Speckle Noise Reduction in Ultrasound Images for Improving the Metrological Evaluation of Biomedical Applications: An Overview

Abstract

In recent years, many studies have examined filters for eliminating or reducing speckle noise, which is inherent to ultrasound images, in order to improve the metrological evaluation of their biomedical applications. In the case of medical ultrasound images, said noise can produce uncertainty in the diagnosis because details, such as limits and edges, should be preserved. Most algorithms can eliminate speckle noise, but they do not consider the conservation of these details. This paper describes, in detail, 27 techniques that mainly focus on the smoothing or elimination of speckle noise in medical ultrasound images. The aim of this study is to highlight the importance of improving said smoothing and elimination, which are directly related to several processes (such as the detection of regions of interest) described in other articles examined in this study. Furthermore, the description of this collection of techniques facilitates the implementation of evaluations and research with a more specific scope. This study initially covers several classical methods, such as spatial filtering, diffusion filtering, and wavelet filtering. Subsequently, it describes recent techniques in the field of machine learning focused on deep learning, which are not yet well known but greatly relevant, along with some modern and hybrid models in the field of speckle-noise filtering. Finally, five Full-Reference (FR) distortion metrics, common in filter evaluation processes, are detailed along with a compensation methodology between FR and Non-Reference (NR) metrics, which can generate greater certainty in the classification of the filters by considering the information of their behavior in terms of perceptual quality provided by NR metrics. INDEX TERMS Diffusion filtering, image pre-processing, metrological evaluation, spatial filtering, speckle noise, ultrasound images, wavelet filtering.

High-Resolution Ultrasound Imaging Enabled by Random Interference and Joint Image Reconstruction

“…We can express an ultrasound echo signal received at the i th transducer element as where is the received RF signal, is the transmission matrix, and is a vectorized image, as observed by the i th array element. According to the principle of superposition [ 9 ], the received ultrasound echo signal is the sum of individual signals reflected off the scatterers . Our goal is to find an estimate of the image given the RF signals and transmission matrices for , this can be accomplished using CS theory [ 27 , 28 , 29 ].…”

“…In Figure 4, we show how RF signals from a single pulse-echo transmission can be rearranged into 18 different sets of subarrays signals. Then, we use subarray signals to reconstruct a subarray image using Equations (5) and (9). In Figure 5a, we show a diagram of a tissue-mimicking phantom that represents the ROI.…”

“…In the past, different techniques have been used to enhance the spatial and contrast resolution of ultrasound systems [9,10]. For instance, in [11,12], speckle reduction techniques using statistical models have been proposed.…”

Optimized Hybrid Model for Gaussian Noise Reduction images

“…Salt and Pepper noise, on the other hand, is an impulse noise where it only has two distinct values: high or low. The values of the noisy pixels are either set too high or too low resulting in salt and pepper noise [16].…”

Towards to Optimal Wavelet Denoising Scheme—A Novel Spatial and Volumetric Mapping of Wavelet-Based Biomedical Data Smoothing

“…Here, we recognize the local filters, considering a certain finite representative pixel neighbourhood. Based on the local statistical features they are capable of modifying the pixels values [ 24 , 25 ]. Such principle is utilized in average, median, Gaussian filter or bilateral filter.…”