الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 94 |  |  |
| | | from | 94 | | |
Abstract
Respiratory infections are a confusing and time-consuming task of constantly looking at clinical pictures of patients. Therefore, there is a need to develop and improve the respiratory case prediction model as soon as possible to control the spread of disease. Deep learning (DL) makes it possible to discover a virus such as COVID and Viral Pneumonia which can be effectively detected using classification tools as Convolutional Neural Network (CNN). Mel Frequency Cepstral Coefficients (MFCC) is a common and effective classification tool. A Proposed MFCC - CNN’s learning model is used to speed up the prediction process and assists medical professional’s classification. MFCC is used to extract image features that are related to the presence of respiratory virus or not. Prediction is based on CNN. This makes virus detection more robust, easier and faster and less time-consuming process with more accurate results reducing the spread of the virus and saves lives. Experimental results show that using a Computed Tomography (CT) image converted to Mel-frequency cepstral spectrogram as an input to CNN can perform better results; with the validation data that include 99.08% accuracy for appropriate COVID and Viral Pneumonia categories and images with the healthy labels (no respiratory virus present). Thus, it can probably be used to detect in CT images the presence of respiratory virus. The work here provides evidence of the idea that high accuracy can be achieved with a trusted dataset, which can have a significant impact on this area. A new proposed combined network that consists of CNN and Probabilistic Neural Network (PNN) is also introduced in this thesis which provides a classification of COVID images from NON COVID images with an accuracy of 100%.