COBIx: Multi-site validation study of the COBIx reporting tool
Research type
Research Study
Full title
COBIx: Multi-site validation study of the Colon and Rectal Endoscopic Biopsy (COBIx) reporting tool
IRAS ID
330776
Contact name
David Snead
Contact email
Sponsor organisation
Univeristy Hospitals Coventry and Warickshire NHS Trust
Duration of Study in the UK
2 years, 0 months, 30 days
Research summary
The diagnosis of serious large bowel diseases such as colitis, Crohn’s disease and cancer, is done by examining tissue samples (biopsies) taken by endoscopic camera examination of the intestine. Large bowel biopsies of this type create a large volume of laboratory workload, comprising approximately 10% of all tissue requests. A significant percentage of these samples are normal (between 30-40%) and contain no evidence of disease. The samples are currently examined manually by a pathologist (a doctor trained to examine tissue), using a microscope. Recent investment means that more laboratories can now scan the microscope slides into a computer as a digital image. The COBIx algorithm takes advantage of digitisation by using computers to analyse biopsy image pixel data to find any irregularities that indicate the presence of disease. This project will fully optimize the COBIx algorithm to a design freeze and then test it more widely across more sites and with a greater number of cases. This is important because different labs have slightly different equipment, stain characteristics and patient populations; we need to ensure that the COBIx algorithm works equally well across different sites, despite these variations. Eleven NHS Hospital Trusts from England and Scotland have been chosen. Over the next 3 years, we will examine 11,000 large bowel biopsies from these centres, comparing the pathologists’ reports with the results of the COBIx algorithm. The results will be compared and analysed statistically. The goal is to see if COBIx accurately separates normal large bowel biopsies from the abnormal biopsies. This would enable normal biopsies to be solely reported by the computer program. Secondly, we wish to see if the detection of serious disease by COBIx is helpful in ensuring cases containing diseases such as cancer or severe inflammation can be prioritised for urgent pathologist review.
REC name
Wales REC 5
REC reference
23/WA/0317
Date of REC Opinion
6 Nov 2023
REC opinion
Favourable Opinion