Convolutional Neural Networks for Optical Discrimination Between Histological Types of Colorectal Polyps Based on White Light Endoscopic Images-Reference-Cited by-同舟云学术

Convolutional Neural Networks for Optical Discrimination Between Histological Types of Colorectal Polyps Based on White Light Endoscopic Images

Published:2023-05-18 Issue: Volume: Page:
ISSN:0926-9630
Container-title:Caring is Sharing – Exploiting the Value in Data for Health and Innovation
language:
Short-container-title:

Author:

Panteris Vasileios¹²,Feretzakis Georgios³⁴,Karantanos Panagiotis¹,Kalles Dimitris³,Verykios Vassilios V.³,Panoutsakou Maria¹,Karagianni Eirini¹,Zoubouli Christina⁵,Vgenopoulou Stefani⁵,Pierrakou Aikaterini⁵,Theodorakopoulou Maria⁵,Papalois Apostolos E.⁶²,Thomaidis Thomas⁷²,Dalainas Ilias⁸,Kouroumalis Elias⁹

Affiliation:

1. Gastroenterology Department, Sismanogleio General Hospital, Marousi, Greece

2. Hellenic Society of Gastrointestinal Oncology

3. School of Science and Technology, Hellenic Open University, Patras, Greece

4. Department of Quality Control, Research and Continuing Education, Sismanogleio General Hospital, Marousi, Greece

5. Pathology Department, Sismanogleio General Hospital, Marousi, Greece

6. Special Unit for Biomedical Research and Education, School of Medicine, Aristotle University of Thessaloniki, Greece

7. 2nd Gastroenterology Department, Hygeia Hospital, Athens, Greece

8. Administration, Sismanogleio General Hospital, Marousi, Greece

9. Department of Gastroenterology, University of Crete Medical School, Heraklion, Greece

Abstract

The objective of this study was to compare different convolutional neural networks (CNNs), as employed in a Python-produced deep learning process, used on white light images of colorectal polyps acquired during the process of a colonoscopy, in order to estimate the accuracy of the optical recognition of particular histologic types of polyps. The TensorFlow framework was used for Inception V3, ResNet50, DenseNet121, and NasNetLarge, which were trained with 924 images, drawn from 86 patients.

Publisher

IOS Press

Link

https://ebooks.iospress.nl/pdf/doi/10.3233/SHTI230208