Endoscopy 2021; 53(S 01): S255-S256
DOI: 10.1055/s-0041-1724969
Abstracts | ESGE Days
ESGE Days 2021 Digital poster exhibition

The Impact Of Artificial Intelligence On The Adenoma Detection Rate (ADR): A Comparison Between Experienced And Trainee Endoscopists’ ADR

S Bernhofer
1   Universitätsklinikum St.Pölten, Gastroenterology and Hepatology, St. Pölten, Austria
,
A Maieron
1   Universitätsklinikum St.Pölten, Gastroenterology and Hepatology, St. Pölten, Austria
› Author Affiliations
 
 

    Aims Colorectal cancer (CRC) is the second most common cause of cancer death in the European Union. Most cases of CRC arise out of adenomas, which could be easily treated in a timely manner by endoscopic removal. Many studies have proven that a high adenoma detection rate (ADR) reduces the incidence of CRC. Artificial Intelligence (AI) is a promising new tool to achieve a high ADR. The aim of this study is to evaluate the impact of AI on the frequency of adenoma detection by endoscopists in training in comparison to the frequency of detection by senior endoscopists.

    Methods Data from all patients who underwent colonoscopy with a GI Genius enhanced colonoscope were collected within a period of two months. Endoscopists were divided into two groups, a trainee group (7 endoscopists; < 500 colonoscopies) and an expert group (4 endoscopists; > 1000 colonoscopies). The polyp and adenoma detection rates of both groups were calculated and compared using cross tabulation and the chi-square test.

    Results So far 150 patients (77 male, 73 female), mean age: 60 (SD± 16) years have been included. The most common indications for colonoscopy were screening (21,3 %), surveillance (17,3 %), GI disturbances (12,7 %), bleeding/anemia (12 %), CED (10,7 %) and elective polypectomy (10 %). Most procedures were done by the trainee group (n = 94). In total 311 polyps were removed. The polyp detection rate (PDR) was 69,1 % in the trainee group and 67,9 % in the expert group. The adenoma detection rate was 48,9 % in the trainee group and 46,4 % in the expert group. There was no significant difference between both groups in terms of PDR (p = 0,869) and ADR (p = 0,766).

    Conclusions Our interim analysis shows that Artificial Intelligence can help to minimize the difference of ADR between experienced and trainee endoscopists.

    Zoom Image
    Fig.1

    Citation Bernhofer S, Maieron A eP480 THE IMPACT OF ARTIFICIAL INTELLIGENCE ON THE ADENOMA DETECTION RATE (ADR): A COMPARISON BETWEEN EXPERIENCED AND TRAINEE ENDOSCOPISTS’ ADR. Endoscopy 2021; 53: S255.


    #

    Publication History

    Article published online:
    19 March 2021

    © 2021. European Society of Gastrointestinal Endoscopy. All rights reserved.

    Georg Thieme Verlag KG
    Rüdigerstraße 14, 70469 Stuttgart, Germany

     
    Zoom Image
    Fig.1