Research scholarships
  1. Macquarie University
  2. Research
  3. PhD and research degrees
  4. How to apply
  5. Graduate research scholarships
  6. Scholarship search
  7. Privacy-aware medical data analytics with federated learning

Privacy-aware medical data analytics with federated learning

This project aims to develop privacy-aware medical data analytics frameworks using federated learning to balance data utility and privacy.

This scholarship is suitable for candidates with strong interests in leveraging private medical data of multi-model for training advanced machine learning models in a privacy-preserved manner.

Key details

  • 20257522
  • MRes
  • Applications close on 31 October 2025
  • International
  • Information technologies
  • $39,100 p.a.

About the scholarship

Medical data analytics plays a critical role in modern healthcare by leveraging machine learning techniques to extract meaningful insights. However, the sensitive nature of medical data, such as clinical data and medical images, poses significant privacy concerns, especially when sharing data across institutions. Federated Learning (FL) has emerged as a promising solution that enables collaborative model training without exposing raw data.

Objectives

The primary objectives of this research project are:

  • to design and implement a federated learning framework for medical data analytics that ensures privacy preservation
  • to  evaluate different privacy-enhancing techniques (eg differential privacy, secure multiparty computation) within FL models
  • to conduct empirical analysis on benchmark medical datasets while complying with privacy regulations (eg CMS Medicare Part D Prescriber Data (USA), World Health Organisation (WHO) Guidelines, MIMIC-IV and MIMIC-CXR).
Research plan and methodology
  • develop a multi-modal federated learning framework that enables knowledge extraction from sensitive medical data across multiple hospitals and institutes
  • integrating data from diverse modalities (eg medical imaging, clinical notes, and patient vitals)
  • addressing the challenge of heterogeneous data fusion, ensuring that data privacy and learning accuracy are both prioritised.
Expected outcomes
A scalable federated learning framework that supports multi-modal healthcare data integration while preserving privacy.

Availability

This scholarship is available to eligible candidates to undertake a one-year MRes program.

Components

The scholarship comprises:

  • a tuition fee offset/scholarship
  • a living allowance stipend.

The value of each stipend scholarship is $39,100 per annum (full time, indexed) for one year.