{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,2,17]],"date-time":"2026-02-17T02:24:17Z","timestamp":1771295057348,"version":"3.50.1"},"posted":{"date-parts":[[2023,6,27]]},"group-title":"Health Informatics","reference-count":41,"publisher":"openRxiv","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2023,6,27]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>The challenge of irregular temporal data, which is particularly prominent for medication use in the critically ill, limits the performance of predictive models. The purpose of this evaluation was to pilot test integrating synthetic data within an existing dataset of complex medication data to improve machine learning model prediction of fluid overload.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Materials and Methods<\/jats:title>\n                  <jats:p>\n                    This retrospective cohort study evaluated patients admitted to an ICU\n                    <jats:italic>\u2265<\/jats:italic>\n                    72 hours. Four machine learning algorithms to predict fluid overload after 48-72 hours of ICU admission were developed using the original dataset. Then, two distinct synthetic data generation methodologies (synthetic minority over-sampling technique (SMOTE) and conditional tabular generative adversarial network (CT-GAN)) were used to create synthetic data. Finally, a stacking ensemble technique designed to train a meta-learner was established. Models underwent training in three scenarios of varying qualities and quantities of datasets.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>Training machine learning algorithms on the combined synthetic and original dataset overall increased the performance of the predictive models compared to training on the original dataset. The highest performing model was the metamodel trained on the combined dataset with 0.83 AUROC while it managed to significantly enhance the sensitivity across different training scenarios.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>The integration of synthetically generated data is the first time such methods have been applied to ICU medication data and offers a promising solution to enhance the performance of machine learning models for fluid overload, which may be translated to other ICU outcomes. A meta-learner was able to make a trade-off between different performance metrics and improve the ability to identify the minority class.<\/jats:p>\n                <\/jats:sec>","DOI":"10.1101\/2023.06.20.23291680","type":"posted-content","created":{"date-parts":[[2023,6,27]],"date-time":"2023-06-27T16:05:12Z","timestamp":1687881912000},"source":"Crossref","is-referenced-by-count":7,"title":["Improving irregular temporal modeling by integrating synthetic data to the electronic medical record using conditional GANs: a case study of fluid overload prediction in the intensive care unit"],"prefix":"10.64898","author":[{"given":"Alireza","family":"Rafiei","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Milad Ghiasi","family":"Rad","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Andrea","family":"Sikora","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8366-4811","authenticated-orcid":false,"given":"Rishikesan","family":"Kamaleswaran","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2023063002450901000_2023.06.20.23291680v1.1","doi-asserted-by":"publisher","DOI":"10.1177\/1060028020959042"},{"issue":"Supplement 2","key":"2023063002450901000_2023.06.20.23291680v1.2","first-page":"S34","article-title":"Development and validation of a medication regimen complexity scoring tool for critically ill patients","volume":"76","year":"2019","journal-title":"American Journal of Health-System Pharmacy"},{"issue":"1","key":"2023063002450901000_2023.06.20.23291680v1.3","doi-asserted-by":"crossref","first-page":"47","DOI":"10.1002\/jac5.1146","article-title":"Medication regimen complexity is associated with pharmacist interventions and drug-drug interactions: A use of the novel MRC-ICU scoring tool","volume":"3","year":"2020","journal-title":"Journal of the American College of Clinical Pharmacy"},{"issue":"4","key":"2023063002450901000_2023.06.20.23291680v1.4","doi-asserted-by":"crossref","first-page":"573","DOI":"10.1177\/0897190021999792","article-title":"Medication regimen complexity score as an indicator of fluid balance in critically ill patients","volume":"35","year":"2022","journal-title":"Journal of Pharmacy Practice"},{"issue":"9","key":"2023063002450901000_2023.06.20.23291680v1.5","doi-asserted-by":"crossref","first-page":"1318","DOI":"10.1097\/CCM.0000000000005585","article-title":"Impact of pharmacists to improve