Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan
Abstract Background Machine learning (ML) model is increasingly used to predict short-term outcome in critically ill patients, but the study for long-term outcome is sparse. We used explainable ML approach to establish 30-day, 90-day and 1-year mortality prediction model in critically ill ventilated...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2022-03-01
|
Series: | BMC Medical Informatics and Decision Making |
Subjects: | |
Online Access: | https://doi.org/10.1186/s12911-022-01817-6 |
_version_ | 1811335563991580672 |
---|---|
author | Ming-Cheng Chan Kai-Chih Pai Shao-An Su Min-Shian Wang Chieh-Liang Wu Wen-Cheng Chao |
author_facet | Ming-Cheng Chan Kai-Chih Pai Shao-An Su Min-Shian Wang Chieh-Liang Wu Wen-Cheng Chao |
author_sort | Ming-Cheng Chan |
collection | DOAJ |
description | Abstract Background Machine learning (ML) model is increasingly used to predict short-term outcome in critically ill patients, but the study for long-term outcome is sparse. We used explainable ML approach to establish 30-day, 90-day and 1-year mortality prediction model in critically ill ventilated patients. Methods We retrospectively included patients who were admitted to intensive care units during 2015–2018 at a tertiary hospital in central Taiwan and linked with the Taiwanese nationwide death registration data. Three ML models, including extreme gradient boosting (XGBoost), random forest (RF) and logistic regression (LR), were used to establish mortality prediction model. Furthermore, we used feature importance, Shapley Additive exPlanations (SHAP) plot, partial dependence plot (PDP), and local interpretable model-agnostic explanations (LIME) to explain the established model. Results We enrolled 6994 patients and found the accuracy was similar among the three ML models, and the area under the curve value of using XGBoost to predict 30-day, 90-day and 1-year mortality were 0.858, 0.839 and 0.816, respectively. The calibration curve and decision curve analysis further demonstrated accuracy and applicability of models. SHAP summary plot and PDP plot illustrated the discriminative point of APACHE (acute physiology and chronic health exam) II score, haemoglobin and albumin to predict 1-year mortality. The application of LIME and SHAP force plots quantified the probability of 1-year mortality and algorithm of key features at individual patient level. Conclusions We used an explainable ML approach, mainly XGBoost, SHAP and LIME plots to establish an explainable 1-year mortality prediction ML model in critically ill ventilated patients. |
first_indexed | 2024-04-13T17:27:06Z |
format | Article |
id | doaj.art-899814f1a8274e338159c2c4d2a9c1af |
institution | Directory Open Access Journal |
issn | 1472-6947 |
language | English |
last_indexed | 2024-04-13T17:27:06Z |
publishDate | 2022-03-01 |
publisher | BMC |
record_format | Article |
series | BMC Medical Informatics and Decision Making |
spelling | doaj.art-899814f1a8274e338159c2c4d2a9c1af2022-12-22T02:37:46ZengBMCBMC Medical Informatics and Decision Making1472-69472022-03-0122111110.1186/s12911-022-01817-6Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central TaiwanMing-Cheng Chan0Kai-Chih Pai1Shao-An Su2Min-Shian Wang3Chieh-Liang Wu4Wen-Cheng Chao5Division of Critical Care and Respiratory Therapy, Department of Internal Medicine, Taichung Veterans General HospitalCollege of Engineering, Tunghai UniversityArtificial Intelligence Center, Tunghai UniversityArtificial Intelligence Studio, Taichung Veterans General HospitalArtificial Intelligence Studio, Taichung Veterans General HospitalDepartment of Critical Care Medicine, Taichung Veterans General HospitalAbstract Background Machine learning (ML) model is increasingly used to predict short-term outcome in critically ill patients, but the study for long-term outcome is sparse. We used explainable ML approach to establish 30-day, 90-day and 1-year mortality prediction model in critically ill ventilated patients. Methods We retrospectively included patients who were admitted to intensive care units during 2015–2018 at a tertiary hospital in central Taiwan and linked with the Taiwanese nationwide death registration data. Three ML models, including extreme gradient boosting (XGBoost), random forest (RF) and logistic regression (LR), were used to establish mortality prediction model. Furthermore, we used feature importance, Shapley Additive exPlanations (SHAP) plot, partial dependence plot (PDP), and local interpretable model-agnostic explanations (LIME) to explain the established model. Results We enrolled 6994 patients and found the accuracy was similar among the three ML models, and the area under the curve value of using XGBoost to predict 30-day, 90-day and 1-year mortality were 0.858, 0.839 and 0.816, respectively. The calibration curve and decision curve analysis further demonstrated accuracy and applicability of models. SHAP summary plot and PDP plot illustrated the discriminative point of APACHE (acute physiology and chronic health exam) II score, haemoglobin and albumin to predict 1-year mortality. The application of LIME and SHAP force plots quantified the probability of 1-year mortality and algorithm of key features at individual patient level. Conclusions We used an explainable ML approach, mainly XGBoost, SHAP and LIME plots to establish an explainable 1-year mortality prediction ML model in critically ill ventilated patients.https://doi.org/10.1186/s12911-022-01817-6Mortality predictionCritical illnessMechanical ventilationMachine learningInterpretability |
spellingShingle | Ming-Cheng Chan Kai-Chih Pai Shao-An Su Min-Shian Wang Chieh-Liang Wu Wen-Cheng Chao Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan BMC Medical Informatics and Decision Making Mortality prediction Critical illness Mechanical ventilation Machine learning Interpretability |
title | Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan |
title_full | Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan |
title_fullStr | Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan |
title_full_unstemmed | Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan |
title_short | Explainable machine learning to predict long-term mortality in critically ill ventilated patients: a retrospective study in central Taiwan |
title_sort | explainable machine learning to predict long term mortality in critically ill ventilated patients a retrospective study in central taiwan |
topic | Mortality prediction Critical illness Mechanical ventilation Machine learning Interpretability |
url | https://doi.org/10.1186/s12911-022-01817-6 |
work_keys_str_mv | AT mingchengchan explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan AT kaichihpai explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan AT shaoansu explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan AT minshianwang explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan AT chiehliangwu explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan AT wenchengchao explainablemachinelearningtopredictlongtermmortalityincriticallyillventilatedpatientsaretrospectivestudyincentraltaiwan |