An efficient adaptive thresholding function optimized by a cuckoo search algorithm for a despeckling filter of medical ultrasound images

In this research paper, we propose an efficient adaptive thresholding function optimized by a cuckoo search algorithm for despeckling filters to avoid boundaries created by more than a few current despeckling filters. The speckle noise contamination caused at some points in ultrasound picture acquisition systems compromises the level of its visuality, which presents a diagnostic challenge for medical doctors. Therefore, to enhance the visual quality, despeckling filters are normally employed in the processing of such pictures. However, several disadvantages have developed within current despeckling filters, which have discouraged the utilization of modern despeckling filters for minimizing impacts from speckle noise. The proposed despeckling filter was developed through a combination of an adaptive thresholding function and a cuckoo search algorithm. Specifically, the aforementioned cuckoo search algorithm optimizes the coefficients of the thresholding function. Additionally, the proposed approach it can be used in a no-reference as well as in a full-reference image quality assessment which the assessment of PSNR, SSIM, MSE and MAE values was used to evaluate the proposed despeckling filter, as well as others, in medical ultrasound images. The findings from the research show that the visual outcome acquired by the proposed filter is inferior to those of filters with a despeckling effect. In addition, the numerical results revealed that the proposed despeckling filter was effective in creating ultrasound imageries for clinical use.