Deep CNN Model for COVID-19 Infection Detection


  • Fatin Zulaikha Zulkifli FKEE-BEJ
  • Audrey Huong


COVID-19, X-Ray, AlexNet


This paper focuses on the use of deep learning model in diagnosing Coronavirus disease (COVID-19) lung infection. For this purpose, this work employed pretrained AlexNet model for two-class classification problem (i.e., “COVID” and “NONCOVID”). It is also the objective of this work to provide quantitative assessment of the classifier’s reliability based on the performance metrics. We used 80 chest X-ray images (i.e., 40 images of COVID-19 infected patients and 40 of healthy individuals) in this preliminary study. This work evaluates the performance of the trained classifier and observed relatively good performance in the prediction of validation data using outcomes from three best training runs. The mean accuracy, specificity, sensitivity, precision, error rate and false positive error are given by 87.5%, 83.3%, 91.67%, 85%, 12.5% and 16.67% respectively. It is concluded that the trained model is able to perform relatively well in classifying COVID-19 X ray images. In future, more data could be used in the training to increase accuracy of the model. Further possible attempt also includes the use of chest images from other imaging technologies in the training to give higher variability on the choice of dataset using which a deep learning model can train on.




How to Cite

Zulkifli, F. Z. ., & Huong, A. (2021). Deep CNN Model for COVID-19 Infection Detection. Evolution in Electrical and Electronic Engineering, 2(1), 213–220. Retrieved from