Patients suffering with inflammatory bowel disease were invited to take part in a focused workshop at the University of Birmingham to assess their experiences.
Participants recognised that human error can lead to missed lesions and that limited operators could delay diagnosis. However, there was reluctance to have a computer alone make a diagnosis.
Artificial intelligence is fast becoming a valuable tool to assist clinicians in making a diagnosis at endoscopy. Despite its interest from a clinician perspective, little is known from the angle of our patient population.
The workshop was held virtually in view of COVID-19 restrictions. Patients were initially shown a video of capsule endoscopy and asked if they had any experience of this modality.
A concise explanation of capsule endoscopy was provided focused on the size of the device, the process involved and the limitations of its use; primarily human error in terms of missed lesions and insufficient operators available to view recordings.
Participants were then asked if they would prefer the use of artificial intelligence in making a diagnosis and any concerns they may have with its use.
Participants recognised that human error can lead to missed lesions and that limited operators could delay diagnosis. However, there was reluctance to have a computer alone make a diagnosis. A combined approach was preferred.
Furthermore, there was a desire to ensure that if diagnosed by a computer the management stage would be human led as opposed to technology led. Clinicians were preferred to help guide patients through the array of possible treatments.
This patient focused workshop highlighted the positives of using AI at capsule endoscopy. It is clear that a combined approach with the use of clinicians and computer aided detection would be most ideal.
With thanks to Laura Chapman for her steadfast organisation, and Dr Rachel Cooney and Dr Gerard Cummins for their expert oversight.
