About the Project
For a full project proposal and details on how to apply using our online recruitment portal please see icr.ac.uk/phds. Please note we only accept applications via the online application system apply.icr.ac.uk
Clinical imaging with CT and MRI is used in nearly all late phase trials of cancer therapies to monitor disease and determine survival endpoints such as progression-free survival (PFS). Methods that enable this, such as the Response Evaluation Criteria in Solid Tumours (RECIST) are objective and can be deployed uniformly across tumour types, participating clinical trial sites and different clinical trials. In addition, various functional imaging methods derived from MRI, PET and other modalities are used in an emerging number of trials that evaluate novel therapies.
In all of these cases, spatially complex data are collected that image each entire tumour at multiple time points. Further, many patients have multiple tumours and these show variation (heterogeneity) in their pathophysiology. However, imaging data is usually reduced to single composite readouts that fail to capture the full potential of this spatially heterogeneous data.
In this PhD we determine the optimum study design for imaging data in cancer clinical trials. We will explore the role for tracking the dynamic changes of single lesions over time and incorporate this into assessment of PFS. We will evaluate how data from multiple lesions may increase the power for a given number of patients, while accounting for within-subject effects that arises from patients having multiple lesions. We will assess how the performance of data rich functional imaging biomarkers may differ from anatomical (RECIST) biomarkers. To do this, it is critical to understand the correlated nature of data emerging from imaging studies, and apply appropriate advanced statistical analyses.
The successful applicant can expect to gain (1) experience of conducting comprehensive literature review, (2) knowledge of current use of imaging in clinical trials and the pitfalls of these approaches, (3) detailed experience of data simulation, mathematical modelling and advanced statistical methods and their implementation in software such as R or STATA, (4) knowledge and experience in the design and execution of data science experiments. Finally, it is expected that the research will result in peer reviewed accepted publications.
This PhD is multi-disciplinary with supervisors from radiology and medical statistics background. The appointee will join a large interactive research group and will benefit further from interaction with external leading experts in the field.
Keywords /Subject Areas
- Imaging Biomarkers
- Statistical Modelling
- Correlated data
- Clinical Trials
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.