Acute Kidney Injury Prognosis Prediction Using Machine Learning Methods: A Systematic Review.

Although machine learning showed a trend toward higher model discrimination, the difference was not statistically significant (AUROC, 0.781; 95% CI, 0.674-0.861 vs. 0.734; 95% CI, 0.547-0.863) for different models, which might be because of the limited number of studies that predicted kidney functional recovery ( Fig S8 ).

also in 53,322 other papers