patient care in the critically ill: a large multicenter analysis using meaningful metrics with the Medication Regimen Complexity-ICU (MRC-ICU) score","volume":"50","year":"2022","journal-title":"Critical care medicine"},{"issue":"1","key":"2023063002450901000_2023.06.20.23291680v1.6","first-page":"1","article-title":"Pharmacophenotype identification of intensive care unit medications using unsupervised cluster analysis of the ICURx common data model","volume":"27","year":"2023","journal-title":"Critical Care"},{"issue":"22","key":"2023063002450901000_2023.06.20.23291680v1.7","doi-asserted-by":"crossref","first-page":"2077","DOI":"10.1093\/ajhp\/zxab237","article-title":"Optimization of critical care pharmacy clinical services: A gap analysis approach","volume":"78","year":"2021","journal-title":"Am J Health Syst Pharm"},{"key":"2023063002450901000_2023.06.20.23291680v1.8","doi-asserted-by":"crossref","unstructured":"Sikora A , Jeong H , Yu M , Chen X , Murray B , Kamaleswaran R. Cluster analysis driven by unsupervised latentfeature learning of intensive care unit medications to identify novel pharmaco-phenotypes of critically ill patients. 2022","DOI":"10.21203\/rs.3.rs-1745568\/v1"},{"key":"2023063002450901000_2023.06.20.23291680v1.9","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMra1814259"},{"issue":"2","key":"2023063002450901000_2023.06.20.23291680v1.10","doi-asserted-by":"crossref","first-page":"444","DOI":"10.1109\/JPROC.2015.2501978","article-title":"Machine learning and decision support in critical care","volume":"104","year":"2016","journal-title":"Proceedings of the IEEE"},{"key":"2023063002450901000_2023.06.20.23291680v1.11","doi-asserted-by":"publisher","DOI":"10.1186\/S40537-019-0217-0\/FIGURES\/6"},{"key":"2023063002450901000_2023.06.20.23291680v1.12","doi-asserted-by":"crossref","unstructured":"Winter JS , Davidson E. Governance of artificial intelligence and personal health information. Digital policy, regulation and governance 2019","DOI":"10.1108\/DPRG-08-2018-0048"},{"key":"2023063002450901000_2023.06.20.23291680v1.13","doi-asserted-by":"crossref","unstructured":"Hernandez M , Epelde G , Alberdi A , Cilla R , Rankin D. Synthetic data generation for tabular health records: A systematic review. Neurocomputing 2022","DOI":"10.1016\/j.neucom.2022.04.053"},{"key":"2023063002450901000_2023.06.20.23291680v1.14","doi-asserted-by":"crossref","first-page":"134466","DOI":"10.1109\/ACCESS.2022.3230324","article-title":"Improving Sepsis Prediction Performance Using Conditional Recurrent Adversarial Net-works","volume":"10","year":"2022","journal-title":"IEEE Access"},{"key":"2023063002450901000_2023.06.20.23291680v1.15","article-title":"Synthetic Data in Healthcare","journal-title":"arXiv preprint"},{"issue":"6","key":"2023063002450901000_2023.06.20.23291680v1.16","doi-asserted-by":"crossref","first-page":"493","DOI":"10.1038\/s41551-021-00751-8","article-title":"Synthetic data in machine learning for medicine and healthcare","volume":"5","year":"2021","journal-title":"Nature Biomedical Engineering"},{"key":"2023063002450901000_2023.06.20.23291680v1.17","doi-asserted-by":"crossref","first-page":"100546","DOI":"10.1016\/j.cosrev.2023.100546","article-title":"Synthetic data generation: State of the art in health care domain","volume":"48","year":"2023","journal-title":"Computer Science Review"},{"issue":"1","key":"2023063002450901000_2023.06.20.23291680v1.18","doi-asserted-by":"crossref","first-page":"e0000082","DOI":"10.1371\/journal.pdig.0000082","article-title":"Synthetic data in health care: A narrative review","volume":"2","year":"2023","journal-title":"PLOS Digital Health"},{"key":"2023063002450901000_2023.06.20.23291680v1.19","unstructured":"Conditional synthetic data generation for robust machine learning applications with limited pandemic data. Proceedings of the AAAI Conference on Artificial Intelligence; 2022."},{"issue":"5","key":"2023063002450901000_2023.06.20.23291680v1.20","doi-asserted-by":"crossref","first-page":"769","DOI":"10.1177\/08971900211008261","article-title":"Fluid stewardship of maintenance intravenous fluids","volume":"35","year":"2022","journal-title":"Journal of Pharmacy Practice"},{"key":"2023063002450901000_2023.06.20.23291680v1.21","doi-asserted-by":"publisher","DOI":"10.1186\/s13054-020-02926-2"},{"key":"2023063002450901000_2023.06.20.23291680v1.22","first-page":"11795468221095875","article-title":"Early diuretics for deresuscitation in septic patients with left ventricular dysfunction","volume":"16","year":"2022","journal-title":"Clinical Medicine Insights: Cardiology"},{"issue":"12","key":"2023063002450901000_2023.06.20.23291680v1.23","doi-asserted-by":"crossref","first-page":"984","DOI":"10.1093\/ajhp\/zxab453","article-title":"From