{
  "abstract": "Introduction In recent years, the availability of Artificial Intelligence (AI) in endoscopy has increased. Numerous studies have been published demonstrating an increase in both polyp and adenoma detection rate (PDR & ADR) of between 10-25%. 1–3 AI therefore shows promise at improving the standards of colonoscopy. Despite this data, uptake is still variable amongst endoscopists with mixed perceptions of its benefit. We aimed to perform an up-to-date assessment of current usage, perceived benefits and barriers to use of AI in colonoscopy by independent endoscopists across four NHS hospital trusts.Methods We created a google form asking endoscopists about their current AI usage. We then asked them to select from a pre-formulated list of options, describing the benefits they associate with AI and the barriers that reduced their usage. Recipients were allowed to choose multiple selections from the list of options.Results In total 35 endoscopists responded to the survey. 20 gastroenterology consultants, 7 gastroenterology registrars or clinical fellows with full accreditation, 6 clinical endoscopists and 2 surgical consultants. 25.7% (n=9) of the participants were Bowel Cancer Screen Program (BCSP) accredited. Over half of the recipients had been practising for over 10 years. 85.7% (n=30) of recipients were aware that their endoscopy stacks were equipped with an endoscopic AI assistant, however only 20% (n=7) claimed to use it routinely. Of the routine users, 5 were gastroenterology consultants and 2 clinical endoscopists. Of these, all had over 5 years’ experience as an independent endoscopist with the majority practising for over 10 years. Only 11.4% (n=4) had had any formal training on using AI. When asked to comment on benefits in using AI 36.8% (n=7) felt it helped them identify more adenomas. The most commonly quoted barriers to usage, was that AI was too distracting (34.3%, n=12), that there were too many false positives (31.4%, n=11) or that the endoscopist had not received any formal training (22.9%, n=8) ( figure 1). The primary limitation for this study was that overall participant number is low.Conclusions Despite support for its usage in large scale trials, everyday incorporation of AI in diagnostic colonoscopy has not been readily taken up. In part this may be due to the lack of formal training. This likely limits the tools perceived benefit whilst simultaneously making it more distracting. This lends scope to the concept that with more widespread, formalised AI training, usage could increase, thereby increasing both polyp and adenoma detection rate.References Wang P, Berzin TM, Brown JR, Bharadwaj S, Becq A, Xiao X, Liu P, Li L, Song Y, Zhang D, Li Y. Real-time automatic detection system increases colonoscopic polyp and adenoma detection rates: a prospective randomised controlled study. Gut. 2019 Oct 1;68(10):1813-9.Repici A, Spadaccini M, Antonelli G, Correale L, Maselli R, Galtieri PA, Pellegatta G, Capogreco A, Milluzzo SM, Lollo G, Di Paolo D. Artificial intelligence and colonoscopy experience: lessons from two randomised trials. Gut. 2022 Apr 1;71(4):757-65.Seager A, Sharp L, Neilson LJ, Brand A, Hampton JS, Lee TJ, Evans R, Vale L, Whelpton J, Bestwick N, Rees CJ. Polyp detection with colonoscopy assisted by the GI Genius artificial intelligence endoscopy module compared with standard colonoscopy in routine colonoscopy practice (COLO-DETECT): a multicentre, open-label, parallel-arm, pragmatic randomised controlled trial. The Lancet Gastroenterology & Hepatology. 2024 Oct 1;9(10):911-23.Abstract P10 Figure 1",
  "authors": [
    {
      "affiliations": [
        "West Hertfordshire NHS Teaching Trust, Watford, United Kingdom"
      ],
      "name": "Dylan Angel"
    },
    {
      "affiliations": [
        "Bedfordshire NHS Foundation Trust, Luton, United Kingdom"
      ],
      "name": "Noam Roth"
    },
    {
      "affiliations": [
        "East and North Hertfordshire NHS Trust, Stevenage, United Kingdom"
      ],
      "name": "Adedeji Oyefeso"
    },
    {
      "affiliations": [
        "East and North Hertfordshire NHS Trust, Stevenage, United Kingdom"
      ],
      "name": "Sabina Beg"
    }
  ],
  "title": "P10 An up-to-date review of uptake and perceived barriers in using artificial intelligence in diagnostic colonoscopy. A four hospital multi-centre analysis",
  "uid": "ad908cb6-e909-5b27-ac2b-9cd1f2896e67"
}