Construction method of renal function progress prediction model of diabetes-related chronic kidney diseases

The invention provides a method for constructing a renal function progress prediction model for diabetes-related chronic kidney diseases, which comprises the following steps of: acquiring sample data of type 2 diabetes and related chronic kidney diseases, and constructing a data set by follow-up visiting whether a kidney composite endpoint event occurs or not; and evaluating importance ranking of prediction outcomes of various baseline variables, combining hierarchical clustering and single-factor Logistic regression to identify prediction factors from multi-dimensional data, and constructing a clinical prediction model by adopting a sequential forward selection strategy. Respectively adding different multi-modal functional MRI parameters on the basis of the clinical prediction model according to a result of single-factor Logistic regression of functional magnetic resonance imaging parameters to an outcome, and constructing a renal function progress prediction model of the diabetes-related chronic kidney dise