Active not recruitingNAINTERVENTIONAL

Small Bowel Deep Learning Algorithm Project

This study is looking at a new computer program, called a 'deep learning algorithm', to help doctors diagnose and monitor Crohn's disease in the small bowel using MRI scans. Crohn's disease affects many people in the UK, often young, and it can be hard to spot precisely on scans. Currently, doctors spend a lot of time carefully looking at each scan. This computer program learns from scans already looked at by expert doctors. The aim is to see if this program can accurately identify areas of inflammation in the small bowel, making the process quicker and more reliable for future patients. This could help doctors better manage Crohn's disease.

At a glance

Status

Active not recruiting

Phase

Sponsor

London North West Healthcare NHS Trust

Enrolment target

226

Start

01 Mar 2019

Estimated completion

01 Dec 2028

Crohn Disease

What is this study about?

Imagine your doctor needs to look closely at detailed images of your insides, like an MRI scan, to check for problems like Crohn's disease. Crohn's can be tricky to find because it affects different people in different ways, and it can be hard to tell healthy areas from unhealthy ones. This study is exploring a new way to help doctors with this vital task.

Researchers are developing a clever computer program, sometimes called 'artificial intelligence', that can learn to 'see' and understand these MRI scans. Think of it like teaching a computer to spot patterns. Expert doctors have already looked at many MRI scans from patients with Crohn's and marked out the healthy and unhealthy parts. The computer program then learns from these examples, just like a student learns from a teacher.

The main goal of this study is to test how well this new computer program can find and highlight areas of inflammation caused by Crohn's disease in the small bowel, especially the last part called the terminal ileum. They will compare the computer's findings with what experienced doctors have already identified. If successful, this technology could make it quicker and easier for doctors to diagnose Crohn's and see if treatments are working, ultimately helping people get the best care for their condition.

Key takeaways

Researchers are developing a computer program to help spot Crohn's disease on MRI scans.
The program learns from scans previously reviewed by expert doctors.
The study uses existing, anonymous MRI images, so no new tests or involvement from patients is needed.
The goal is to make diagnosing and monitoring Crohn's disease quicker and more accurate for future patients.

Who may be eligible?

To be included in this study, patients must be aged 16 years or older. This study looks back at MRI scans that were taken as part of your usual care, so you wouldn't need to have any new scans just for this research. The MRI scans being used must include specific types of images that provide a detailed view of the body.

If you have already had an MRI scan that showed no signs of Crohn's disease in your small or large bowel, those scans might be used as examples of 'normal' scans for the computer to learn from. On the other hand, if your MRI showed Crohn's disease specifically in the last part of your small bowel (the terminal ileum), those scans could also be included.

However, some scans won't be used. For instance, if the MRI pictures aren't clear enough, or if you've had surgery on your bowel that might change its normal shape. Also, if your scan showed problems that aren't related to Crohn's disease, or if your Crohn's was only in your large bowel and not connected to the terminal ileum, those scans wouldn't be suitable for this particular study.

Could this study suit you?

Answer these quick questions to see if you may be eligible. This is a guide only — the research team makes the final call.

Are you 16 years old or older?
Did your MRI scan include specific detailed image types?
Was your MRI scan clear enough for doctors to read well?
If you had Crohn's, was it in your terminal ileum (last part of the small bowel)?
Have you NOT had surgery on your bowels that would change their shape?

Answer every question to see your result.

What does participation involve?

You don't need to do anything to take part in this study. The researchers are looking back at MRI scans that have already been taken as part of people's regular medical care. This means no extra hospital visits, no additional tests, and no new medications or follow-up appointments for you. Your past images will be used anonymously, meaning your name and personal details won't be linked to them, ensuring your privacy.

Potential risks and benefits

Since this study only uses existing, anonymous MRI scans, there are no direct risks to the patients whose scans are being used. You won't have any additional procedures or treatments. The potential benefit is that if this computer program proves successful, it could help doctors in the future to diagnose Crohn's disease more accurately and quickly, leading to better care for others. As the study uses past, anonymised data, there is no option to withdraw your specific data once it has been included.

Locations (1)

Some site locations are approximate. We're improving this — please verify with the trial team before travelling.

St Mark's Hospital
Verified postcode
London, United Kingdom

Common questions

What is 'deep learning'?

It's a type of powerful computer program that can 'learn' from information, like medical images, to spot patterns and help doctors.

Will this study use my personal information?

No, all patient identities from the MRI scans are completely removed, so the information used is anonymous.

Do I need to do anything to be part of this study?

No, this study uses MRI scans that have already been taken, so you don't need to do anything.

How might this help people with Crohn's disease?

It could make it quicker and easier for doctors to diagnose Crohn's and see if treatments are working, leading to better care.

Is this computer program replacing doctors?

No, the aim is to assist doctors, helping them interpret scans more consistently and efficiently, not to replace their expertise.