Identification and Classification of COVID-19 using Radiological Images

  • Unique Paper ID: 155873
  • Volume: 9
  • Issue: 2
  • PageNo: 277-288
  • Abstract:
  • The outbreak of the COVID-19 has put the whole world in an unpleasant situation. There is a surge in number of people affected by this disease day by day. Many people have lost their lives to this pandemic disease. At present, the detection of corona virus disease 2019 (COVID-19) is one of the main challenges in the world. This paper implements an artificial intelligence technique based on a deep Convolution Neural Network (CNN) models to detect COVID-19 automatically using real-world datasets. A qualitative analysis is performed to inspect the decisions made by CNN model using a technique known as Class Activation Mapping (CAM). Using CAM, the activations contributed most to the decision of CNN models can be mapped to visualize the most discriminating regions on the input image. Chest CT scan plays a very crucial role in determining the severity of the disease in the infected patients. An AI based image-assisted system is formulated to extract COVID-19 infected sections from lung CT scans (axial view).Threshold filter is applied to extract the lung region and eliminate possible artifacts. Then image segmentation is performed using graph-cut method to extract the region of interest (lung area and infected regions). The binary images of region of interest are then employed to identify the pixel ratio between the lung area and infection sections to calculate the severity level of infection. The classification of COVID-19 infected cases based on severity assist the pulmonologist not only to detect but also to help plan the treatment process efficiently.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 9
  • Issue: 2
  • PageNo: 277-288

Identification and Classification of COVID-19 using Radiological Images

Related Articles