COVID-19 Detection using Image Modality: A Review

Kukreja, P.; Gupta, D.

COVID-19 or Coronavirus is a pandemic that has spread and has affected many people around the world. It is important that the disease is identified at an early stage only so that an infected individual can be isolated. RT-PCR (Reverse transcription PCR testing) is the tool used to analyze and detect viral RNA therefore is used for the detection of SARS-COV-2 but it is very time-consuming. This paper discusses various image modalities like CT-Scan, X-rays, Ultrasound which are used for detection. Deep learning methods have been demonstrated. to be a strong weapon in the arsenal used by clinicians. Insights of various data sets for training the network and the performance measures used by the researchers are highlighted. In this paper, a complete survey of techniques of Deep Learning for the diagnosis of COVID-19 is discussed using various types of medical imaging modalities. Results indicate that imaging characteristics can play an important role in the detection of COVID-19. Finally, we conclude by discussing the challenges related to the use of deep learning methods for identification of COVID-19 and probable future trends in this field of study. © 2021 IEEE.