theory to bedside: Implementation of fluid stewardship in a medical ICU pharmacy practice","volume":"79","year":"2022","journal-title":"American Journal of Health-System Pharmacy"},{"key":"2023063002450901000_2023.06.20.23291680v1.24","doi-asserted-by":"crossref","unstructured":"Qin X , Zhang W , Hu X , Zhou W. A deep learning model to identify fluid overload status in critically ill patients based on chest X-ray images. Polish Archives of Internal Medicine 2023:16396\u201396","DOI":"10.20452\/pamw.16396"},{"key":"2023063002450901000_2023.06.20.23291680v1.25","doi-asserted-by":"publisher","DOI":"10.1101\/2023.06.16.23291493"},{"key":"2023063002450901000_2023.06.20.23291680v1.26","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2013.281053"},{"key":"2023063002450901000_2023.06.20.23291680v1.27","doi-asserted-by":"publisher","DOI":"10.1016\/s0140-6736(07)61602-x"},{"issue":"6","key":"2023063002450901000_2023.06.20.23291680v1.28","doi-asserted-by":"crossref","first-page":"863","DOI":"10.1177\/0897190019853979","article-title":"Fluid stewardship during critical illness: a call to action","volume":"33","year":"2020","journal-title":"Journal of Pharmacy Practice"},{"key":"2023063002450901000_2023.06.20.23291680v1.29","doi-asserted-by":"publisher","DOI":"10.1002\/mpr.329"},{"key":"2023063002450901000_2023.06.20.23291680v1.30","doi-asserted-by":"crossref","first-page":"321","DOI":"10.1613\/jair.953","article-title":"SMOTE: synthetic minority over-sampling technique","volume":"16","year":"2002","journal-title":"Journal of artificial intelligence research"},{"key":"2023063002450901000_2023.06.20.23291680v1.31","unstructured":"Xu L , Skoularidou M , Cuesta-Infante A , Veeramachaneni K. Modeling tabular data using conditional gan. Advances in Neural Information Processing Systems 2019;32"},{"key":"2023063002450901000_2023.06.20.23291680v1.32","doi-asserted-by":"publisher","DOI":"10.3102\/10769986027001077"},{"key":"2023063002450901000_2023.06.20.23291680v1.33","unstructured":"Van der Maaten L , Hinton G. Visualizing data using t-SNE. Journal of machine learning research 2008;9(11)"},{"key":"2023063002450901000_2023.06.20.23291680v1.34","unstructured":"Lundberg SM , Lee S-I. A unified approach to interpreting model predictions. Advances in neural information processing systems 2017;30"},{"key":"2023063002450901000_2023.06.20.23291680v1.35","doi-asserted-by":"crossref","first-page":"156","DOI":"10.1016\/j.jcrc.2020.07.004","article-title":"A narrative review of pharmacologic de-resuscitation in the critically ill","volume":"59","year":"2020","journal-title":"Journal of Critical Care"},{"issue":"2","key":"2023063002450901000_2023.06.20.23291680v1.36","doi-asserted-by":"crossref","first-page":"336","DOI":"10.3390\/jcm11020336","article-title":"Fluid Overload Phenotypes in Critical Illness\u2014A Machine Learning Approach","volume":"11","year":"2022","journal-title":"Journal of clinical medicine"},{"key":"2023063002450901000_2023.06.20.23291680v1.37","doi-asserted-by":"publisher","DOI":"10.1186\/s13054-019-2472-z"},{"issue":"1","key":"2023063002450901000_2023.06.20.23291680v1.38","first-page":"1","article-title":"Principles of fluid management and stewardship in septic shock: it is time to consider the four D\u2019s and the four phases of fluid therapy","volume":"8","year":"2018","journal-title":"Annals of intensive care"},{"issue":"1","key":"2023063002450901000_2023.06.20.23291680v1.39","first-page":"1","article-title":"Fluid overload in the ICU: evaluation and management","volume":"17","year":"2016","journal-title":"BMC nephrology"},{"issue":"4","key":"2023063002450901000_2023.06.20.23291680v1.40","doi-asserted-by":"crossref","first-page":"803","DOI":"10.1016\/j.ccc.2015.06.013","article-title":"Fluid overload","volume":"31","year":"2015","journal-title":"Critical care clinics"},{"key":"2023063002450901000_2023.06.20.23291680v1.41","unstructured":"Synthetic examples improve generalization for rare classes. Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision; 2020."}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2023.06.20.23291680","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T19:35:57Z","timestamp":1768505757000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2023.06.20.23291680"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,27]]},"references-count":41,"URL":"https:\/\/doi.org\/10.1101\/2023.06.20.23291680","relation":{"is-preprint-of":[{"id-type":"doi","id":"10.1016\/j.compbiomed.2023.107749","asserted-by":"subject"}]},"subject":[],"published":{"date-parts":[[2023,6,27]]},"subtype":"preprint"}}