Small Bowel Deep Learning Algorithm Project

NCT03706664 | Active Not Recruiting

• 1 other identifier: IRAS No:238924
• Study Type: interventional
• Target: 226
• Locations: 1 country
• Sites: 1

Brief Summary

Crohn's disease affects 200,000 people in the UK (\~1 in 500), most are young (diagnosed \< 35 years) with costs of direct medical care exceeding £500 million. Crohn's disease is caused by an auto-immune response and affects any part of the digestive tract, most commonly the last segment of the small bowel (the terminal ileum). Magnetic resonance imaging (MRI) plays a role in 3 areas: Crohn's disease diagnosis , monitoring treatment response \& assessing development of complications. To evaluate the small bowel using MRI, Radiologists visually examine the scan slice-by-slice. The interpretation is time consuming and error-prone because of disease presentation variability and differentiation of diseased segments from collapsed segments. Deep learning for image analysis is based on a computer algorithm "learning" from human (Radiologist) generated training data. This method has been successfully applied to medical imaging, for example computer detection of lung cancer on chest X-rays. This pilot study investigates if a deep learning algorithm can identify and score segments of inflamed terminal ileum affected by Crohn's disease. To our knowledge this is the first project attempting to develop such an algorithm.The study will retrospectively review MR images obtained as part of standard care from patients being investigated for, Crohn's or being followed up with Crohn's disease. 226 patients' images will be used for the study. On fully anonymised images two Radiologists working at Northwick Park Hospital will score and outline normal and abnormal loops of terminal ileum. Imperial College computer science department will then develop a deep learning algorithm from imaging features of normal and abnormal loops. The study end-point is algorithm performance vs. images labelled by Radiologists. The eventual aim is to develop an algorithm that assists Radiologists in the accurate diagnosis and follow-up of patients with Crohn's disease.

Conditions

Crohn Disease

Outcome Measures

Primary Outcomes (1)

• Machine learning algorithm's ability to accurately localize the terminal ileum.

  Study will compare manually segmented regions of interest by Radiologists with predictions by machine learning localisation algorithm.

  24 months

Secondary Outcomes (2)

• Data processing time until a diagnosis reported by algorithm.

  24 months

• Machine learning algorithm's ability to accurately distinguish abnormal and normal terminal ileum.

  24 months

Study Arms (2)

Training of machine learning algorithm

OTHER

113 MR Enterography images labelled by Radiologists will be used to develop a machine learning algorithm to (1) localise the terminal ileum, (2) classify the terminal ileum as normal or abnormal.

Other: Machine learning algorithm

Testing of machine learning algorithm

OTHER

113 MR Enterography images labelled by Radiologists will be used to test the accuracy of the machine learning algorithm to (1) localise the terminal ileum, (2) classify the terminal ileum as normal or abnormal compared to Radiologists opinion. Cross Validation analysis will be used for data analysis.

Other: Machine learning algorithm

Interventions

Machine learning algorithm OTHER

Study will develop and test a machine learning algorithm using MR Enterography images labelled by Radiologists.

Testing of machine learning algorithm; Training of machine learning algorithm

Eligibility Criteria

• Age: 16 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Child (0-17), Adult (18-64), Older Adult (65+)

You may qualify if:

• Patient's age \>16 years of age, (this age cut off has been used in the recent METRIC trial investigating imaging in Crohn's disease)
• MRI sequences obtained include axial T2 weighted images; coronal T2 weighted images and axial post contrast MRI images.
• Normal MR Enterography studies reviewed in consensus by two Radiologists (UP \& PL). Normal is defined as no sites of small or large bowel Crohn's disease.
• MR Enterography studies reviewed in consensus by two Radiologists shows terminal ileal Crohn's disease. Patients with more than one segment of small bowel Crohn's disease including terminal ileum are eligible. Patients with terminal ileal Crohn's disease continuous with large bowel are eligible.
• Diagnosis of Crohn's disease of terminal ileum based on endoscopic, histological and radiological findings. (This criteria has been used in the recent METRIC trial investigating imaging in Crohn's disease).

You may not qualify if:

• Poor quality MRI images as judged by consensus Radiologist opinion.
• No more than 3 MRI scans will come from the same patient.
• MR Enterography shows any bowel abnormality not due to Crohn's.
• Patient has undergone previous small or large bowel resection (this will distort anatomy and is beyond the scope of the present project). Patients' with other previous surgeries are eligible.
• Patients with large bowel Crohn's disease not continuous with the terminal ileum.

Sponsors & Collaborators

• London North West Healthcare NHS Trust: lead
• Imperial College London: collaborator

Study Sites (1)

St Mark's Hospital

London, Harrow, HA13UJ, United Kingdom

Location

MeSH Terms

Conditions

Crohn Disease

Interventions

Machine Learning Algorithms

Condition Hierarchy (Ancestors)

Inflammatory Bowel Diseases; Gastroenteritis; Gastrointestinal Diseases; Digestive System Diseases; Intestinal Diseases

Intervention Hierarchy (Ancestors)

Algorithms; Mathematical Concepts

Study Officials

• Uday Patel, FRCR MBBS

  London NorthWest Healthcare NHS Trust

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: NONE
• Masking Details: Neither the Radiologists nor the Computer scientists/outcomes assessors will be masked to the image labels or if a given MR Enterography has been used in the training or validation dataset.
• Purpose: DIAGNOSTIC
• Intervention Model: SINGLE GROUP
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Model Details: Radiologists will label 226 MR Enterography images as normal or abnormal. The labelled images will be randomised between training and validation sets. The training dataset will be used to develop machine learning algorithm to localise the terminal ileum \& classify the terminal ileum as normal or abnormal. The validation dataset will test the accuracy of the algorithm against the Radiologists labels.

Study Record Dates

• First Submitted: October 11, 2018
• First Posted: October 16, 2018
• Study Start: March 1, 2019
• Primary Completion: August 1, 2025
• Study Completion: December 1, 2025
• Last Updated: May 23, 2025

Record last verified: 2025-05

Data Sharing

• IPD Sharing: Will not share

Locations

GB (1)