TrustGuard (AI in Healthcare)
AI applications in healthcare, federated learning, explainable AI and medical imaging.
Project overview
TrustGuard is the flagship initiative that brings the centre's healthcare and trustworthy AI research together under one programme. It is co-led by Dr. Saw Shier Nee and Dr. Erma Rahayu Mohd Faizal Abdullah, whose research covers medical image and signal analysis, clinical risk assessment, computer vision and ethical AI. The initiative's current focus spans AI applications in healthcare, federated learning, explainable AI and medical imaging. Those strands are closely connected. Federated learning addresses how clinical models can be trained across institutions without moving sensitive patient records; explainable AI addresses whether a clinician can understand and question what a model produced. Medical imaging provides the concrete setting in which both questions are tested, drawing on the segmentation, detection and diagnostic imaging work already published by Malaya AIR members.
Research focus
Current focus
Research areas
Leadership
Why this work matters
Clinical adoption of machine learning depends on more than predictive accuracy. Models often have to be trained on data that cannot easily leave the institution holding it, and their outputs have to be interpretable by the clinicians who act on them. TrustGuard treats those two constraints as research questions in their own right rather than as deployment details to be settled later.
Research connections
Publications by the researchers involved, or in the research areas this project spans.
- Accepted paper MICCAI 2025
- Reconstructed U-Net for Gastric-Cancer Segmentation IEEE Transactions on Fuzzy Systems 2024/2025
- Deep Learning for Dental CBCT Detection BMC Oral Health 2024
- Radiology Report Generation Cognitive Computation 2025
- Explainable Goal-driven Agents & Robots ACM Computing Surveys (CSUR) 2023
Collaboration
Interested in research collaboration, technical exchange or translational opportunities related to this work?
Discuss a Collaboration