref011","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1161\/STROKEAHA.107.493262","article-title":"Sex differences in acute stroke care in a statewide stroke registry","volume":"39","author":"JW Gargano","year":"2008","journal-title":"Stroke"},{"key":"pone.0353683.ref012","first-page":"73","article-title":"Moving toward explainable decisions of artificial intelligence models for the prediction of functional outcomes of ischemic stroke patients","volume-title":"Exon Publications","author":"E Zihni","year":"2022"},{"key":"pone.0353683.ref013","doi-asserted-by":"crossref","first-page":"975380","DOI":"10.3389\/fepid.2022.975380","article-title":"Modeling non-linear relationships in epidemiological data: The application and interpretation of spline models","volume":"2","author":"NA Schuster","year":"2022","journal-title":"Front Epidemiol"},{"issue":"7","key":"pone.0353683.ref014","doi-asserted-by":"crossref","first-page":"1159","DOI":"10.1002\/sim.6782","article-title":"Review and evaluation of penalised regression methods for risk prediction in low-dimensional data with few events","volume":"35","author":"M Pavlou","year":"2016","journal-title":"Stat Med"},{"issue":"5","key":"pone.0353683.ref015","doi-asserted-by":"crossref","DOI":"10.1002\/bimj.202300245","article-title":"Penalized Regression Methods With Modified Cross-Validation and Bootstrap Tuning Produce Better Prediction Models","volume":"66","author":"M Pavlou","year":"2024","journal-title":"Biom J"},{"issue":"4","key":"pone.0353683.ref016","first-page":"183","article-title":"Acute Intravenous Tissue Plasminogen Activator Therapy does not Impact Community Discharge after Inpatient Rehabilitation","volume":"2","author":"NL Ifejika","year":"2015","journal-title":"Int J Neurorehabil"},{"issue":"1","key":"pone.0353683.ref017","doi-asserted-by":"crossref","first-page":"8","DOI":"10.1136\/svn-2015-000002","article-title":"Use of intravenous tissue plasminogen activator and hospital costs for patients with acute ischaemic stroke aged 18-64\u2005years in the USA","volume":"1","author":"H Joo","year":"2016","journal-title":"Stroke Vasc Neurol"},{"issue":"2","key":"pone.0353683.ref018","doi-asserted-by":"crossref","DOI":"10.5812\/aapm-116358","article-title":"Evaluating the Efficacy of the National Early Warning Score in Predicting the Mortality of Stroke Patients Admitted to Intensive Care Units","volume":"12","author":"A Mohammadian Erdi","year":"2022","journal-title":"Anesth Pain Med"},{"issue":"3","key":"pone.0353683.ref019","doi-asserted-by":"crossref","first-page":"202","DOI":"10.1111\/j.1747-4949.2011.00696.x","article-title":"Too good to treat? Outcomes in patients not receiving thrombolysis due to mild deficits or rapidly improving symptoms","volume":"7","author":"JZ Willey","year":"2012","journal-title":"Int J Stroke"},{"issue":"1","key":"pone.0353683.ref020","doi-asserted-by":"crossref","first-page":"17","DOI":"10.3988\/jcn.2014.10.1.17","article-title":"Outcomes of endovascular mechanical thrombectomy and intravenous tissue plasminogen activator for the treatment of vertebrobasilar stroke","volume":"10","author":"W Brinjikji","year":"2014","journal-title":"J Clin Neurol"},{"issue":"1","key":"pone.0353683.ref021","doi-asserted-by":"crossref","first-page":"44","DOI":"10.2214\/AJR.18.20260","article-title":"JOURNAL CLUB: Use of Gradient Boosting Machine Learning to Predict Patient Outcome in Acute Ischemic Stroke on the Basis of Imaging, Demographic, and Clinical Information","volume":"212","author":"Y Xie","year":"2019","journal-title":"AJR Am J Roentgenol"},{"issue":"4","key":"pone.0353683.ref022","doi-asserted-by":"crossref","first-page":"1053","DOI":"10.1177\/23969873241253366","article-title":"Artificial intelligence to predict individualized outcome of acute ischemic stroke patients: The SIBILLA project","volume":"9","author":"P Caliandro","year":"2024","journal-title":"Eur Stroke J"},{"issue":"5","key":"pone.0353683.ref023","article-title":"Optimizing acute stroke outcome prediction models: Comparison of generalized regression neural networks and logistic regressions","volume":"17","author":"S Qu","year":"2022","journal-title":"PLoS One"},{"issue":"1","key":"pone.0353683.ref024","doi-asserted-by":"crossref","first-page":"10071","DOI":"10.1038\/s41598-021-89434-7","article-title":"Random forest-based prediction of stroke outcome","volume":"11","author":"C Fernandez-Lozano","year":"2021","journal-title":"Sci Rep"},{"issue":"8","key":"pone.0353683.ref025","doi-asserted-by":"crossref","first-page":"3145","DOI":"10.1111\/jocn.16999","article-title":"Machine learning decision support model for discharge planning in stroke patients","volume":"33","author":"Y Cui","year":"2024","journal-title":"J Clin Nurs"},{"issue":"6","key":"pone.0353683.ref026","doi-asserted-by":"crossref","first-page":"1386","DOI":"10.1227\/neu.0000000000003248","article-title":"Comprehensive Morphometric Analysis to Identify Key Neuroimaging Biomarkers for the Diagnosis of Adult Hydrocephalus Using Artificial Intelligence","volume":"96","author":"S Gholampour","year":"2025","journal-title":"Neurosurgery"},{"issue":"6","key":"pone.0353683.ref027","doi-asserted-by":"crossref","first-page":"1569","DOI":"10.3174\/ajnr.A9166","article-title":"Preoperative Neuroimaging