Smith S W, Trahey G E, Hubbard S M, Wagner R F
Center for Devices and Radiological Health, Food and Drug Administration, Rockville, MD 20857.
Ultrason Imaging. 1988 Jan;10(1):29-51. doi: 10.1177/016173468801000103.
In recent years, analysis of the second order statistics of ultrasound speckle has led to accurate prediction and measurements of the average speckle size in the transducer focal zone. In this paper, that work has been extended to the average speckle size as determined by the normalized autocovariance in the presence of transducer phase aberrations. In general, a phase aberration causes a narrowing of the main lobe of the normalized autocovariance in the lateral direction. However, the lateral speckle autocovariance also showed significant side lobes in the presence of phase aberrations, indicating that individual speckles in a region of interest are not independent but are correlated so that less information is present for the task of signal detection when a transducer phase aberration exists. The same evidence of correlated speckle was found in the near field of a transducer in the region of fine speckle texture. This explanation satisfies the quandary of poor detectability in the near field region where the speckle is fine but the lateral resolution is quite degraded. The axial speckle in the presence of phase aberrations showed a small increase in main lobe widths and no evidence of side lobes. Beginning in 1978, the analysis of the second order statistics of speckle images for the purpose of spatial compounding led to accurate measurement and prediction of the cross-correlation curve as a function of transducer aperture translation for purposes of spatial compounding. In this paper, that work has been extended to the presence of transducer phase aberrations. The existence of transducer phase aberrations causes significant increases in the rate of decorrelation of speckle interference patterns as a transducer is translated. This indicates that spatial compounding will result in quite significant improvements in area-wise SNR and low contrast lesion detection for the case of severe random aberrators or focal point errors.
