Analysis of sepsis heterogeneity and subtype identification: a multidimensional clinical feature clustering study based on machine learning

The custom R code used for data preprocessing, k-means clustering, cluster validation, survival analysis, and all sensitivity analyses (including the incremental-value assessment of cluster labels beyond SOFA score and age using the DeLong test, NRI/IDI, SOFA-quartile comparison, and mice-based multiple imputation) to facilitate the analysis of sepsis heterogeneity and subtype identification: a multidimensional clinical feature clustering study based on machine learning