An Improvement for Medical Image Analysis Using Data Enhancement Techniques in Deep Learning
Authors:
- 1 September 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-3
Abstract
A huge number of applications and algorithms have been suggested to analyze medical images. Recent developments in deep learning especially, deep convolutional neural networks (CNN), improved the performance of medical image classification methods. However, training a deep CNN from scratch with medical images is complicated task as it requires a large amount of labelled data. In this paper, we show the role of using different data augmentation techniques to solve such problems. Firstly, we created a deep CNN model with twelve layers for image classification. Then we trained this network with original computed tomography scan (CT) image dataset and some new datasets which are created by generating new images using our original image data. By comparing the classification results of our network on different datasets, we show that using data augmentation techniques can help to improve the medical image classification results and to boost up the network performance.Keywords
This publication has 7 references indexed in Scilit:
- Recurrent Topic-Transition GAN for Visual Paragraph Generation2017
- Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks2017
- Deep Residual Learning for Image Recognition2016
- Going deeper with convolutions2015
- Multi-column deep neural networks for image classification2012
- Chinese Handwriting Recognition Contest 20102010
- Document Image Defect Models1992