FEIT Research Project Database

Radiomics for cancer diagnosis and patient outcome prediction

Project Leader: Leigh Johnston
Collaborators: Nick Hardcastle (Peter MacCallum Cancer Foundation), James Korte (Peter MacCallum Cancer Foundation), Price Jackson (Peter MacCallum Cancer Foundation),
Primary Contact: Leigh Johnston (l.johnston@unimelb.edu.au)
Keywords: cancer; imaging; machine learning; magnetic resonance imaging MRI
Disciplines: Biomedical Engineering

Radiomics is an emerging field that aims to extract vital information from growing collections of medical imaging and patient data. A large number of quantitative features are extracted from medical images, from basic features such as shape and size to more advanced texture analysis metrics. Radiomic signatures are constructed from these features across a patient cohort and are being investigated for tumor staging, treatment response and patient outcome prediction.

Whilst a range of radiomics feature analysis tools are available through the research community, a bottleneck for many studies is manual contouring or segmentation of the tumor volume. We propose the application of Deep Learning, which has recently shown excellent performance in many image analysis tasks, to develop an auto-contouring tool. Such a tool would automate our radiomics analysis and allow us to mine a much larger portion of the existing medical image database.

The project would suit a student with a background in medical imaging and analysis. The initial project has potential to focus on MRI data and developing a prediction model for brain tumors. 

Further information: Please contact Dr James Korte (James.Korte@petermac.org) for further information about this project.