Gastric Section Detection Based on Decision Fusion of Convolutional Neural Networks

Abstract

To provide accurate histological parameter assessment of each gastric section from endoscopic images, gastric sections need to be correctly identified in advance. In this paper, we propose a novel CNN based ensemble learning method to detect gastric sections from endoscopic images by fusing decisions of multiple convolutional neural network (CNN) models which provide initial decision probability of the endoscopic image. The decision probability is concatenated and classified by a decision fusion network to achieve effective and efficient gastric section detection. In the experiments, we compare the proposed method with state-of-the-art CNN and CNN based ensemble learning methods and conclude that the proposed method owns the best testing accuracy.