Pseudo Zernike Moment and Deep Stacked Sparse Autoencoder for COVID-19 Diagnosis

(Aim) COVID-19 is an ongoing infectious disease. It has caused more than 107.45 m confirmed cases and 2.35 deaths till 11/Feb/2021. Traditional computer vision methods have achieved promising results on the automatic smart diagnosis. (Method) This study aims to propose a novel deep learning method that can obtain better performance. We use pseudo-Zernike moment (PZM), derived from Zernike moment, as extracted features. Two settings are introducing: (i) image plane over unit circle; (ii) inside circle. Afterward, we deep-stacked sparse autoencoder (DSSAE) classifier. Besides, multiple-way data augmentation chosen overcome overfitting. The based Gaussian noise, salt-and-pepper speckle horizontal vertical shear, rotation, Gamma correction, random translation scaling. (Results) 10 runs of 10-fold cross validation shows our PZM-DSSAE achieves sensitivity 92.06% ± 1.54%, specificity 92.56% 1.06%, precision 92.53% 1.03%, accuracy 92.31% 1.08%. Its F1 score, MCC, FMI arrive at 92.29% ±1.10%, 84.64% 2.15%, 1.10%, respectively. AUC model 0.9576. (Conclusion) demonstrate “image circle” get circle.” this proposed eight state-of-the-art approaches.