ABSTRACT
Gastric cancer is one of the most widely reported problems in the world, causing high modality rates in recent times. Gastroscopy is an efficient method that is widely used to analyze gastric problems. The advent of deep learning helps doctors to detect gastric cancer in the early stages. The performance of the existing methods in detecting gastric cancer from the images is not accurate. This study proposes a novel deep-learning framework that can be used to detect gastric cancer from gastric slice images. The proposed method is based on a patch-based analysis of the given input image. Specifically, the model selects and extracts the features from the images in the training phase and evaluates the genuine risk of the patients. This is one of the novel contributions of the proposed work. The bag-of-features technique is applied to the extracted features in the proposed network for the selected patches for better analysis. Experimental results prove that the proposed framework can detect gastric cancer from the images effectively and efficiently. The model is robust enough to detect the minute lesions that can cause the gastric tumor in the further stages. The dataset used in this analysis is publicly available, and the results achieved by this model are higher than the other conventional models that use the same dataset. The proposed framework gives higher accuracy scores compared with existing frameworks.
