Data Sharing With Collaborative Partners to Develop Computer Aided Detection for the Assessment of the Small Bowel Using MRI
Crohn's disease is characterised by an abnormal immunological response within the bowel wall leading to abnormal wall thickening, stricturing (narrowing), fistulation (abnormal connections) to adjacent organs and strictures (narrowing), abnormal motility, and local sepsis (infection). Radiological imaging of the small bowel defines diagnosis, disease extent, biological activity and complications and is vital for timely and efficacious clinical management. Small bowel magnetic resonance imaging (MRI) does not impart ionising radiation and is proving to be a safe, well tolerated and robust method of assessing the small bowel and is widely implemented in the NHS and Europe. MRI evaluates multiple disease related features such as bowel wall thickness and motility which are proving increasing reliable for disease identification, staging, therapeutic guidance and assessment of treatment response. Currently however radiologists must manually make these measurements which are time consuming and difficult. There are no computer tools that can quickly and accurate make relevant measurements on MRI to guide patient management. Radiologists at University College Hospital have formed collaborations with groups within University College London (UCL) and around Europe. The investigators have received grant funding to develop computer software to accurately assess the small bowel using MRI over 3 years. To develop this software, it is necessary to use anonymised datasets from patients with and without Crohn's disease undergoing small bowel MRI. Part of this project will require prospective collection of MRI data which has been granted ethical approval as a major amendment to a currently running project (09/H0714/62). The department of Radiology at UCLH has been running a clinical small bowel MRI service since 2005 and have several hundred datasets on its PACS. This current ethics application seeks permission to datashare with the collaborative partners fully anonymised MRI datasets and relevant clinical data from patients previously undergoing small bowel MRI for clinical indications at UCLH.
• Had a small bowel MRI at UCLH between 2005 and Sept 2010 as part of routine clinical practice who:
• Have a clinical standard of reference within 6 weeks of the MRI scan
• Patients with normal small bowel MRI studies and no clinical evidence of bowel pathology on any additional diagnostic investigations