AbstractIntroduction: Ultrasound (US) is a nonionizing radiation capable of real time imaging at low cost. Its most attractive application is quantitative tissue characterization with the objective of differentiating normal tissues from diseased tissues. In this study, an automated method using singular spectrum analysis (SSA) to estimate the mean scatterer space (MSS) of US signals is proposed. Methods Entropy was used to determine the optimal number of components for the SSA. Subsequently, this number was compared with the results using a fixed number of components. A method based on the spectrum of the original signal was also used for comparison. The method was evaluated by using 24,000 simulated US signals, i.e., echoes and jitters backscattered from samples with different ratios of regular-to-irregular structure, as well as with 152 signals obtained from a phantom made of nylon wires. Results For the simulated signals, the proposed method for estimating the MSS presented results similar to the other methods that were tested. However, the magnitude-of-the-spectrum method loses the phase information, and hence, does not allow the characterization of irregular structures. For the signals recorded from the phantom, the methods using SSA and entropy achieved better results. Conclusion In this study, the combination of SSA with entropy to estimate the MSS of a periodic or quasi-periodic medium was proposed. The proposed method achieved similar or better results compared with two other methods found in the scientific literature. The novelty of the proposed method is the application of entropy as a quantitative criterion for selecting the SSA periodic components, allowing it to become independent of heuristic criteria.

Abstract Introduction Various signal-processing techniques have been proposed to extract quantitative information about internal structures of tissues from the original radio frequency (RF) signals instead of an ultrasound image. The quantifiable parameter called the mean scatterer spacing (MSS) can be useful to detect changes in the quasi-periodic microstructure of tissues such as the liver or the spleen, using ultrasonic signals. Methods We evaluate and compare the performance of three classic methods of spectral estimation to calculate the MSS without operator intervention: Tufts-Kumaresan, SAC (Spectral Autocorrelation) and MUSIC (MUltiple SIgnal Classification). Initially the evaluations were performed with 10,000 signals simulated from a model in which the variables of interest are controlled, and then, real signals from sponge phantoms were used. Results For the simulated signals, the performance of all three methods decreased with increasing Ad or jitter levels. For the sponges, none of the methods accurately estimated the pore size. Conclusion For the simulated signals, Tufts-Kumaresan had the lowest performance, whereas SAC and MUSIC had similar results. For sponges, only Tufts-Kumaresan was able to detect the increase in the size of the pores of the sponge, although most often, it estimated sizes larger than expected.

A interpretação da imagem ultrassônica, por ocorrer de modo visual e qualitativa, traz uma variação inter e intra-observador importante. A adoção de métodos quantitativos é uma forma de diminuir esta dependência. Entre tais métodos está a quantificação do espaçamento médio entre espalhadores (Mean Scatterer Spacing - MSS), que pode ser útil para detectar mudanças na microestrutura quasi-periódica de tecidos como o hepático ou o esplênico. Neste trabalho foram avaliados três métodos clássicos de estimação espectral para cálculo do MSS (sem intervenção do operador): BURG, WIENER e MUSIC. O intuito é comparar suas potencialidades para a estimação automática de espaçamento médio de espalhadores ultrassônicos. Inicialmente as avaliações foram realizadas com 10.000 sinais simulados a partir de um modelo em que se tem controle das variáveis de interesse, e em seguida foram utilizados sinais reais de phantoms de fios de nylon imersos em água. O método de BURG não conseguiu estimar adequadamente o espaçamento em sinais de phantom, tendo apresentado resultados equivalentes aos outros métodos deste trabalho somente para sinais simulados. O método de WIENER para os sinais simulados apresentou resultados de menor percentual de acerto, ficando em segundo lugar, para os sinais dos phantoms. O método de subespaço MUSIC apresentou melhor desempenho global em relação a BURG e WIENER, com resultados de 100% de acerto para o phantom de fio de nylon de 1,2 mm e 91,45% para 0,8 mm considerando uma janela de acerto de 10%.

The interpretation of ultrasound imaging is essentially visual and qualitative, so there are important inter and intra-observer variations. Quantification methods aim at decreasing this dependency. Among those, the quantification of the Mean Scatterer Spacing (MSS) can be useful to detect changes in the microstructure of quasi-periodic tissues, such as liver or spleen. This study evaluated the following methods of spectral estimation for calculating the MSS (without requiring operator intervention): BURG, WIENER and MUSIC. The aim is to compare their potential for automatic estimation of MSS from ultrasonic scattering signals. Initially, the evaluation has been carried out using 10,000 simulated signals, with the aim of studying the behavior of the methods using a model in which the variables of interest can be controlled. Then, the methods have been applied to real signals of nylon phantoms immersed in water. The BURG method could not estimate the spacing of US phantom signals, presenting results similar to the other methods only for simulated signals. The WIENER method for the simulated signals was in second place in terms of percentage of success, when considering signals from the phantoms. The subspace method MUSIC had the best performance from all three methods.