Deep Learning Based Thoracic X-ray Image Classification

Abstract

Radiology Image Analysis is a critical sector and this job mostly being done by medical specialists and people expect highest level of care and service regardless of cost. Due to complexity and subjectivity of images it is limited. Widespread variation exits across different interpreters and labour in terms of image interpretation by human experts. My objective is to analyze medical X-ray images using deep learning and utilize images using Pandas, Keras, OpenCV, Tensorflow etc to obtain classification of images like Atelectasis, Consolidation, Cardiomegaly, Edema, Effusion, Emphysema, Fibrosis, Her-nia, Infiltration, Mass, Nodule, Pleural, Pneumonia, Pneumothorax, Thickening etc. I have used Convolution Neural Networks(CNN) algorithm because compared to other image classification algorithms CNN have ability to automatically extract the high level representations from big data using little pre- processing. Ultimately, a simple and efficient model will lead clinicians towards better diagnostic decision for patients to provide them solutions with good accuracy.