Artificial intelligence system using white light for real-time optical characterization of colonic polyps

Aims To prospectively evaluate the clinical feasibility as well as diagnostic performances of AI-alone and AI-assisted OD of DCPs in a real-life setting.

Methods Consecutive outpatients referred for colonoscopy with at least one DCP were evaluated. DCPs were real-time classified by AI (AI-alone OD) and by the endoscopist with the assistance of AI (AI-assisted OD).The histopathology was the reference standard [1] [2] [3] [4] [5].

Results Overall 480 DCPs were detected, and 460 retrieved. AI provided a clinically relevant outcome in 81.4% DPCs (“adenoma” or “non-adenoma” in 71.0% and 10.4%, respectively), while 19.6% of DPCs were labelled as “no prediction”. Sensitivity, specificity, PPV, NPV and overall accuracy of AI-alone OD were 97.0% (95%CI: 94.0-98.6), 38.1% (95%CI: 28.9-48.1), 80.1% (95%CI: 75.2-84.2), 83.3% (95%CI: 69.2-92.0) and 80.5% (95%CI: 68.7-82.8%), respectively. The same figures for AI-assisted OD were: 94.8% (95%CI: 91.1-97.1), 58.9% (95%CI: 49.7-67.5), 82.4% (95%CI: 77.4-86.5), 84.9% (95%CI: 75.2-91.4) and 83.0% (95% CI: 78.8-86.6), respectively. Clinical performances of AI-assisted OD experts and non-experts were: sensitivity (96.1% vs. 93.6%), specificity (65.0% vs.52.5%), positive predictive value (84.7% vs. 80.1%), negative predictive value (89.1% vs. 80.0%) and overall accuracy (85.8% vs. 80.1%).

Conclusions AI-alone OD is feasible in >80% of DCPs in clinical practice. AI-alone showed a high sensitivity and suboptimal specificity. The human-machine interaction results in improved diagnostic performances, especially when experts are involved.

Conflicts of interest

Medtronic CoFujifilm Co

• 1 Houwen BBSL, Hassan C, Coupé VMH. et al. Definition of competence standards for optical diagnosis of diminutive colorectal polyps: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2022; 54 (01) 88-99
• 2 Pecere S, Antonelli G, Dinis-Ribeiro M. et al. Endoscopists performance in optical diagnosis of colorectal polyps in artificial intelligence studies. United European Gastroenterol J 2022; 10 (08) 817-826
• 3 Rondonotti E, Hassan C, Tamanini G. et al. Artificial intelligence-assisted optical diagnosis for the resect-and-discard strategy in clinical practice: the Artificial intelligence BLI Characterization (ABC) study. Endoscopy. 2022
• 4 Hassan C, Balsamo G, Lorenzetti R. et al. Artificial Intelligence Allows Leaving-In-Situ Colorectal Polyps. Clin Gastroenterol Hepatol 2022; 20 (11) 2505-2513
• 5 Biffi C, Salvagnini P, Dinh NN. et al. Standalone performance of a novel intelligent device for real-time optical characterization of colorectal polyps: A multi-reader, prospective study. Springer Nature. 2021 LATEX template

Publication History

Article published online:
14 April 2023