Data integration
Solving the scale problem
The paradigm of data integration has emerged as one of the objectives of personalised precision oncology. One of the exciting challenges it presents is that different data types live in entirely different physical scales: from the tens of centimeters of medical images, down to the micrometers of subcellular structure and beyond. While medical imaging provides an all-encompassing frame of reference, as well as a vital quantitative and non-invasive measurement of the disease, integrating the complementary biopsy-based measurements is non-trivial from a spatial and analytical point of view.
In our research, we develop new techniques to enable the spatial co-registration of multi-scale data and extract insights about tumour heterogeneity. For example, we have designed software to automatically generate 3D printed moulds that allow us to take biopsies from specific locations that are co-registered with the patient's CT or MRI images [JCO CCI]. We have also developed a technique to achieve to achieve similar co-registration for in-vivo biopsies [European Radiology], in collaboration with our radiology colleagues. We are using these techniques to acquire multi-scale data with unprecedented spatial information, on which we will be able to apply exciting data analysis techniques.
