Ultrasonic Backscatter and Attenuation (11–27 MHz) Variation with Collagen Fiber Distribution in Ex Vivo Human Dermis

Abstract

This ex vivo study explores the relationship of ultrasonic attenuation and backscatter to dermal microarchitecture by comparing ultrasonic measurements of these parameters (11–27 MHz) to a microscopic analysis of three parameters describing the collagen distribution (mean thickness and spacing of collagen bundles along the insonification direction and the percent area occupied by collagen). Skin samples (N = 31) were obtained from patients undergoing breast or abdominal reduction surgery. Radio-frequency (rf) signals were acquired in a B-scan format using an ultrasound system developed for skin imaging (Ultrasons Technologies, Tours, France). Ultrasonic data were analyzed to calculate average integrated backscatter (IBS in dB) and frequency dependence of backscatter (n, dimensionless) of each specimen at depths centered approximately 370, 620 and 880 μm beneath the skin surface. Average integrated attenuation coefficient (IA in dB.cm⁻¹) and frequency dependence of attenuation coefficient (β in dB.cm⁻¹.MHz⁻¹) were estimated across the depth between 240 and 1,000 μm. The three collagen distribution parameters were estimated using digitized microcopic fields from matched regions of histological sections stained with hematoxylin-eosin-saffron. No significant correlation was identified between collagen distribution parameters and IA or β. For the most superficial depth studied in abdominal skin, n was inversely correlated to collagen bundle thickness (r = −0.67, p = 0.002) and percent area (r = −0.65, p = 0.003). At the same depth, IBS was inversely correlated to percent area of collagen (r = −0.51, p = 0.03). The rather high collagen packing (48 to 82% area) measured in histological sections and the inverse relationship observed between IBS and percent area of collagen suggest that a packing factor should be included in models relating skin collagen distribution to ultrasound spectral parameters. A better understanding of the relationship between ultrasound parameters and the microarchitecture of the dermis should help to interpret changes in ultrasonic parameters observed during in vivo ultrasonic skin examinations.