Development and validation of an explainable neural network model for predicting progression in type 2 diabetic kidney disease

1

Type 2 diabetic kidney disease (T2DKD) affects 20-40% of type 2 diabetes patients and is a leading cause of end-stage renal disease globally.
2

A machine learning model was developed using four core predictors: MASLD, AST, DPN, and age, to forecast T2DKD progression.
3

The neural network model demonstrated optimal performance with an AUC of 0.742 and satisfactory calibration in predicting T2DKD progression.
4

SHAP analysis confirmed stable feature importance rankings and revealed synergistic interactions among the identified predictors.
5

This cost-effective neural network model is suitable for clinical practice, aiding early intervention and improving prognosis in T2DKD patients.

1