Development and Validation of a Federated Learning Model for Antimicrobial Resistance Prediction Using Multi-Source Global AMR Data Deployed Across an 11-Branch Ugandan Hospital Network
Abstract
Topic: Development and Validation of a Federated Learning Model for Antimicrobial Resistance Prediction Using Multi-Source Global AMR Data Deployed Across an 11-Branch Ugandan Hospital Network Background Antimicrobial resistance (AMR) is a growing global health crisis disproportionately affecting low- and middle-income countries where diagnostic capacity is limited and empiric prescribing is common. In Uganda, AMR threatens management of malaria co-infections, recurrent UTIs, and pneumonia all dependent on effective antibiotic therapy. Large AMR datasets exist through repositories such as Vivli, yet local networks lack tools to translate this into actionable clinical intelligence. C-Care IMC, an 11-branch hospital network in Kampala, operates an Agentic Federated Learning platform across 1.29 million inpatient records without patient data leaving any facility. We propose to treat multiple independent Vivli AMR datasets as separate federated nodes, each contributing encrypted model updates without sharing raw data, to train a globally-informed resistance prediction model deployed into C-Care’s platform as a clinical decision support tool, establishing a reproducible, privacy-preserving AMR prediction framework for East Africa. Research Objectives 1. To train a federated AMR prediction model using multiple independent Vivli datasets as federation nodes, each contributing encrypted gradient updates without raw data leaving its source. 2. To validate the federated model against…
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.