Title :
Rician inverse Gaussian model of scattering in ultrasound contrast media
Author :
Slavik, Vladimir ; Kolar, R. ; Jirik, Radovan ; Harabis, Vratislav
Author_Institution :
Dept. of Biomed. Eng., Brno Univ. of Technol., Brno, Czech Republic
Abstract :
A statistical analysis of 1st and 2nd harmonic echoes backscattered from contrast media using Rician inverse Gaussian (RiIG) envelope model is described. Knowledge of statistics of the scattered signal may contribute to perfusion analysis. A range of ultrasound contrast agent (UCA) concentrations is examined. The RiIG distribution was fitted to image histograms, using Matlab Global optimization toolbox. Influence of simulated attenuation was also tested. Analysis of resulting RiIG distribution parameters shows that while its parameters α and β do not follow any properties of the changing concentration, parameter δ shows logarithmic dependence on concentration in range of 3.125 - 50 mg/l. Taking into account simulated attenuation, we observed that the range and shape of the dependency remain stable, although the actual range of parameter δ decreases with increasing attenuation. This analysis contributes to our previous results, where the Nakagami distribution showed similar dependence of the shape parameter on UCA concentration in a smaller, partially overlapping, concentration range.
Keywords :
Gaussian processes; biomedical ultrasonics; inverse problems; mathematics computing; medical image processing; optimisation; statistical analysis; ultrasonic imaging; 1st harmonic echo backscattering; 2nd harmonic echo backscattering; Matlab Global optimization toolbox; Nakagami distribution; RiIG distribution parameters; Rician inverse Gaussian model; image histograms; logarithmic dependence; perfusion analysis; scattered signal; statistical analysis; ultrasound contrast agent concentrations; ultrasound contrast media scattering; Attenuation; Harmonic analysis; Histograms; Mathematical model; Nakagami distribution; Rician channels; Ultrasonic imaging; Attenuation; Concentration estimation; Histogram fitting; Rician inverse gaussian distribution;
Conference_Titel :
Ultrasonics Symposium (IUS), 2013 IEEE International
Conference_Location :
Prague
Print_ISBN :
978-1-4673-5684-8
DOI :
10.1109/ULTSYM.2013.0082