High-frequency ultrasound skin thickness: Comparison of manual reading and automatic border detection includes assessment of interobserver variation of measurement

BACKGROUND: Twenty megahertz ultrasound is the preferred method for in vivo measurement of skin thickness. This study on methodology aimed to compare operator-managed or "manual" measurement of skin thickness (MM) with measurement by automatic border detection (ABD).

METHODS: The skin thickness of the flexor mid-forearm was measured in 48 subjects (subgroups 32 and 16) with the DermaScan C® (Cortex Technology). MM and ABD were performed on the same scans by two experienced operators. MM is measured on the peak of interface A-mode echoes representing the epidermal surface and the dermis-subcutis interface, while ABD measures on the slope of the echoes, for example, the acoustic contour.

RESULT: Automatic border detection measured the skin about 20% thicker (mean: 0.94/0.89 mm, SD: 0.09/0.11) in comparison with MM (mean: 1.25/1.22 mm, SD: 0.09/0.12), P < .001. The two methods correlated, r = .9443 and .8663. Two individual researchers reached very close results using MM, that is, mean 0.91 and 0.93 mm, and the interobserver variation was nonsignificant.

CONCLUSION: MM and ABD for ultrasound skin thickness measurement rely on different principles and consequently measure skin thickness differently, with ABD overestimating the thickness with about 20%. MM measures anatomically more correct and is preferred in research. MM has been validated against area-based thickness measurement without contour artifact and produced identical results. ABD is less operator-dependent and applicable to field studies, and useful in the beauty industry to characterize aged skin. The two methods of recording each have a scene of use.