Utility of a Three-dimensional Convolutional Neural Network-based Renal Function Prediction Using Kidney MRI in Chronic Kidney Disease

Abstract

Abstract A three-dimensional convolutional neural network model was developed to predict renal function in patients with chronic kidney disease (CKD) using magnetic resonance imaging (MRI) Dixon-based T1-weighted in-phase (IP)/opposed-phase (OP)/water-only (WO) images. Seventy-three patients with severe renal dysfunction (estimated glomerular filtration rate [eGFR] <30 mL/min/1.73 m2, CKD stage G4–5); 172 with moderate renal dysfunction (30≤ eGFR <60 mL/min/1.73 m2, CKD stage G3a/b); and 76 in the control (eGFR ≥60 mL/min/1.73 m2, CKD stage G1–2) groups participated in this study. The model was applied to the right, left, and both kidneys, as well as for each imaging method (T1-weighted IP/OP/WO images). The best performance was obtained when using bilateral kidneys and IP images, with an accuracy of 0.862 ± 0.036. The overall accuracy was better for bilateral kidney models than that for unilateral kidney models. Our deep-learning approach using kidney MRI could apply to the evaluation of renal function in patients with CKD.