摘要
Lung infiltration is a non-communicable condition where materials with higher density than air exist inthe parenchyma tissue of the lungs. Lung infiltration can be hard to be detected in an X-ray scan even for aradiologist, especially at the early stages making it a leading cause of death. In response, several deeplearning approaches have been evolved to address this problem. This paper proposes the Slide-Detecttechnique which is a Deep Neural Networks (DNN) model based on Convolutional Neural Networks (CNNs)that is trained to diagnose lung infiltration with Area Under Curve (AUC) up to 91.47%, accuracy of 93.85%and relatively low computational resources.
作者简介
Corresponding author:Ahmed E.Mohamed(E-mail:Ahmed.E.Mohamed@eng1.cu.edu.eg,ORCID:0000-0002-0030-8780).
