Barrett's oesophagus is a condition in which the area between the oesophagus and stomach no longer closes, allowing acidic contents of the stomach to enter the oesophagus and damage the lining. The project aims to assist clinicians by offering informed biopsy process, in which the system presents the operator with clinical outcomes of patients with visually similar GI tracts. The project goal is to assess the use of artificial intelligence-based similarity detection systems to better inform biopsy placement, increasing the reliability of bi-yearly inspections.

The project intention is to acquire secondary endoscopy captured data from clinicians to create a sufficient dataset to train a learning system to achieve the stated objective. The data required will feature footage (images/videos) from inside the oesophagus. The preferred data would contain associated diagnosis notes and description, however, data without diagnosis can still be used. All data can and will be anonymised, as personal information will not be useful in this work.

The data will be labelled by either researchers, or professional clinicians and prepared ready to feed into a chosen learning AI system. Through an iterative learning process, the chosen AI system will learn to discriminate between severity of Barrett's oesophagus and output optimal biopsy target.

Inclusion Criteria:

• Adult
• Attending Southmead Hospital for on OGD
• Has capacity
• Initially any indication, later patients with known Barrett's oesophagus

Exclusion Criteria:

• Children (<18)
• Patients without capacity
• Urgent and emergency procedures
• Limited English

Attending NBT clinic for on OGD phase 1, then later attending NBT clinic due to suspected Barrett's.