Noise filtering in ultrasound images using Gravitational Search Algorithm

Ultrasound Imaging plays a great role in the real-time medical imaging for the visualization and diagnostic purposes. Such a coherent imaging system is usually influenced by the presence of the speckle noise. Consequently, image denoising emerges as a fundamental trend to decline the effect of such interferences. This paper demonstrates an adaptive noise filtering method in ultrasound images. Gravitational Search Algorithm (GSA) is exploited for optimization of the filter coefficients of 2D FIR filters. The mean square error (MSE) and signal to noise ratio (SNR) metrics are used for assessment of the gained results. Based on the quantitative comparisons, the proposed approach outperforms the Gaussian, and median filters.