ATC Abstracts

American Transplant Congress abstracts

  • Home
  • Meetings Archive
    • 2020 American Transplant Congress
    • 2019 American Transplant Congress
    • 2018 American Transplant Congress
    • 2017 American Transplant Congress
    • 2016 American Transplant Congress
    • 2015 American Transplant Congress
    • 2013 American Transplant Congress
  • Keyword Index
  • Resources
    • 2016 Resources
      • 2016 Welcome Letter
      • ATC 2016 Program Planning Committees
      • ASTS Council 2015-2016
      • AST Board of Directors 2015-2016
    • 2015 Resources
      • 2015 Welcome Letter
      • ATC 2015 Program Planning Committees
      • ASTS Council 2014-2015
      • AST Board of Directors 2014-2015
      • 2015 Conference Schedule
  • Advanced Search

Machine Learning to Predict Deceased Donor Kidney Biopsy Results

C. Martinez, A. Placona

United Network for Organ Sharing, Richmond, VA

Meeting: 2020 American Transplant Congress

Abstract number: B-013

Keywords: Area-under-curve (AUC), Biopsy, Kidney, Prediction models

Session Information

Date: Saturday, May 30, 2020

Session Name: Poster Session B: Kidney Deceased Donor Allocation

Session Time: 3:15pm-4:00pm

 Presentation Time: 3:30pm-4:00pm

Location: Virtual

Related Abstracts
  • Machine Learning to Predict Deceased Donor Liver Biopsy Results
  • Using Ensembles of Machine Learning Classifiers to Maximize the Accuracy and Stability of Molecular Biopsy Interpretation

*Purpose: Despite inconclusive evidence that kidney biopsy findings are associated with graft outcomes, biopsies are often cited as a key factor on whether to accept a deceased kidney offer. Motivated by the influential role of biopsy findings and to determine if biopsies are representations of other clinical data, we developed and evaluated a machine learning model to predict biopsy results given only clinical and lab data available from the OPTN at the time of match.

*Methods: We collected OPTN data, at the time of match, for deceased donors since 1999 with a kidney glomerulosclerosis (GS) (N = 63819 kidneys from 32968 donors) assessment. Next, we identified variables that could be associated with a kidney biopsy. Our dataset included mixed continuous/discrete demographic variables (e.g. age and ethnicity), discrete medical history indicators (e.g. history of hypertension), and continuous lab measurements (e.g. serum creatinine levels). After we generated the dataset, we trained an XGBoost machine learning model to predict the following GS categories: 0-5%, 6-10%, 11-15%, 15-20% or >20%. The model was trained on 90% of the data; the remaining 10% were reserved for model assessment via ROC analysis. We calculated ROC curves, Figure 1a, in a one-versus-rest manner (e.g. 20%+ GS versus all other categories).

*Results: We observed markedly different performance across categories. The GS model performs relatively well at rank-ordering on the extreme ends (0-5% and >20%; AUC 0.73 and 0.75) but poorly for intermediate categories (AUCs from 0.54 to 0.59). Among the top features, as determined from XGBoost feature importances, were age and BMI followed by a number of features derived from lab measurements (Figure 1b).

*Conclusions: Distinct differences in performance across GS categories may indicate challenges for automated biopsy prediction models. A key challenge is that the reliability between pathologists may vary depending on the level of GS. Hence, there may be lack of agreement from a single biopsy; however, due to a phenomenon known as restricted variance, one could expect better agreement at the extremes values and increased uncertainty toward the middle as observed in our model. Further investigation is needed to assess confidence of biopsy results stratified by severity and the impact on predictive models derived from those results.

 border=

  • Tweet
  • Email
  • Print

To cite this abstract in AMA style:

Martinez C, Placona A. Machine Learning to Predict Deceased Donor Kidney Biopsy Results [abstract]. Am J Transplant. 2020; 20 (suppl 3). https://atcmeetingabstracts.com/abstract/machine-learning-to-predict-deceased-donor-kidney-biopsy-results/. Accessed January 23, 2021.

« Back to 2020 American Transplant Congress

Most Viewed Abstracts

  • This Week
  • This Month
  • All Time
    • Subtherapeutic Low Tacrolimus Trough Levels (≤3.5 Ng /ml) Are A Risk Factor For Acute Rejection And Creatinine Doubling.
    • Penis Transplantation: First U.S. Experience.
    • Low GFR after Kidney Donation Is Not Chronic Kidney Disease
    • Is There a Difference Between DCD and DBD Kidney Transplantation with Similar KDPI?
    • Live Related Kidney Transplant Experience in Abuja, Nigeria – First Eight Cases Ever.
    • Home
    • Search
    • Penis Transplantation: First U.S. Experience.
    • Is There a Difference Between DCD and DBD Kidney Transplantation with Similar KDPI?
    • Low GFR after Kidney Donation Is Not Chronic Kidney Disease

    Visit Our Partner Sites

    American Transplant Congress (ATC)

    Visit the official site for the American Transplant Congress »

    American Journal of Transplantation

    The official publication for the American Society of Transplantation (AST) and the American Society of Transplant Surgeons (ASTS) »

    American Society of Transplantation (AST)

    An organization of more than 3000 professionals dedicated to advancing the field of transplantation. »

    American Society of Transplant Surgeons (ASTS)

    The society represents approximately 1,800 professionals dedicated to excellence in transplantation surgery. »

    Copyright © 2013-2021 by American Society of Transplantation and the American Society of Transplant Surgeons. All rights reserved.

    Privacy Policy

    loading Cancel
    Post was not sent - check your email addresses!
    Email check failed, please try again
    Sorry, your blog cannot share posts by email.
    This site uses cookies: Find out more.