Data analytics in radiation oncology


The variable response of tumours to radiation therapy presents a significant challenge in radiation oncology.  This variability occurs both within a tumour and between tumours of the same classification. Dr. Jeff Andrews and Dr. John Braun are exploring methods for characterizing these variations that pose a significant challenge when developing effective treatment plans for cancer. While classical tools typically ignore the spatial aspects of the data, they will incorporate spatial information from multiple image types into the analyses, with the ultimate goal of improving the understanding of the progression of the disease and aid in the delivery and therefore efficacy of radiation therapy.

Secondly, in collaboration with Dr. Carlone and Dr. Hyde (Clinical Physics BCCA- CSI), this team is investigating use of data analytic techniques and machine learning to improve imaging quality assurance. They will specifically investigate the imaging precision in stereotactic ablative radiotherapy used in the treatment of lung cancer and identify correlates of this to treatment outcomes.  This team also aims to build new optimization and verification techniques for 4 Pi dynamic trajectory stereotactic ablative radiotherapy.  Finally, the group will determine the accuracy and reproducibility of two-dimensional and three-dimensional approaches for quantitative imaging of tumour angiogenesis, including dual energy, temporal subtraction, and tomosynthesis approaches. This analysis will enable the development of an optimized protocol for quantitative imaging of breast tumour angiogenesis, the application of which is the development of novel statistical approaches for precision breast cancer therapies.