All in one cancer imaging optimisation using an integrated mathematical and deep learning approach

Our goal is to develop a bespoke automated platform to track tumour dynamics. The computational pipeline will include computerised tomography (CT) image reconstruction, segmentation and radiomics analyses all performed jointly and returning an analysis that is independent of scanner manufacturer (please see the diagram).

This will extend the usefulness of standard of care CT beyond the traditional qualitative assessment of the disease burden into robust quantification of dynamics of tumour volume and heterogeneity. It will allow real-time precise treatment decisions in both the clinic setting and multidisciplinary team meetings.