Accepting PhD Students

PhD projects

Pancreatic cancer is a high mortality cancer which progresses quickly following diagnosis. The average median survival from diagnosis ranges from 6 months for metastatic pancreatic cancer to 16 months for non-metastatic pancreatic cancer.
Patients with non-metastatic disease are classified according to whether their tumours can be removed by surgery as either (i) resectable upfront (R), (ii) locally advanced and unresectable (LA) or (iii) borderline resectable (BR). This classification is based on radiology reports and is crucial for patient management.
One of the problems that doctors face is adequately categorising these patients based on radiological criteria. Our group is currently undertaking a clinical trial (SCANPatient study) that evaluates the use of structured radiology reports to improve the accuracy of classifying patients’ tumours of the pancreas. The SCANPatient randomised trial will compare the standard narrative style radiological report with structured synoptic radiology report for reporting CT scans of the pancreas in suspected pancreas cancer.
As part of this project, you will use an existing algorithm and AI to capture the key information in the existing narrative pancreas radiology report to generate an automated structured synoptic radiology report in a checklist-style report template.