Image pseudo-coloring for the analysis of COVID-19 using chest X-ray images

An assessment of COVID-19 used laboratories for pathogen testing and non-laboratory test for thorax imaging. Chest X-ray is the initial modality for screening, widespread availability, and low cost. Therefore, a Chest X-ray is an effective diagnostic tool, for example, detecting fracture, given the ground-glass characteristic of COVID-19. However, ground glass is difficult to identify. Pseudo-coloring helps identify ground-glass images because the human eye easily distinguishes many types of colors. The parameters used in data collection of chest X-rays were the COVID-19 patients that were confirmed by PT-PCR test and approved suspected by specialist doctors and non-COVID-19 patients. We evaluated the chest X-ray of 50 adults suspected of COVID-19 pneumonia and 22 patients with non-COVID diagnoses. The patients were between 20 and 49 years of age. Multix Fusion performed the radiological images for X-ray Images in Andalas University Hospital. All coding was carried out with MATLAB 2017b software on the operating system macOS. The initial abnormalities suggesting COVID-19 pneumonia on a chest x-ray are loss of the standard black appearance in the lung. This image is seen as increased whiteness because of increased density but insufficient to obscure lung marking, giving a ground-glass appearance. PSNR non-COVID and COVID-19 have compared the alpha level in the two-sample t-test assuming equal variances window (α=0.05). The result of p-value<0.05 shows the data differ significantly. Then, we applied the histogram equalization to the image, and the p-value=0.474, so the non-COVID-19 and COVID-19 have PSNR (dB) equal. The results show that pseudo-coloring give different color density to image and can help easily distinguishes the ground glass opacities. © 2023 Author(s).