Student: Renyang Gu
1st supervisor: Vicky Goh, King’s College London
2nd supervisor: Isabel Dregely, King’s College London
Myeloma is a haematological cancer. It is debilitating causing unremitting bone pain & pathological fractures with a 5-year survival rate of 47%. Since myeloma can affect any marrow-containing bone, assessment of the whole skeleton is required to determine the optimum treatment.
The Challenge: No imaging techniques currently available allow for accurate whole-body skeletal delineation and fat quantitation. No standardised automated post processing tools are available for assessment of whole skeletal metrics. Manual skeleton delineation for quantifying skeletal fat fraction is time consuming (5+ hours) & requires clinical expertise.
There is an opportunity that quantitative whole-body imaging may address the real clinical need to rethink how we acquire and quantify marrow disease. This project aims to develop integrated whole-body UTE multi-echo Dixon MRI combined with Deep Learning of quantitative whole skeleton imaging metrics of bone marrow disease. The overall hypothesis is that whole body quantitative skeletal metrics will improve the diagnosis, characterisation of tumour burden as well as improve therapy response assessment. In addition, we will define whether quantitative CT is comparable to MRI for disease burden as a cost-effective alternative technique.
This project spans multiple disciplines: Physics/Mathematics (image acquisition, reconstruction) Computer Science (image analysis, machine learning), & Cancer Biology (biology of myeloma). It will generate new medical imaging approach, addresses a real clinical problem, and its success, the development of a novel streamlined acquisition and analysis method for quantitation and automated segmentation will impact significantly on current myeloma care and other cancers.