Classification and Region Analysis of Covid-19 Infection Using Lung Ct Images and Deep Convolutional Neural Networks

Abstract

COVID-19 is a global health problem that requires efficient diagnostic techniques. However, the accurate analysis of COVID-19 is challenging due to the contrast and texture variation in infected regions. Therefore, a new two-stage deep CNNs based framework is proposed for segmenting COVID-19 infected regions in Lung CT images. In the first stage, COVID-19 specific features are extracted using a two-level DWT and then classified using a new CNN block-based CoV-CTNet. While in the second stage, the infectious CT images are analyzed using a novel CoV-RASeg segmentation CNN. These models systematically implement region and edge operations to learn COVID-19 infection properties related to region-homogeneity, texture variation, and boundaries. Moreover, static attention and transfer learning are employed to effectively learn the lesion regions and improve model convergence, respectively.  The performance of the proposed CoV-CTNet is evaluated using accuracy (98.80%) and MCC (0.98), while the proposed SA-CoV-RASeg achieves an IoU score of 98.75%.

Note:
Funding Information: This work was conducted with the support of the research grant of National Research Foundation of Korea (2017R1A2B2005065). This study was also supported by the PIEAS IT endowment fund and HEC indigenous Scholarship under the Pakistan Higher Education Commission (HEC).

Declaration of Interests: Authors declared no conflict of interest.

Keywords: COVID-19, CT image, Convolutional Neural Networks, Transfer learning, Classification, and Segmentation