Computer aided-image recognition in Barrett’s oesophagus
Most of the oesophageal cancers in the Western world arise from Barrett’s oesophagus. This is a condition where inflammation triggered by longstanding reflux induces an intestinal epithelium with metaplasia that is genetically instable and premalignant. Barrett’s oesophagus is detected endoscopically by colour changes from light pink to salmon-red. Dysplastic areas and early cancers are seen as subtle alterations in surface and microvasculature pattern. However, up to 25% of cases of early oesophageal Barrett’s cancer are missed at endoscopy.1,2
In a joined project with the Ludwig Institute and medical engineering, we work on computer-aided automated recognition of dysplastic and neoplastic lesions in Barrett’s oesophagus.
Figure: Principle of narrow band imaging. (Source: www.olympus-europa.com)
Frequency ablation of Barrett’s oesophagus
We participate in the national UK Halo registry, a longterm study on outcome of radiofrequency ablation to prevent dysplastic Barrett’s epithelium from progressing to cancer. Unsolved questions are predictors of treatment failure, surveillance intervals and outcome of this method in dysplastic squamous epithelium.
Figure: Using optical filters such as narrow band
|Figure: Radiofrequency ablation in the oesophagus|
Prof. Xin Lu, Ludwig Institute for Cancer Research, Oxford
Prof. Jens Rittscher, Ludwig Institute and Oxford Insitute for Biomedical Engineering
Prof. Christoph Dietrich, Caritas Hospital, Bad Mergentheim, Germany
Prof. Stefan Zeuzem, University Frankfurt/Main, Germany