Dr Yinyin Yuan

Using image processing techniques, 2D/3D tumour models can be reconstructed from histopathological/immunohistochemistry images of tumour sections or in vitro cells. The spatial patterns and interactions of different cell types can be automatically quantified and visualised. This approach allows us to explore higher order parameters of the tumour microenvironment, providing valuable tools for correlating complex phenotype landscapes with genomics and clinical outcome.

Her team work on developing such image processing pipelines for two main purposes:

1. In pathology, quantified image patterns of tumour models can be directly integrated with genomic, transcriptomics and other types of molecular signals to develop predictive and diagnostic tools. Integrating rich information from both phenotypic and molecular data allows us to draw on the power of both histopathological and molecular analysis.
   
2. In cancer biology, image patterns of in vivo and in vitro models can be used to advance our understanding of the tumour microenvironment, in particular, the mechanisms and consequences of interactions between different types of cells. Linking image patterns with experimental interventions will enhance our ability to detect important patterns that would otherwise have been missed by eye.

Click to enlarge image

Yinyin joined ICR in 2012 as the leader of the Computational Pathology and Integrative Genomics team. Trained as a computer scientist, she finished a 5-year BSc degree within 4 years (2003) at the University of Science and Technology of China, before obtaining her MSc (2005) and PhD (2009) at University of Warwick. At Warwick she became interested in studying genetic regulation in plant disease by leveraging statistical analysis tools originally developed for other disciplines such as economics.

Intrigued by the immense biological complexity underlying diseases, she started her postdoctoral research at the Cambridge Research Institute Cancer Research UK to characterise the molecular landscape of breast cancer. Together with an international research team, she discovered 10 new molecular subtypes of breast cancer in 2,000 patients. As a Junior Research Fellow at Wolfson College, University of Cambridge, she interacted with researchers from various backgrounds such as physics, engineering and humanity substantially benefitting her future research.

During her time in Cambridge, she realised how quantitative analysis can be developed into an elegant yet powerful approach to provide objective solutions in pathology, and how computer science can help achieve this. By training a computer to automatically identify cancer cells in pathological specimens just like how cameras recognise faces, a brand new modern framework can be built upon pathological observations with a wealth of knowledge accumulated over centuries. In doing so, the characteristics of diseases in hundreds of patients and their treatment strategies can be objectively evaluated to gain understanding about causal mutations and accordingly the best treatment.

At ICR, her aim is to continue bring in techniques from a broader range of fields to formulate unique approaches for linking genetic mutations, pathological observations and patient treatment.

Outside work she is an active rock climber and hiker. She also enjoys dancing having performed as a contemporary dancer, and live music having played caixa in a samba band.