HOG + CNN Net: Diagnosing COVID-19 and Pneumonia by Deep Neural Network from Chest X-Ray Images

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Diseases

Background:

• Coronavirus disease 2019 (COVID-19) is a highly contagious respiratory illness caused by SARS-CoV-2, leading to a global pandemic.
• The rapid spread of COVID-19 necessitates effective and timely diagnostic tools for disease control.
• Medical radiography, including chest X-rays, is a crucial modality for identifying pneumonia associated with COVID-19.

Purpose of the Study:

• To develop and evaluate an automated system for the early detection and classification of COVID-19 from frontal chest X-ray images.
• To differentiate between COVID-19 positive, pneumonia positive, and normal chest X-ray cases using advanced image analysis techniques.
• To establish the efficacy of a proposed deep learning model as an early warning system for COVID-19.

Main Methods:

• Utilized a combination of Histogram of Oriented Gradients (HOG) features and a deep convolutional neural network (CNN).
• The model was trained and tested on frontal chest X-ray images.
• Image classification was performed to categorize images into three classes: COVID-19 positive, pneumonia positive, and normal.

Main Results:

• The proposed HOG and deep CNN-based model demonstrated effective performance in detecting abnormalities in chest X-rays.
• The system achieved high accuracy in classifying chest X-ray images into the specified categories.
• The model proved capable of serving as an effective tool for early detection of COVID-19.

Conclusions:

• The developed deep learning model shows significant promise for the early and accurate detection of COVID-19 using chest X-rays.
• This approach can contribute to public health efforts by enabling faster identification of infected individuals.
• The system's effectiveness in classification supports its potential role in managing the ongoing COVID-19 pandemic.