الكشف عن مرض الكوفيد- 19 باستخدام تقنيات الذكاء الاصطناعي و الصور الطبية الشعاعية

Abstract EN

This study aims to emphasize the importance of screening COVID-19 disease using Artificial Intelligence and radiological images (chest X-ray and CT images), by extracting the features from the images and classifying them using standard classification algorithms.

Finally, we combined the classification task with the retrieving task for its importance in facilitating comparison and diagnosis .We proposed a novel method for extracting features by combines both Handcraft and Automatic feature extraction methods, and we got better results from using one of the two methods.

Our proposed method combines three levels of feature extraction, the Local Binary Pattern for extracting low-level feature extraction, Bag of Visual Word for mid-level, and the AlexNet, which relies on deep learning for high-level features.

Finally, we weighted these features according to their importance, and we used Principal Component Analysis for dimension reduction.