EVALUATING THE PERFORMANCE OF INVARIANT MOMENTS AND SHAPE FEATURES TO CLASSIFY NORMAL AND COVID SUBJECTS
Keywords:
COVID-19, 2D shape analysis, Invariant moments, Shape featureAbstract
Alteration in lung shape and texture are considered as the significant pathological cores for analysing the severity of COVID-19 disease. The main objective of this analysis is to estimate the shape deformation of lung in CT images due to the incidence of COVID-19. Totally 176 samples are considered for the analysis. Comparison between invariant moments and shape features is carried out to analyse the geometric variations of lungs. The extracted features are used to classify normal and COVID subjects using neural network classifier. From the results, it is observed that feature values obtained using invariant moments are found to be statistically significant (p<0.05) compared to shape features. Inter (normal and COVID) subjects mean and STD value, for invariant moments M5 feature is (0.76±0.07, 0.02±0.10) and for shape features extent feature is (0.61±0.09, 0.54±0.14). Invariant moments obtained 0.993 of accuracy and 0.992 of F1 score using neural network classifier. In this study, it is found that invariant moments perform better in comparison with shape features demonstrating significant differences between normal and COVID CT scan images.
References
Al-Quteimat, Osama, M., and Amer Mustafa Amer. (2020). The Impact of the COVID-19 Pandemic on Cancer Patients. American J. of Clinical Oncology., 43(6):452-455
Asodekar, Bhagyashri H., Sonal, A. Gore, and A.D. Thakare. (2019). Brain Tumor Analysis Based on Shape Features of MRI Using Machine Learning. 2019 5th International Conference On Computing, Communication, Control And Automation (ICCUBEA)., 19-21 doi.org/10.1109/iccubea 47591. 2019.9129512.
Dmitriev, E.A., A.A. Borodinov, A.I. Maksimov and S.A. Rychazhkov. (2019). Automatic Detection of Constructions Using Binary Image Segmentation Algorithms Information Technology and Nanotechnology., https://doi.org/10.18287/1613-0073-2019-2391-264-268.
Dogantekin, Esin., Mustafa Yilmaz., Akif Dogantekin., Engin Avci., and Abdulkadir Sengur. (2008). A Robust Technique Based on Invariant Moments-ANFIS for Recognition of Human Parasite Eggs in Microscopic Images. Expert Systems with Applications. https://doi.org/10.1016/j.eswa.2007.07.020.
Faul, Franz., Edgar Erdfelder., Axel Buchner., and Albert-Georg Lang. (2009). Statistical Power Analyses Using G*Power 3.1: Tests for Correlation and Regression Analyses. Behavior Research Methods., 41(4):1,149-60.
George, D.N., Jehlol, H.B., and Oleiwi, A.S.A. (2015). Brain tumor detection using shape features and machine learning algorithms. Int. J. of Advanced Research in Computer Science and Software Eng., 5(10):454-459.
Huang, Zhihu and Jinsong Leng. (2010). Analysis of Hu’s Moment Invariants on Image Scaling and Rotation. 2010 2nd International Conference on Computer Engineering and Technology., https://doi.org/10.1109/iccet.2010. 5485542.
Li, Meng. (2020). Chest CT Features and Their Role in COVID-19. Radiology of Infectious Diseases., 7(2):51-54
Ma Jun., Ge Cheng., Wang Yixin., An Xingle., Gao Jiantao., Yu Ziqi., and He Jian. (2020). COVID-19 CT Lung and Infection Segmentation Dataset (Version Verson 1.0) [Data set]. Zenodo. http://doi.org/10.5281/zenodo. 3757476.
Mingqiang, Yang., Kpalma Kidiyo., and Ronsin Joseph. (2008). A Survey of Shape Feature Extraction Techniques. Pattern Recognition Techniques, Technology and Applications., https://doi.org/10.5772/6237.
Öztürk, Şaban., Umut Özkaya., and Mücahid Barstuğan. (2021). “Classification of Coronavirus (COVID‐19) from X‐ray and CT Images Using Shrunken Features. Int. J. of Imaging Systems and Technol., 31(1):5-15
Ramteke, R.J. (2010). Invariant Moments Based Feature Extraction for Handwritten Devanagari Vowels Recognition. Int. J. of Computer Applications., 1(18):975-8,887
