IN-VIVO ULTRASONIC PARAMETRIC IMAGING OF CAROTID ATHEROSCLEROTIC PLAQUE BY VIDEODENSITOMETRIC TECHNIQUE

Extensive experimental and clinical data show that the ultrasonic imag e conveys information on the biochemical composition of the atheroscle rotic plaque, ie, the relative content of lipids (hypoechoic), fibrous tissue (hyperechoic), and calcific deposits (very echogenic with shad owing). A more dishomogeneous echo structure of the plaque is also mor e often associated with clinically complicated carotid plaques. To dat e, however, the assessment of plaque density and homogeneity by transc utaneous B-mode imaging remains subjective and qualitative. The aim of this study was to assess whether plaque echodensity and homogeneity m ight be established on a more objective and quantitative basis by desc ription of the spatial distribution of echo amplitude (referred to as tissue texture) applied to digitized images, obtained with commerciall y available B-mode transcutaneous imaging systems. A total of 47 B-mod e images derived from echotomographic studies in 10 patients were digi tized off line. For each region of interest, a set of first-order (mea n gray level, standard deviation, skewness, kurtosis: mathematical des criptors of the shape of the frequency distribution of gray-level hist ogram) and of second-order (entropy, angular moment: mathematical desc riptors of the spatial distribution of gray levels within the region o f interest) textural parameters were evaluated. The visual, concordant reading by two independent, experienced observers assigned the plaque s on the basis of qualitatively assessed echodensity to three groups: ''soft'' (n = 18), ''fibrotic'' (n = 20), ''calcific'' (n = 9). Regard ing spatial gray-level distribution, 46 plaques would be separated int o ''homogeneous'' (n = 17) and ''dishomogeneous'' (n = 29). On digitiz ed images, the normalized mean gray level was the most effective first -order textural parameter for distinguishing soft (24.2 +/- 12.4 arbit rary units in a zero to 255 scale) from fibrotic (64.5 +/- 16.4) and c alcific plaques (125.3 +/- 24.5), P < 0.01 for all intergroup differen ces. ''Homogeneous'' plaques were separated from ''heterogeneous'' one s on the basis of entropy (5 +/- 1 vs 7.9 +/- 9.7; P < 0.01), whereas the values of angular second moment overlapped (1.542E-3 + 1.334E-3 vs 5.181E-4 +/-2.5615E-4, P = ns). In conclusion, quantitative texture a nalysis of ultrasonic images derived from transcutaneous, high-resolut ion, commercially available B-scan systems is feasible in man and prov ides a quantitative operator-independent assessment of plaque echodens ity and homogeneity.