Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Chronic Kidney Disease I: Introduction01:25

Chronic Kidney Disease I: Introduction

928
Chronic Kidney Disease (CKD) arises when the kidneys progressively lose their ability to function, ultimately leading to end-stage renal disease. At this advanced stage, the kidneys can no longer filter waste or maintain essential body functions, requiring renal replacement therapy (RRT) through dialysis or a kidney transplant for survival.Early-stage chronic kidney disease and detection challengesIn CKD's early stages, symptoms often remain absent because healthy nephrons compensate for...
928
Drug Dosing in Renal Diseases: Estimation of Glomerular Filtration Rate Based on Serum Creatinine Concentration01:28

Drug Dosing in Renal Diseases: Estimation of Glomerular Filtration Rate Based on Serum Creatinine Concentration

285
Glomerular filtration rate (GFR) can be estimated from serum creatinine using the modification of diet in renal disease (MDRD) formula or the chronic kidney disease–epidemiology collaboration (CKD–EPI) equation. Both methods are widely used in clinical practice to assess kidney function and guide treatment decisions.The MDRD equation does not require weight or height measurements and is normalized to the body surface area of 1.73 m², considered the average adult surface area.
285

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Intragraft donor-specific antibodies reflect histologic heterogeneity in kidney allografts with concurrent serum DSA.

Frontiers in immunology·2026
Same author

Interfacial stability of organic-based hole transport layers in perovskite photovoltaics for space-like thermal environments.

Communications engineering·2026
Same author

Ethnicity-specific molecular subtypes and a machine-learning risk model in Asian patients with non-muscle-invasive bladder cancer.

Scientific reports·2026
Same author

Pathomic analysis of 5-year surveillance biopsies as predictors of kidney allograft loss.

American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons·2026
Same author

Reinforcement Learning for Intraoperative Hypotension Management with Consideration to Postoperative Acute Kidney Injury.

Kidney360·2026
Same author

MySurgeryRisk Model Predictions of Postoperative Complications and Mortality.

JAMA surgery·2026

Related Experiment Video

Updated: Mar 10, 2026

Comparative Proteomic Analysis of Whole Kidney, Medulla, and Cortical Tubules in Diabetic Pathogenesis of Kidney Injury in Mice
10:31

Comparative Proteomic Analysis of Whole Kidney, Medulla, and Cortical Tubules in Diabetic Pathogenesis of Kidney Injury in Mice

Published on: May 2, 2025

792

Explainable Feature Embeddings from Histopathology Foundation Models: A Case Study for End Stage Kidney Disease Risk

Harishwar Reddy Kasireddy1,2, Nicholas Lucarelli3, Donghwan Yun4

  • 1Dept. of Electrical and Computer Engineering, Univ. of Florida, Gainesville, FL.

Proceedings of Spie--The International Society for Optical Engineering
|March 9, 2026
PubMed
Summary

Foundational models in pathology image analysis are enhanced for explainability by correlating handcrafted features with deep learning embeddings. This approach improves prediction of end-stage kidney disease from kidney biopsies.

Keywords:
Diabetic nephropathyclassificationdigital pathologyend stage kidney diseaseexplainabilityfeature embeddingsfoundation models

More Related Videos

Supervised Machine Learning for Semi-Quantification of Extracellular DNA in Glomerulonephritis
09:16

Supervised Machine Learning for Semi-Quantification of Extracellular DNA in Glomerulonephritis

Published on: June 18, 2020

7.4K
Assessment of Kidney Function in Mouse Models of Glomerular Disease
09:16

Assessment of Kidney Function in Mouse Models of Glomerular Disease

Published on: June 30, 2018

18.7K

Related Experiment Videos

Last Updated: Mar 10, 2026

Comparative Proteomic Analysis of Whole Kidney, Medulla, and Cortical Tubules in Diabetic Pathogenesis of Kidney Injury in Mice
10:31

Comparative Proteomic Analysis of Whole Kidney, Medulla, and Cortical Tubules in Diabetic Pathogenesis of Kidney Injury in Mice

Published on: May 2, 2025

792
Supervised Machine Learning for Semi-Quantification of Extracellular DNA in Glomerulonephritis
09:16

Supervised Machine Learning for Semi-Quantification of Extracellular DNA in Glomerulonephritis

Published on: June 18, 2020

7.4K
Assessment of Kidney Function in Mouse Models of Glomerular Disease
09:16

Assessment of Kidney Function in Mouse Models of Glomerular Disease

Published on: June 30, 2018

18.7K

Area of Science:

  • Computational Pathology
  • Artificial Intelligence in Medicine
  • Digital Pathology

Background:

  • Foundational models (FMs) offer improved performance in pathology image analysis but lack clinical transparency due to their black box nature.
  • Explainability is crucial for integrating FMs into clinical workflows, requiring understanding of features learned by these models for specific tasks.

Purpose of the Study:

  • To develop a computational pipeline that enhances the explainability of Foundational Models (FMs) in pathology image analysis.
  • To correlate handcrafted features (HFs) with feature embeddings (FEs) from FMs to improve transparency and performance.
  • To predict end-stage kidney disease (ESKD) risk using explainable AI methods on kidney biopsy images.

Main Methods:

  • A pipeline was developed correlating HFs from segmented kidney tissue units (arteries, tubules, glomeruli) with FEs from Prov-Gigapath (PG) and UNI FMs.
  • Pearson correlation coefficient was used to identify correspondences between HFs and FEs.
  • Logistic Regression (LR) and k-Nearest Neighbors (kNN) classifiers were trained on combined feature sets for ESKD prediction using 56 diabetic nephropathy kidney biopsy whole slide images (WSIs).

Main Results:

  • The LR model trained on the combined feature set achieved improved accuracy (0.8393), balanced accuracy (0.7938), and Matthew's correlation coefficient (0.5993) compared to models using individual feature sets.
  • Prov-Gigapath (PG) demonstrated superior specificity (1.000) and AUROC (0.8281), while UNI showed higher AUPRC (0.7813).
  • Feature explainability maps were generated, correlating specific learned features with pathological structures.

Conclusions:

  • Correlating handcrafted features with FM-derived embeddings significantly enhances explainability and predictive performance in computational pathology.
  • The proposed methodology facilitates a more transparent integration of advanced AI models into clinical practice for kidney disease diagnosis.
  • This approach holds promise for improving the accuracy and reliability of AI-driven diagnostic tools in pathology.