1,091
Views
0
CrossRef citations to date
0
Altmetric
Critical Care Nephrology and Continuous Kidney Replacement Therapy

Prediction of successful weaning from renal replacement therapy in critically ill patients based on machine learning

ORCID Icon, , , & ORCID Icon
Article: 2319329 | Received 23 Oct 2023, Accepted 10 Feb 2024, Published online: 28 Feb 2024

Figures & data

Figure 1. The research flow chart and schematic diagram of sampling time range and outcome of RRT withdrawal with variables.

Figure 1. The research flow chart and schematic diagram of sampling time range and outcome of RRT withdrawal with variables.

Table 1. Baseline characteristic and variables of weaning from RRT in severe acute kidney injury patients in ICU.

Table 2. Parameters in the multivariable logistic regression model of weaning from RRT in severe AKI patients.

Figure 2. The AUROC curve of validation set and test set in prediction model of weaning from RRT.

Figure 2. The AUROC curve of validation set and test set in prediction model of weaning from RRT.

Figure 3. Visual histogram of variable importance of XGBoost.

Figure 3. Visual histogram of variable importance of XGBoost.

Table 3. Model validation results with machine learning algorithms.

Supplemental material

Supplemental Material

Download JPEG Image (1.3 MB)

Availability of data and materials

The datasets used and/or analyzed during this study are available from the corresponding author on reasonable request.