Despeckling of Prostate Ultrasound Images using Computational Optimization Approaches

J.Ramesh*, R.Manavalan

Prostate cancer is a serious disease found only in the reproductive system of men. Prostate cancer is caused by the continuous deposition of protein in prostate gland. Ultrasound (US) imaging is an important modality in the early recognition and analysis of prostate cancer. Generally, medical ultrasound images are affected by speckle noise during the acquisition process. Speckle noise produces a granular pattern in the US images which leads to failure in identifying the exact infected area of prostate gland. It is necessary to eliminate the speckle from the US image without affecting edges and content of the information. In this article, various metaheuristic optimization based algorithms are proposed for despeckling the prostate US images. The performance of the proposed algorithms is analyzed by standard parameters such as MSE and PSNR. Further, the algorithms are tested using time and space complexities. The experimental results clearly shows that one of the methods Ant Colony Optimization (ACO) is superior to other methods in despeckling as well as conserving the edges.

Volume 11 | 02-Special Issue

Pages: 142-156