Analysis of In Vivo Skin Anisotropy Using Elastic Wave Measurements and Bayesian Modelling

Abstract

In vivo skin exhibits viscoelastic, hyper-elastic and non-linear characteristics. It is under a constant state of non-equibiaxial tension in its natural configuration and is reinforced with oriented collagen fibers, which gives rise to anisotropic behaviour. Understanding the complex mechanical behaviour of skin has relevance across many sectors including pharmaceuticals, cosmetics and surgery. However, there is a dearth of quality data characterizing the anisotropy of human skin in vivo. The data available in the literature is usually confined to limited population groups and/or limited angular resolution. Here, we used the speed of elastic waves travelling through the skin to obtain measurements from 78 volunteers ranging in age from 3 to 93 years old. Using a Bayesian framework allowed us to analyse the effect that age, gender and level of skin tension have on the skin anisotropy and stiffness. First, we propose a new measurement of anisotropy based on the eccentricity of angular data and conclude that it is a more robust measurement when compared to the classic “anisotropic ratio”. Our analysis then concluded that in vivo skin anisotropy increases logarithmically with age, while the skin stiffness increases linearly along the direction of Langer Lines. We also concluded that the gender does not significantly affect the level of skin anisotropy, but it does affect the overall stiffness, with males having stiffer skin on average. Finally, we found that the level of skin tension significantly affects both the anisotropy and stiffness measurements employed here. This indicates that elastic wave measurements may have promising applications in the determination of in vivo skin tension. In contrast to earlier studies, these results represent a comprehensive assessment of the variation of skin anisotropy with age and gender using a sizeable dataset and robust modern statistical analysis. This data has implications for the planning of surgical procedures and questions the adoption of universal cosmetic surgery practices for very young or elderly patients.

A machine learning approach to predict in vivo skin growth

“…For simplicity of implementation and interpretation, the skin growth in this study was modelled to be transversely isotropic. This is motivated by clinical observations of tissue expansion in paediatric patients [13] and supported by our previous work, which indicated that younger patients have more isotropic skin tension [25]. Han et al relaxed this assumption and presented an anisotropic version of this model containing two growth rates for the transverse and longitudinal directions [29].…”

“…The goal of this study is to present a non-invasive method to measure important material properties of the skin, including natural pre-stretch, stiffness, and growth rate, and to predict the quantity of skin growth 7 days after expansion. Motivated by our previous work [25] [26], a non-invasive wave propagation technique is used in the following procedure:…”