Markers, Clinical Severity Measures, and Shunt Characteristics for Predicting Shunt Revision in Idiopathic Intracranial Hypertension: An Explainable Machine Learning Study","volume":"47","author":"S Gholampour","year":"2026","journal-title":"AJNR Am J Neuroradiol"},{"issue":"2","key":"pone.0353683.ref028","doi-asserted-by":"crossref","first-page":"827","DOI":"10.3390\/make6020039","article-title":"Impact of Nature of Medical Data on Machine and Deep Learning for Imbalanced Datasets: Clinical Validity of SMOTE Is Questionable","volume":"6","author":"S Gholampour","year":"2024","journal-title":"MAKE"},{"issue":"1","key":"pone.0353683.ref029","doi-asserted-by":"crossref","first-page":"134","DOI":"10.1161\/01.STR.0000048217.44714.02","article-title":"Utility of the NIH Stroke Scale as a predictor of hospital disposition","volume":"34","author":"D Schlegel","year":"2003","journal-title":"Stroke"},{"issue":"5","key":"pone.0353683.ref030","doi-asserted-by":"crossref","first-page":"307","DOI":"10.1177\/15459683231166935","article-title":"Predictors of Discharge Destination After Stroke","volume":"37","author":"T Schrage","year":"2023","journal-title":"Neurorehabil Neural Repair"},{"issue":"1","key":"pone.0353683.ref031","first-page":"11","article-title":"Predictive Clinical Factors of In-Hospital Mortality in Women Aged 85 Years or More with Acute Ischemic Stroke","volume":"54","author":"S Torres-Riera","year":"2025","journal-title":"Cerebrovasc Dis"},{"issue":"2","key":"pone.0353683.ref032","doi-asserted-by":"crossref","first-page":"198","DOI":"10.1177\/1747493015609776","article-title":"Quality adjusted life year gains associated with administration of recombinant tissue-type plasminogen activator for treatment of acute ischemic stroke: 1998-2011","volume":"11","author":"DP Lubeck","year":"2016","journal-title":"Int J Stroke"},{"issue":"1","key":"pone.0353683.ref033","doi-asserted-by":"crossref","first-page":"107489","DOI":"10.1016\/j.jstrokecerebrovasdis.2023.107489","article-title":"Machine learning prediction of hospital discharge disposition for inpatients with acute ischemic stroke following mechanical thrombectomy in the United States","volume":"33","author":"MS Jabal","year":"2024","journal-title":"J Stroke Cerebrovasc Dis"},{"issue":"4","key":"pone.0353683.ref034","doi-asserted-by":"crossref","first-page":"368","DOI":"10.1016\/S1474-4422(15)70017-7","article-title":"Alteplase versus tenecteplase for thrombolysis after ischaemic stroke (ATTEST): a phase 2, randomised, open-label, blinded endpoint study","volume":"14","author":"X Huang","year":"2015","journal-title":"Lancet Neurol"},{"issue":"10","key":"pone.0353683.ref035","doi-asserted-by":"crossref","first-page":"2404","DOI":"10.1007\/s00415-018-9009-2","article-title":"Initiating anticoagulant therapy after ICH is associated with patient characteristics and treatment recommendations","volume":"265","author":"JA Sembill","year":"2018","journal-title":"J Neurol"},{"issue":"4","key":"pone.0353683.ref036","doi-asserted-by":"crossref","first-page":"508","DOI":"10.1007\/s11239-014-1144-8","article-title":"Comparison of costs and discharge outcomes for patients hospitalized for ischemic or hemorrhagic stroke with or without atrial fibrillation in the United States","volume":"39","author":"X Pan","year":"2015","journal-title":"J Thromb Thrombolysis"},{"issue":"8","key":"pone.0353683.ref037","article-title":"A Retrospective Analysis of the Underlying Health Status of Patients Treated for Stroke in the Emergency Department of a Community Hospital Situated in a Health Professional Shortage Area","volume":"16","author":"EB Voura","year":"2024","journal-title":"Cureus"},{"issue":"4","key":"pone.0353683.ref038","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0196216","article-title":"FSI simulation of CSF hydrodynamic changes in a large population of non-communicating hydrocephalus patients during treatment process with regard to their clinical symptoms","volume":"13","author":"S Gholampour","year":"2018","journal-title":"PLoS One"},{"issue":"3","key":"pone.0353683.ref039","doi-asserted-by":"crossref","first-page":"386","DOI":"10.1134\/S0021894417030026","article-title":"Numerical simulation of cerebrospinal fluid hydrodynamics in the healing process of hydrocephalus patients","volume":"58","author":"S Gholampour","year":"2017","journal-title":"J Appl Mech Tech Phy"},{"key":"pone.0353683.ref040","doi-asserted-by":"crossref","first-page":"50","DOI":"10.1016\/j.jocn.2016.09.012","article-title":"Evaluating the effect of hydrocephalus cause on the manner of changes in the effective parameters and clinical symptoms of the disease","volume":"35","author":"S Gholampour","year":"2017","journal-title":"J Clin Neurosci"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0353683","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T17:36:44Z","timestamp":1784137004000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0353683"}},"subtitle":[],"editor":[{"given":"Jiawen","family":"Deng","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,7,15]]},"references-count":40,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2026,7,15]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0353683","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,15]]}}}