Developing Multimodal AI for Personalized Cancer Diagnosis

1:2 people will get cancer in their lifetimes. Performing diagnosis earlier, more accurately, and in ways personalized to patient will improve treatment, life expectancy, and patent health.

The future of diagnosis is ‘multi-modal’. The ability to acquire multiple different types of data regarding a patient’s tumour is driving the next generation of oncology discovery and care. The key challenge is to develop frameworks that maximally leverage these different diagnostic modalities to best enable clinical decisions.

It is now the time to leverage artificial intelligence (AI) based tools for multimodal data analysis and integration. But while AI has already transformed several domains such as language processing, object/facial recognition, and consumer profiling; its use in the development of diagnostics has lagged behind.

We seek a PhD student to work as part of an interdisciplinary team of AI specialists, clinician scientists, and pathologists to develop AI-based diagnostic tools including: i) Deep learning (DL) digital pathology models capable of automated analysis of histopathology images; ii) Integrative technologies which that combine histopathology, omics, and patient data to generate multimodal diagnostic signatures.