{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"volume":["9(2)"],"submitter":["Zou H"],"pubmed_abstract":["<h4>Background and aims</h4>Accurately predicting length of stay (LOS) is considered a challenging task for health care systems globally. In previous studies on LOS range prediction, researchers commonly pre-classified the LOS ranges, which were the same for all patients in the same classification, and then utilized a classifier for prediction. In this study, we innovatively aimed to predict the specific LOS range for each patient (the LOS range was different for each patient).<h4>Methods</h4>In the modified deep neural network (DNN), the overall sample error (root mean square error (RMSE) method), the estimated sample error (ERR<sub>pred</sub> method), the probability distribution with different loss functions (Dis<sub>pred</sub>_Loss1, Dis<sub>pred</sub>_Loss2, and Dis<sub>pred</sub>_Loss3 method), and the generative adversarial networks (WGAN-GP for LOS method) are used for LOS range prediction. The Medical Information Mart for Intensive Care III (MIMIC-III) database is used to validate these methods.<h4>Results</h4>The RMSE method is convenient for LOS range prediction, but the predicted ranges are all consistent in the same batch of samples. The ERR<sub>pred</sub> method can achieve better prediction results in samples with low errors. However, the prediction effect is worse in samples with larger errors. The Dis<sub>pred</sub>_Loss1 method encounters a training instability problem. The Dis<sub>pred</sub>_Loss2 and Dis<sub>pred</sub>_Loss3 methods perform well in making predictions. Although WGAN-GP for LOS method does not show a substantial advantage over other methods, this method might have the potential to improve the predictive performance.<h4>Conclusion</h4>The results show that it is possible to achieve an acceptable accurate LOS range prediction through a reasonable model design, which may help physicians in the clinic."],"journal":["Heliyon"],"pagination":["e13573"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC9958433"],"repository":["biostudies-literature"],"pubmed_title":["Predicting length of stay ranges by using novel deep neural networks."],"pmcid":["PMC9958433"],"pubmed_authors":["Zou H","Wang M","Yang W","Tang L","Liang H","Zhu Q","Wu H"],"additional_accession":[]},"is_claimable":false,"name":"Predicting length of stay ranges by using novel deep neural networks.","description":"<h4>Background and aims</h4>Accurately predicting length of stay (LOS) is considered a challenging task for health care systems globally. In previous studies on LOS range prediction, researchers commonly pre-classified the LOS ranges, which were the same for all patients in the same classification, and then utilized a classifier for prediction. In this study, we innovatively aimed to predict the specific LOS range for each patient (the LOS range was different for each patient).<h4>Methods</h4>In the modified deep neural network (DNN), the overall sample error (root mean square error (RMSE) method), the estimated sample error (ERR<sub>pred</sub> method), the probability distribution with different loss functions (Dis<sub>pred</sub>_Loss1, Dis<sub>pred</sub>_Loss2, and Dis<sub>pred</sub>_Loss3 method), and the generative adversarial networks (WGAN-GP for LOS method) are used for LOS range prediction. The Medical Information Mart for Intensive Care III (MIMIC-III) database is used to validate these methods.<h4>Results</h4>The RMSE method is convenient for LOS range prediction, but the predicted ranges are all consistent in the same batch of samples. The ERR<sub>pred</sub> method can achieve better prediction results in samples with low errors. However, the prediction effect is worse in samples with larger errors. The Dis<sub>pred</sub>_Loss1 method encounters a training instability problem. The Dis<sub>pred</sub>_Loss2 and Dis<sub>pred</sub>_Loss3 methods perform well in making predictions. Although WGAN-GP for LOS method does not show a substantial advantage over other methods, this method might have the potential to improve the predictive performance.<h4>Conclusion</h4>The results show that it is possible to achieve an acceptable accurate LOS range prediction through a reasonable model design, which may help physicians in the clinic.","dates":{"release":"2023-01-01T00:00:00Z","publication":"2023 Feb","modification":"2025-04-22T09:45:51.959Z","creation":"2025-04-05T23:08:58.061Z"},"accession":"S-EPMC9958433","cross_references":{"pubmed":["36852025"],"doi":["10.1016/j.heliyon.2023.e13573"]}}