Roopa Kumari
Department of Computer Science, Gurukula Kangri Vishwvidyalya, Haridwar, Uttarakhand, India.

Neena Gupta
Department of Computer Science, Gurukula Kangri Vishwvidyalya, Haridwar, Uttarakhand, India.

Narender Kumar
Department of Computer Science, Doon University, Dehradun, Uttarakhand, India.

DOI https://doi.org/10.33889/IJMEMS.2021.6.5.076

  ;

Abstract

Covid-19, a disease that originated in the Chinese city of Wuhan, has spread across almost the entire globe. Pneumonia, which infects the lungs, is one of the symptoms of this disease. In the past X-ray images were used to segment various diseases such as pneumonia, tuberculosis, or lung cancer. Recent studies showed that Covid-19 affects the lungs. As a result, an X-ray imaging could help to detect and diagnose Covid-19 infection. This study presents a novel hybrid algorithm (CHDPSOK) for segmenting a Covid-19 infected X-ray image. To find Covid-19 contamination in the lungs, we use a segmentation-based approach using K-means and Dynamic PSO algorithm. In the present paper, segmentation of infected regions in the X-ray image uses a cumulative histogram to initialize the population of the PSO algorithm. In a dynamic PSO algorithm, the velocity of the particle changes dynamically which is useful to avoid the local minima. K-means is used to change the position of the particle dynamically for better convergence. To validate the segmentation performance of our algorithm, we used the Kaggle dataset in our experiments. The performance of the proposed method is analyzed both qualitatively and quantitatively. The results explicitly demonstrate the outperformance of the proposed algorithm.

Keywords- Covid-19, Chest X-ray, Image segmentation, Particle swarm optimization, K-means.

Citation

Kumari, R., Gupta, N., & Kumar, N. (2021). Segmentation of Covid-19 Affected X-Ray Image using K-means and DPSO Algorithm. International Journal of Mathematical, Engineering and Management Sciences, 6(5), 1255-1275. https://doi.org/10.33889/IJMEMS.2021.6.5.076.