Why this study is needed
Pancreatic cancer is one of the most difficult cancers to treat. For some patients, surgery offers the only chance of cure — but deciding on the best treatment option can be demanding, even for experts.
Doctors use CT scans and specialist team discussions to decide whether a tumour can be safely removed.
This assessment can be uncertain — and different specialist teams sometimes reach different conclusions from the same scan.
In some cases, patients undergo exploratory surgery but the tumour cannot be removed. This is distressing and can delay other treatments.
What PREDICT-PanC will do
The study evaluates a computer-based tool that analyses CT scans using artificial intelligence (AI) — a computer programme designed to support doctors, not replace them.
The AI tool estimates whether a pancreatic cancer appears directly removable by surgery, whether chemotherapy should be considered first, or whether non-surgical treatments are likely to be the most appropriate option. It provides an additional source of information to help specialist teams during their discussions.
The AI system has previously been evaluated in retrospective research studies and is now being tested prospectively in real clinical decision-making.
Silent Evaluation
The AI tool analyses CT scans in the background. Doctors make their decisions as usual, without seeing the AI's results. Researchers will compare the AI's predictions with what the clinical team decided — and with what was later found during treatment or surgery. The information gained during this phase will be used to update and improve the algorithm.
Decision Support
The AI predictions may be shown to specialist teams as additional information during clinical discussions. Clinicians will ultimately decide how to treat the patient. Researchers will evaluate whether this improves treatment decisions and patient outcomes.
What is the AI tool?
The AI tool is a computer programme that has been trained to analyse medical CT scan images. It looks for patterns in the images that may indicate whether a cancer can be safely removed by surgery.
It is important to understand that the AI does not make decisions about patient care. It produces an assessment — a second opinion, in a sense — that doctors can consider alongside their own expert clinical judgement.
The study is designed to find out whether this kind of tool is genuinely useful in real clinical practice. We are not assuming it will be — that is precisely what we are testing.
Why this matters for patients
These are the potential benefits we are investigating — the study is designed to test whether they are real, not to assume them.
- Reduce unnecessary exploratory surgery and its associated risks
- Help patients access the most appropriate treatment earlier
- Improve consistency of decision-making between hospitals and specialist teams
- Support doctors with additional information when interpreting complex CT scans
Data safety is our priority
All patient data used in this study is handled with the highest standards of care, in full compliance with NHS information governance and UK data protection law.
Anonymised Data
All CT scans and clinical information used in the study are fully anonymised. No names, dates of birth, or other personal identifiers are included in the research dataset.
No Data Sharing
Patient data is not shared with third parties, commercial organisations, or anyone outside the direct research team. Data remains within secure NHS and university systems.
Ethics Review
The study will undergo full NHS Research Ethics Committee review. All data handling follows the UK General Data Protection Regulation (UK GDPR) and the NHS Code of Confidentiality.
Secure Infrastructure
All data is stored on encrypted, access-controlled systems within NHS and university networks. Only authorised members of the research team can access the data.
Patient and public involvement
We believe that research should be shaped with input from people who have experienced pancreatic cancer care. Patient representatives help ensure the study focuses on outcomes that matter most to patients — and that the research is communicated clearly and respectfully.
If you have lived experience of pancreatic cancer, as a patient or as a carer or family member, your perspective is genuinely valuable to us. There is no obligation, no clinical commitment, and no specialist knowledge required.
How you could contribute
- Review study information and patient-facing materials
- Advise on outcomes that matter most to patients and families
- Comment on how AI is explained to patients
- Help ensure the study reflects real patient priorities
- Attend occasional short meetings — online or in person
Who is leading the study
The study is led by clinicians and researchers at the Royal Free London NHS Foundation Trust and University College London, working with collaborators in surgical oncology, radiology, artificial intelligence, and clinical trials methodology.
Mr Sebastian Staubli, MD, Dr. habil.
Senior Clinical Fellow, HPB & Liver Transplant Surgery · Royal Free London NHS Foundation Trust
Honorary Research Associate, University College London
Mr Staubli is a senior clinical fellow specialising in hepatopancreaticobiliary (HPB) surgery and liver transplantation. His research focuses on surgical quality improvement and the application of artificial intelligence to clinical decision-making in complex abdominal surgery.
Get in touch
If you are interested in finding out more about patient involvement, or simply want to have an informal conversation, please get in touch. There is no pressure to commit to anything — we would simply love to hear from you.
We aim to respond to all enquiries within 5 working days.
Senior Clinical Fellow, HPB & Liver Transplant Surgery
Royal Free London NHS Foundation Trust
Honorary Research Associate, UCL
Data privacy notice
By submitting this form, your message is sent via Web3Forms, a third-party form processing service. Your name, email, and message are forwarded to the research team and are not stored by Web3Forms beyond delivery. For study-related data, see the data protection section.