Breast Cancer Foundation NZ
Using deep learning and digital pathology for detailed breast tumour analysis
What is the problem and who is affected?
Digital pathology – looking at tumour samples on a computer screen rather than through a microscope – helps pathologists analyse tumours and report the subtype, which determines what treatment is required, to doctors faster.
What is this research hoping to achieve?
In this project, Dr Harris and his team will take digital pathology even further by developing machine-learning algorithms to automatically identify genomic features of the tumour that would otherwise only be found through expensive molecular testing. Such tests, which indicate how well a breast cancer will respond to treatment, are usually only available at considerable cost to the patient. Dr Harris’s project has the potential to make testing more affordable and accessible, enabling more patients to have tailored treatments that can increase chances of survival.
Research grant given in partnership with Breast Cancer Cure and the Health Research Council.