ASN's Mission

To create a world without kidney diseases, the ASN Alliance for Kidney Health elevates care by educating and informing, driving breakthroughs and innovation, and advocating for policies that create transformative changes in kidney medicine throughout the world.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005


The Latest on Twitter

Kidney Week

Abstract: FR-PO755

Prediction of Pathologic Type of Primary Nephrotic Syndrome Using a Machine Learning Algorithm

Session Information

Category: Glomerular

  • 1004 Clinical/Diagnostic Renal Pathology and Lab Medicine


  • Xu, Hui, Nephrology Department, Xiangya Hospital, Central South University, Changsha, Hunan, China, Changsha, China

Renal biopsy is the gold standard to determine the pathologic type of primary nephrotic syndrome. However, renal biopsy cannot be performed in some cases. Based on this, we tried to using a machine learning algorithm to predict the pathologic type in primary nephrotic syndrome patients without renal biopsy.


Clinical data and pathologic types of 203 patients with primary nephrotic syndrome were collected. We trained and validated a machine learning algorithm using data from 203 patients. Then the model was tested prospectively on another 63 biopsy-confirmed patients with primary nephrotic syndrome. Thirdly, Compared with the pathologic results of renal biopsy, the predictive effectiveness of the model was further verified.


Overall accuracy of prediction from the retrospective set of 203 patients was 62.2% across all types of nephrotic syndrome. Among them, minimal-change disease (MCD), 52.2%; non-IgA mesangial proliferative glomerulonephritis, 64.1% ( Non-IgAN); IgA nephropathy (IgAN), 57.1%; membranous nephropathy (MN), 76.1% and focal segmental glomerulosclerosis (FSGS), 10.5%.The accuracy of model prediction for the prospectively collected dataset of 63 patients was 61.9%. MCD, 85.7%; Non-IgAN, 60%; IgAN, 80%; MN, 63.6% and FSGS, 0%. Meanwhile, the algorithm identified 17 of 33 variables as contributing strongly to type of renal pathology. Among them, age, hemoglobin and serum creatinine were the top three(Figure 1).


The method has high precision and can be used to help those patients who are not suitable for renal biopsy to predict the pathologic type of primary nephrotic syndrome, which can guide the diagnosis, choice of treatment and evaluation of prognosis of primary nephrotic syndrome.


  • Government Support - Non-U.S.