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 the highest level of care and service regardless of cost. Due to the complexity and subjectivity of images, it is limited. Widespread variation exists across different interpreters and labor 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 a classification of images like Atelectasis, Consolidation, Cardiomegaly, Edema, Effusion, Emphysema, Fibrosis, Hernia, 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 the ability to automatically extract high-level representations from big data using little pre-processing. Ultimately, a simple and efficient model will lead clinicians towards the better diagnostic decision for patients to provide them solutions with good accuracy.