CSIRO Postdoctoral Fellowship in AI Based Decision Support in Fracture Diagnosis

• Do you have a PhD in computer science, medical image analysis, computer vision, data science, machine learning?
• Conduct impactful research in delivering digital solutions to Australia’s greatest health challenges.
• Join the AEHRC team in the exciting 3-year postdoctoral opportunity!

CSIRO Early Research Career (CERC) Postdoctoral Fellowships provide opportunities to scientists and engineers who have completed their doctorate and have less than three years of relevant postdoctoral work experience. These fellowships aim to develop the next generation of future leaders of the innovation system.

The Australian e-Health Research Centre is excited to welcome a CERC Fellow to join the team in contributing to research in the development of AI technology, augmenting X-ray clinicians for improved outcomes in fracture diagnosis in emergency settings.  

The focus of this CERC Fellowship will be on developing and evaluating an AI-based system for fast and improved assessment of bone fracture and related infections using X-rays. This would initially include analysis images and meta data which the team has already got access to, experiment on existing machine learning models and development of computer vision/machine learning models specific to the context. 

Currently, fracture diagnosis is reliant on X-rays being assessed manually by qualified experts. However, while X-ray equipment is widely available in rural and remote hospitals, clinics and nursing posts, in the face of staff shortages and increasing patient numbers, getting access to a competent X-ray expert is increasingly difficult. In this context, the integration of an AI tool to support clinical decision-making has significant potential to reduce clinician burden and reduce the time to diagnosis and treatment. Previous studies, including our own, have demonstrated the promising utility of AI assistance in fracture detection on X-rays. However, there remains a gap in the research when it comes to developing large-scale AI models for population-based studies and conducting comprehensive assessments of skeletal injuries. Despite being a cost-effective option, the inherent challenge of poor visible contrast of soft tissue (traditionally problematic for human visual inspection) in X-rays, coupled with the scarcity of longitudinal data, has left this area largely unexplored. This project aims to address this crucial knowledge gap.