Endoscopy 2022; 54(S 01): S18-S19
DOI: 10.1055/s-0042-1744593
Abstracts | ESGE Days 2022
ESGE Days 2022 Oral presentations
1:00–12:00 Thursday, 28 April 2022 Club E. Endoscopic grading and surveillance in IBD

A VIRTUAL CHROMOENDOSCOPY ARTIFICIAL INTELLIGENCE SYSTEM TO DETECT ENDOSCOPIC AND HISTOLOGIC REMISSION IN ULCERATIVE COLITIS

M. Iacucci
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
2   National Institute for Health Research (NIHR) Birmingham Biomedical Research Centre, Birmingham, United Kingdom
3   Division of Gastroenterology and Hepatology, University of Calgary, Calgary, Canada
,
R. Cannatelli
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
4   Gastroenterolgy and Digestive Endoscopy Unit, Department of Biochemical and Clinical Sciences “L. Sacco”, University of Milan, ASST Fatebenefratelli Sacco, Milan, Italy
,
T.L. Parigi
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
5   Department of Biomedical Science, Humanitas University, Milan, Italy
,
A. Buda
6   Gastroenterology Unit, Santa Maria del Prato Hospital, Feltre, Italy
,
N. Labarile
7   Section of Gastroenterology II, National Institute of Research “Saverio De Bellis”, Castellana Grotte, Italy
,
O.M. Nardone
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
8   Gastroenterology, Department of Clinical Medicine and Surgery, University Federico II of Naples, Naplesitit, Italy
,
G.E. Tontini
9   Division of Gastroenterology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
,
A. Rimondi
10   University of Milan, Milan, Italy
,
A. Bazarova
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
11   Institute for Biological Physics, University of Cologne, Cologne, Germany
,
P. Bhandari
12   Division of Gastroenterology, Queen Alexandra Hospital, Portsmouth, United Kingdom
,
R. Bisschops
13   Division of Gastroenterology, University Hospitals Leuven, Leuven, Belgium
,
G. De Hertogh
13   Division of Gastroenterology, University Hospitals Leuven, Leuven, Belgium
,
R. del Amor
14   Instituto de Investigación e Innovación en Bioingeniería, Universitat Politècnica de València, València, Spain
,
J.G Ferraz
3   Division of Gastroenterology and Hepatology, University of Calgary, Calgary, Canada
,
M. Goetz
15   Division of Gastroenterology, Klinikum, Böblingen, Germany
,
X. Gui
16   Division of Gastroenterology, University of Washington, Seattle, Washington, United States
,
B. Hayee
17   Division of Gastroenterology, Kings College London, London, United Kingdom
,
R. Kiesslich
18   Helios HSK Wiesbaden, Wiesbaden, Germany
,
M. Lazarev
19   Division of Gastroenterology, Johns Hopkins Hospital, Baltimore, Maryland, United States
,
V. Naranjo
14   Instituto de Investigación e Innovación en Bioingeniería, Universitat Politècnica de València, València, Spain
,
R. Panaccione
3   Division of Gastroenterology and Hepatology, University of Calgary, Calgary, Canada
,
A. Parra-Blanco
20   Division of Gastroenterology, University of Nottingham, Nottingham, United Kingdom
,
L. Pastorelli
21   Liver and Gastroenterology Unit, Department of Health Sciences, Universita' degli Studi di Milano, DeparmtASST Santi Paolo E Carlo, University Hospital San Paolo, Milan, Italy
,
T. Rath
22   Division of Gastroenterology, University of Erlangen, Erlangen, Germany
,
E.S Røyset
23   Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Norway
,
M. Vieth
24   Klinikum Bayreuth, Bayreuth, Germany
,
V. Villanacci
25   Institute of Pathology, Spedali Civili, Brescia, Italy
,
D. Zardo
26   Department of Pathology, San Bortolo Hospital, Vicenza, Italy
,
S. Ghosh
27   APC Microbiome Ireland, College of Medicine and Health, Cork, Ireland
2   National Institute for Health Research (NIHR) Birmingham Biomedical Research Centre, Birmingham, United Kingdom
1   Institute of Immunology and Immunotherapy, NIHR Wellcome Trust Clinical Research Facilities, University of Birmingham, and University Hospitals Birmingham NHS Trust, Birmingham, United Kingdom
3   Division of Gastroenterology and Hepatology, University of Calgary, Calgary, Canada
,
E. Grisan
28   School of Engineering Computer Science and Informatics, London South Bank University, London, United Kingdom
29   Department of Engineering, University of Padova, Padova, Italy
› Author Affiliations
› Further Information
Also available at
 

Aims We aimed to develop an artificial intelligence (AI) system to assess endoscopic remission (ER) and histologic remission (HR) of ulcerative colitis (UC) in both white light (WL, using Ulcerative Colitis Endoscopic Index of Severity [UCEIS]) and virtual chromoendoscopy (VCE, using Paddington International Virtual ChromoendoScopy ScOre [PICaSSO]).

Methods A convolutional neural network (CNN) was developed based on 559 endoscopy videos, from 302 UC patients prospectively included in the PICaSSO multicentre study. The videos were divided in training (254), validation (62), and testing (243), and comprised 67280 frames in total. The CNN was trained to predict both ER (defined as UCEIS≤1 in WL and as PICaSSO≤3 in VCE) and HR (defined as Robarts Histological Index ≤ 3 with no neutrophils in lamina propria or epithelium) in video clips.

Zoom Image
Fig. 1

Results In the validation cohort, our system predicted ER in WL videos with 82% sensitivity, 94% specificity and an area under the ROC curve (AUROC) of 0.92. In VCE, sensitivity was 74%, specificity 95%, and AUROC 0.95. In the testing cohort, the diagnostic performance remained similar.

The diagnostic performance for the prediction of HR in the validation set had sensitivity, specificity, and accuracy of 92%, 83%, and 85%, respectively, using VCE; and 83%, 87%, and 86% respectively, with WL. In the testing set, these metrics declined modestly while remaining good. Of note, the algorithm’s prediction of histology was similar with VCE and WL videos.

Table 1

Diagnostic performance

PICaSSO≤3

UCEIS≤1

RHI≤3 and no neutrophils in LP

Validation/Testing cohort

VCE 62 videos/243 videos

WL 58 videos/241 videos

VCE 61 videos/241 videos

WL 59 videos/238 videos

Sensitivity

0.74 (0.53 – 0.93)/0.60 (0.45 – 0.74)

0.82 (0.67 – 0.98)/0.68 (0.49 – 0.83)

0.92 (0.79 – 1.06)/0.78 (0.70 – 0.82)

0.83 (0.62 – 1.04) /0.89 (0.84 – 0.93)

Specificity

0.95 (0.89 – 1.02)/0.89 (0.84 – 0.93)

0.94 (0.87 – 1.02)/0.88 (0.83 – 0.92)

0.83 (0.72 – 0.94)/0.77 (0.63 – 0.88)

0.87 (0.78 – 0.97)/0.49 (0.34 – 0.64)

Accuracy

0.89 (0.81 – 0.97)/0.83 (0.78 – 0.88)

0.89 (0.81 – 0.97)/0.85 (0.80 – 0.90)

0.85 (0.76 – 0.94)/0.77 (0.71 – 0.82)

0.86 (0.77 – 0.95)/0.81 (0.75 – 0.86)

Conclusions Our AI system accurately recognizes ER in videos and predicts HR equally well.



Publication History

Article published online:
14 April 2022

© 2022. European Society of Gastrointestinal Endoscopy. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany