Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Adversarial Robust Federated Learning for Secure Mortality Risk Prediction using Multi-Institutional Electronic

Saswati Chatterjee¹, Suneeta Satpathy²

¹Parul University.

Journal of Visualized Experiments : Jove

|June 1, 2026

Summary

A new Adversarially Robust Federated Learning (AR-FL) model predicts patient mortality risk using Electronic Health Records (EHRs) without data sharing. This privacy-preserving approach enhances model robustness and cross-institutional generalization for clinical decision support.

Related Concept Videos

Methods of Documentation VII: EMR

Methods of Documentation VII: EMR

Electronic Medical Records (EMRs) primarily center around electronically documenting patients' health information within a single healthcare organization or practice. They contain essential clinical data related to a patient's medical history, diagnoses, medications, treatment plans, lab results, and other pertinent information relevant to the specific encounter or episode of care. EMRs are designed to streamline documentation and workflow processes within individual healthcare settings,...

You might also read

Related Articles

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

Sort by

Same author

Comparative Evaluation of Ensemble Machine Learning Approaches for Heart Disease Prediction.

Journal of visualized experiments : JoVE·2026

Same author

Respiring cultureware for high-density, scalable, multipurpose cell-based bioproduction.

Research square·2026

Same author

Cloud-based DDoS detection using hybrid feature selection with deep reinforcement learning (DRL).

Scientific reports·2025

Same author

Nuclease-free precise genome editing corrects MECP2 mutations associated with Rett syndrome.

Frontiers in genome editing·2024

Same author

The Role of Recombinant AAV in Precise Genome Editing.

Frontiers in genome editing·2022

Same author

Geospatial multivariate analysis of COVID-19: a global perspective.

GeoJournal·2021

Area of Science:

Health Informatics
Machine Learning
Privacy-Preserving Technologies

Background:

Electronic Health Records (EHRs) are crucial for data-driven clinical decisions.
Centralized training of predictive models is hindered by EHR data sensitivity and privacy regulations.
Need for secure, privacy-preserving methods for cross-institutional healthcare analytics.

Purpose of the Study:

Introduce a secure, reproducible, and scalable Adversarially Robust Federated Learning (AR-FL) model.
Enable privacy-preserving, adversarially resilient predictive modeling across diverse clinical settings.
Develop a standardized pipeline for training and evaluating clinical decision-support systems.

Main Methods:

Utilized a min-max adversarial training approach for enhanced robustness against perturbations.

Related Experiment Videos

Employed a domain-aware attention mechanism to adapt to institutional feature distribution differences.

Implemented privacy-protecting methods for secure aggregation of model updates during federated communication.

Main Results:

The AR-FL model demonstrated superior predictive performance compared to traditional methods.
Achieved significant adversarial robustness, protecting against worst-case data perturbations.
Showcased strong cross-institutional generalization capabilities, adapting to varied clinical settings.

Conclusions:

The AR-FL model offers a secure and effective solution for privacy-preserving, cross-institutional EHR analysis.
Standardized training and evaluation pipelines facilitate the development of reliable clinical decision-support tools.
This approach supports ethical compliance and enhances the utility of EHR data for patient mortality risk prediction.