Intelligent Diagnosis of Covid-19 Based on CNN-PNN

Abstract

Today the whole world suffers and fears the epidemic of the Coronavirus and the developed waves in it, as we have now reached the fourth wave, and this is a serious matter. Where the statistics of the Coronavirus in the current data showed that 213 countries are affected by this epidemic, and about 6 millions of deaths are recorded. This virus spreads rapidly, and patients mainly suffer from breathing. The patient who suffers from pre-existing health problems will be more likely to contract this disease, so there was an urgent need for artificial intelligence to enter to quickly detect this virus, so the world turned to deep learning, which is one of the most powerful methods and techniques for classification because of its use of Bayas Rule, where there is no possibility of error. This paper proposes CNN (Convolutional Neural Networks) and PNN (Proprestitc Neural Networks) mixed tomography scanning model to classify Covid-19 images, the proposed network called the CNN-PNN model. The CNN-PNN model can use CNN to compute the dependency and continuity features of the output of the middle layer of the PNN model, and correlate the properties of these middle levels with the final full network to predict the classification.