Author(s): Dr.Sachi Mohanty
Automated tissue characterization of interstitial lung disease is one of the most important elements of Computer Aided Disease system. The problem remains challenging even though there has been much research in this area. While deep learning has produced brilliant success in image applications over the past few years, the majority of training is with sub-optimal parameters, requiring unnecessary long training time, setting up hyper parameters. In this paper we explore the classification of lung tissue pattern affected with interstitial lung disease (ILD) in high resolution computed tomography (HRCT) scans and evaluate different CNN architectures with and without transfer learning and we examine the effect of using Cyclical learning rates for faster convergence and the hyper-parameters tuning and data augmentation using Med Gift dataset.
Share this article