A neural network algorithm for detection of GI angiectasia during small-bowel capsule endoscopy - 19/12/18

Doi : 10.1016/j.gie.2018.06.036

Romain Leenhardt, MD ^1,^⁎, Pauline Vasseur, MSc ², Cynthia Li, MSc ^1,³, Jean Christophe Saurin, MD, PhD ⁴, Gabriel Rahmi, MD, PhD ⁵, Franck Cholet, MD, PhD ⁶, Aymeric Becq, MD ¹, Philippe Marteau, MD, PhD ¹, Aymeric Histace, PhD ^{^{∗
Drs Histace and Dray are co-last authors.},}², Xavier Dray, MD, PhD ^1,^2,^{^{∗
Drs Histace and Dray are co-last authors.}}
The CAD-CAP Database Working Group^{^†}
Computer-Assisted Diagnosis for Capsule Endoscopy Database Working Group: Sylvie Sacher-Huvelin: MD, Farida Mesli: MD, Chloé Leandri: MD, Isabelle Nion-Larmurier: MD, Stéphane Lecleire: MD, Romain Gerard: MD, Clotilde Duburque: MD, Geoffroy Vanbiervliet: MD, Xavier Amiot: MD, Jean Philippe Le Mouel: MD, Michel Delvaux: MD, Pierre Jacob: MSc, Camille Simon-Shane: MSc, Olivier Romain: PhD.
Sylvie Sacher-Huvelin, MD, Farida Mesli, MD, Chloé Leandri, MD, Isabelle Nion-Larmurier, MD, Stéphane Lecleire, MD, Romain Gerard, MD, Clotilde Duburque, MD, Geoffroy Vanbiervliet, MD, Xavier Amiot, MD, Jean Philippe Le Mouel, MD, Michel Delvaux, MD, Pierre Jacob, MSc, Camille Simon-Shane, MSc, Olivier Romain, PhD

¹ Sorbonne University, Department of Hepato-Gastroenterology, APHP, Saint Antoine Hospital, Paris, France
² ETIS, Université de Cergy-Pontoise, ENSEA, CNRS, Cergy-Pontoise Cedex, France
³ Drexel University, College of Arts & Sciences, Philadelphia, Pennsylvania, USA
⁴ Department of Endoscopy and Gastroenterology, Pavillon L, Hôpital Edouard Herriot, Lyon, France
⁵ Georges Pompidou European Hospital, APHP, Department of Gastroenterology and Endoscopy, Paris, France
⁶ Digestive Endoscopy Unit, University Hospital, Brest, France

^∗Reprint requests: Romain Leenhardt, MD, Sorbonne University, Department of Hepato-Gastroenterology, Paris F-75012, France.Sorbonne UniversityDepartment of Hepato-GastroenterologyParisF-75012France

Abstract

Background and Aims

GI angiectasia (GIA) is the most common small-bowel (SB) vascular lesion, with an inherent risk of bleeding. SB capsule endoscopy (SB-CE) is the currently accepted diagnostic procedure. The aim of this study was to develop a computer-assisted diagnosis tool for the detection of GIA.

Methods

Deidentified SB-CE still frames featuring annotated typical GIA and normal control still frames were selected from a database. A semantic segmentation images approach associated with a convolutional neural network (CNN) was used for deep-feature extractions and classification. Two datasets of still frames were created and used for machine learning and for algorithm testing.

Results

The GIA detection algorithm yielded a sensitivity of 100%, a specificity of 96%, a positive predictive value of 96%, and a negative predictive value of 100%. Reproducibility was optimal. The reading process for an entire SB-CE video would take 39 minutes.

Conclusions

The developed CNN-based algorithm had high diagnostic performances, allowing detection of GIA in SB-CE still frames. This study paves the way for future automated CNN-based SB-CE reading softwares.

Le texte complet de cet article est disponible en PDF.

Abbreviations : CAD, CAD-CAP, CNN, GIA, SB, SB-CE

Plan

Material and Methods

Computer-assisted diagnosis for a capsule endoscopy database

GIA still frames datasets

Control still frames datasets

Development of the detection algorithm

Analysis of still frame datasets

Outcomes

Results

Discussion

	DISCLOSURE: J. Saurin is a consultant for Capsovision, Medtronic, and Intromedic. G. Rahmi is a consultant for Medtronic. X. Dray is a consultant for Boston Scientific, Fujifilm, Pentax, and Medtronic. S. Huvelin is a consultant for Medtronic. All other authors disclosed no financial relationships relevant to this publication.
	If you would like to chat with an author of this publication, you may contact Dr Leenhardt at romainleni@gmail.com.

Export

Vol 89 - N° 1

P. 189-194 - janvier 2019 Retour au numéro

Article précédent

Continuing Medical Education Exam: January 2019
Karthik Ravi, William Ross, Ara Sahakian, Brian Weston, Prasad G. Iyer, Amit Rastogi, Michael B. Wallace

| Article suivant

Artificial intelligence and capsule endoscopy: Is the truly “smart” capsule nearly here?
Michael F. Byrne, Fergal Donnellan

Bienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.

Déjà abonné à cette revue ?

connectez-vous ou créez un compte

A neural network algorithm for detection of GI angiectasia during small-bowel capsule endoscopy - 19/12/18

Abstract

Background and Aims

Methods

Results

Conclusions

Plan

Export citations

Fichier

Contenu

Accès rapides

Mon compte

Aide & support

Plateformes Elsevier Masson

Déclaration CNIL