Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning
Objective Acute kidney injury (AKI) poses a lethal risk in intensive care unit (ICU) patients, where early detection is challenging. This study was to establish a prediction model for AKI 24 hours in advance for ICU patients and to help clinicians monitor patients at an early stage by key features....
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SAGE Publishing
2025-01-01
|
| Series: | Digital Health |
| Online Access: | https://doi.org/10.1177/20552076241311173 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841556520472412160 |
|---|---|
| author | Li Zhang Mingyu Li Chengcheng Wang Chi Zhang Hong Wu |
| author_facet | Li Zhang Mingyu Li Chengcheng Wang Chi Zhang Hong Wu |
| author_sort | Li Zhang |
| collection | DOAJ |
| description | Objective Acute kidney injury (AKI) poses a lethal risk in intensive care unit (ICU) patients, where early detection is challenging. This study was to establish a prediction model for AKI 24 hours in advance for ICU patients and to help clinicians monitor patients at an early stage by key features. Methods In this study, the Medical Information Mart for Intensive Care (MIMIC) databases were used to construct a dataset of critically ill patients. Predictive models were constructed using five machine learning algorithms based on MIMIC-IV data, and the best predictive model was selected by multiple model evaluation metrics. MIMIC-III data were used for external validation. We conducted an interpretability analysis of the model using SHapley Additive exPlanations (SHAP) to clarify key features and decision-making mechanisms. Results A total of 18,186 patient data were included in this study. The analysis combining calibration and decision curves demonstrated that the eXtreme Gradient Boosting (XGBoost) exhibited superior performance among the five algorithms, achieving an area under the receiver operating characteristic curve of 0.88. Interpretability analysis based on the XGBoost model showed diuretic use, mechanical ventilation, vasopressor use, age, and antibiotic use as the most important decision factors of the model. The SHAP summary plot was used to illustrate the effects of the top 19 features attributed to the XGBoost. Conclusions The XGBoost algorithm can predict the occurrence of AKI more accurately. Interpretative analysis of the model reveals the mechanisms of key features, and reflects the individual differences between patients, providing an important clinical reference. |
| format | Article |
| id | doaj-art-aabee957899748e5a92f41c413d27cbd |
| institution | Kabale University |
| issn | 2055-2076 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | SAGE Publishing |
| record_format | Article |
| series | Digital Health |
| spelling | doaj-art-aabee957899748e5a92f41c413d27cbd2025-01-07T08:04:34ZengSAGE PublishingDigital Health2055-20762025-01-011110.1177/20552076241311173Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learningLi Zhang0Mingyu Li1Chengcheng Wang2Chi Zhang3Hong Wu4 School of Medicine and Health Management, , Wuhan, China School of Medicine and Health Management, , Wuhan, China School of Medicine and Health Management, , Wuhan, China Yunnan Provincial Archives, Kunming, China School of Medicine and Health Management, , Wuhan, ChinaObjective Acute kidney injury (AKI) poses a lethal risk in intensive care unit (ICU) patients, where early detection is challenging. This study was to establish a prediction model for AKI 24 hours in advance for ICU patients and to help clinicians monitor patients at an early stage by key features. Methods In this study, the Medical Information Mart for Intensive Care (MIMIC) databases were used to construct a dataset of critically ill patients. Predictive models were constructed using five machine learning algorithms based on MIMIC-IV data, and the best predictive model was selected by multiple model evaluation metrics. MIMIC-III data were used for external validation. We conducted an interpretability analysis of the model using SHapley Additive exPlanations (SHAP) to clarify key features and decision-making mechanisms. Results A total of 18,186 patient data were included in this study. The analysis combining calibration and decision curves demonstrated that the eXtreme Gradient Boosting (XGBoost) exhibited superior performance among the five algorithms, achieving an area under the receiver operating characteristic curve of 0.88. Interpretability analysis based on the XGBoost model showed diuretic use, mechanical ventilation, vasopressor use, age, and antibiotic use as the most important decision factors of the model. The SHAP summary plot was used to illustrate the effects of the top 19 features attributed to the XGBoost. Conclusions The XGBoost algorithm can predict the occurrence of AKI more accurately. Interpretative analysis of the model reveals the mechanisms of key features, and reflects the individual differences between patients, providing an important clinical reference.https://doi.org/10.1177/20552076241311173 |
| spellingShingle | Li Zhang Mingyu Li Chengcheng Wang Chi Zhang Hong Wu Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning Digital Health |
| title | Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| title_full | Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| title_fullStr | Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| title_full_unstemmed | Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| title_short | Prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| title_sort | prediction of acute kidney injury in intensive care unit patients based on interpretable machine learning |
| url | https://doi.org/10.1177/20552076241311173 |
| work_keys_str_mv | AT lizhang predictionofacutekidneyinjuryinintensivecareunitpatientsbasedoninterpretablemachinelearning AT mingyuli predictionofacutekidneyinjuryinintensivecareunitpatientsbasedoninterpretablemachinelearning AT chengchengwang predictionofacutekidneyinjuryinintensivecareunitpatientsbasedoninterpretablemachinelearning AT chizhang predictionofacutekidneyinjuryinintensivecareunitpatientsbasedoninterpretablemachinelearning AT hongwu predictionofacutekidneyinjuryinintensivecareunitpatientsbasedoninterpretablemachinelearning |