Endoscopy 2023; 55(S 02): S84-S85
DOI: 10.1055/s-0043-1765197
Abstracts | ESGE Days 2023
Oral presentation
Engineering Endoscopy: Circuits and bolts, muscles and joints 22/04/2023, 08:30 – 09:30 Liffey Meeting Room 3

Artificial intelligence increases sustainability of colonoscopy by promoting histology-sparing strategies

M. Topa
1   Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
,
G. E. tontini
1   Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
,
N. Nandi
1   Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
,
L. Scaramella
2   Gastroenterology and Endoscopy Unit, Foundation IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy
,
F. Cavallaro
2   Gastroenterology and Endoscopy Unit, Foundation IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy
,
M. Vecchi
1   Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
,
L. Elli
2   Gastroenterology and Endoscopy Unit, Foundation IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy
› Author Affiliations
 
 

    Aims Artificial intelligence (AI) has made great strides over the past years in endoscopy. The ongoing climate crisis forced the healthcare systems to include sustainability in the decision-making process. The aim of our study is to define the CO2 footprint modifications induced by AI use in colorectal polyps’ identification.

    Methods Findings from a recently developed AI–based medical device (GI Genius, Medtronic), allowing a real-time Computer-Aided polyps detection (CADe) and a Computer-Aided-Diagnosis (CADx) were used to define sensitivity, specificity, accuracy, positive and negative predictive values for colon polyps. Data were then adjusted according to the available carbon footprint of routinary histologic examination (0.28 kg CO2, 1 jar).

    Results In the reported cohort, GI Genius CADx provided an optical diagnosis of 454/476 diminutive polyps, 295 in the rectosigmoid tract. According to CADx, 242/295 rectosigmoid diminutive polyps were amenable for a leave-in-situ strategy being diagnosed as non-adenomas with a NPV for CADx adenoma prediction of 97.6% (95%CI, 94.1%-99.1%, p=0.002). 212 polyps overall, both adenomas and non-adenomas (44.5% of all diminutive polyps retrieved) would be amenable for a resect-and-discard approach. Overall, using CADx 95.4% (454/476) of the diminutive polyps could avoid histologic analyses with a reduction of CO2 emission ranging from about 133kg to 13kg (Table 1). Notably, no adenocarcinoma was found ([Table 1]).

    Zoom
    Table 1 Scenarios of different biopsy-sparing approaches.

    Conclusions CADx could efficiently reduce the CO2 footprint of colonoscopy by avoiding histological analyses. From 2016 to 2021, about 2,432,900 subjects underwent screening colonoscopies in Italy, thus the environmental effect could be relevant in supporting a "green endoscopy".


    Conflicts of interest

    Authors do not have any conflict of interest to disclose.

    Publication History

    Article published online:
    14 April 2023

    © 2023. European Society of Gastrointestinal Endoscopy. All rights reserved.

    Georg Thieme Verlag KG
    Rüdigerstraße 14, 70469 Stuttgart, Germany

     
    Zoom
    Table 1 Scenarios of different biopsy-sparing approaches.