Publish In |
International Journal of Electrical, Electronics and Data Communication (IJEEDC)-IJEEDC |
Journal Home Volume Issue |
||||||||
Issue |
Volume-10,Issue-7 ( Jul, 2022 ) | |||||||||
Paper Title |
Two Approaches for Lung Covid-19 Infection Classification on Ct Image | |||||||||
Author Name |
Mustafa Abdullah, Adnan Saher Mohammed | |||||||||
Affilition |
||||||||||
Pages |
1-9 | |||||||||
Abstract |
Abstract - In early 2020, an existential health catastrophe resulted in the worldwide spread of the Coronavirus 2019 (COVID-19). Traditional healthcare strategies for treating COVID-19 may be improved with the use of automated CT imaging identification of lung infections like COVID-19. COVID-19 infections can only be diagnosed by using CT imaging. CT images are obtained from two databases: CC-CCII and Mos MedData, which are used for improvement purposes of lung COVID-19 abnormality classification on CT scans. The significant heterogeneity and low density between infected and normal tissues make it difficult to identify infection using CT scans. The earlier COVID-19 infection predicate is used to perform a range of diagnostic activities, which helps identify pathological (COVID-19 infections) as well as enhances CT diagnostic reporting accuracy. Finally, modified machine learning models (CNN and SVM) were used to classify CT images as COIV-19 infection or normal. Analysis of experimental and clinical data shows that the proposed methodologies for examining the variability of the internal geometric characteristics (classification) of the lung and COVID-19 infections in images are effective. The modified systems showed that the accuracy of SVM with combined LBP with HOG is 98% and modified CNN of 98%. Keywords - Image Processing, COVID-19 Infections, Classification, Feature Extraction, CT Image, Machine Learning. | |||||||||
View Paper |