A novel ultra-light suction device for mechanical characterization of skin

Abstract

Suction experiments have been extensively applied for skin characterization. In these tests the deformation behavior of superficial tissue layers determines the elevation of the skin surface observed when a predefined negative (suction) pressure history is applied. The ability of such measurements to differentiate between skin conditions is limited by the variability of the elevation response observed in repeated experiments. The scatter was shown to be associated with the force exerted by the observer when holding the instrument against the skin. We have developed a novel suction device and a measurement procedure aiming at a tighter control of mechanical boundary conditions during the experiments. The new device weighs only 3.5 g and thus allows to minimize the force applied on the skin during the test. In this way, it is possible to reliably characterize the mechanical response of skin, also in case of low values of suction pressure and deformation. The influence of the contact force is analyzed through experiments on skin and synthetic materials, and rationalized based on corresponding finite element calculations. A comparative study, involving measurements on four body locations in two subjects by three observers, showed the good performance of the new procedure, specific advantages, and limitations with respect to the Cutometer®, i.e. the suction device most widely applied for skin characterization. As a byproduct of the present investigation, a correction procedure is proposed for the Cutometer measurements, which allows to partially compensate for the influence of the contact force. The characteristics of the new suction method are discussed in view of future applications for diagnostic purposes.

Fig 1. Schematic illustration of the Nimble with all components.

The aspiration probe is connected (tubes T1 and T2) to pressure sensors PS1 and PS2 via filters 1 and 2. T1 is connected to a peristaltic pump, which generates the progressive negative pressure in the cavity. The valve releases the pressure at the end of the measurement.

Fig 2. Aspiration probe.

(A) Schematic of aspiration probe: The position of the vertical tube T1 defines the elevation height h. (B) The tissue is drawn into the cavity until it closes T1. (C) The dimensions of the aspiration probe are indicated.

Fig 3. Output curve of Cutometer measurements.

Representative Elevation-Time curve (left) and Pressure-Elevation curve (right) of a Mode 2 Cutometer measurement. The maximum elevation is indicated with R0 and the loading and unloading curves are displayed. From the loading part of the Pressure-Elevation curve the closing pressure at 1 mm elevation is extracted for comparison with the Nimble parameter pcl^nimble. T₀ and T_R0 indicate the start of the experiment and the timepoint when R0 is reached respectively. These graphs were generated with a maximum suction pressure of 400 mbar.

Fig 4. Output curve of Nimble measurements.

Representative Pressure-Time curve of a Nimble measurement for a defined elevation of 1.0 mm. Pressure measured by PS1 and PS2 are indicated. The closing pressure is recorded when Δp = 5 mbar is recognized.

Fig 5. Influence of contact force on volar forearm.

Weight study with Nimble (A) and Cutometer (B) on human volar forearm. The grey blocks indicate added mass of each 20.0 g, corresponding to increased contact force; (C) shows results of measurements on volar forearm with the Cutometer (red) and the Nimble (blue), and corresponding FE calculations on skin (orange).

Fig 6. Representation of the correction procedure.

(A) Cutometer measures tissue elevation from a baseline, defined by the apex of an initial deformation (response to contact force). (B) Nimble measures tissue elevation from skin surface (negligible initial deformation) due to its low weight. (C) The correction scheme accounts for this discrepancy and adds the Offset to the elevation measured by the Cutometer.

Fig 7. Mean stiffness and standard deviation for each location and subject.

(A) and (B): linear regression for a slope of 1.1 between mean stiffness k^cuto and k^nimble (A) and corrected Cutometer data (B), for eight different skin locations. (C): standard deviation of stiffness measurements for each body location separated for two subjects (VO1 and VO2).

Fig 8. Mean values and standard deviation of stiffness values k^R0, k^cuto, kcorrR0,kcorrcuto and k^nimble.

Values are shown for each location separated by subjects. Significant difference between locations measured by each device are indicated. Significance level are indicated as **p < 0.01 and *** p < 0.001.

Fig 9. Intraobserver variability in terms of linear dependency.

Linear dependency of stiffness values k^R0, k^cuto, k^nimble, ${\mathit{k}}_{\mathit{c}\mathit{o}\mathit{r}\mathit{r}}^{\mathit{R}\mathbf{0}}$ and ${\mathit{k}}_{\mathit{c}\mathit{o}\mathit{r}\mathit{r}}^{\mathit{c}\mathit{u}\mathit{t}\mathit{o}}$ based on repeated measurements at each location (VF, FH, BH and LB).

Fig 10. Histogram plot of error between measurements of different observers (n = 3) for stiffness values k^R0, k^cuto, k^nimble, kcorrR0 and kcorrcuto.

The error is defined as the percentage difference of each measured parameter with respect to the mean value for that sample. Root mean square (RMS) is indicated for each parameter (RMS^R0 = 19%, RMS^cuto = 46%, RMS^nimble = 24%, ${\mathit{R}\mathit{M}\mathit{S}}_{\mathit{c}\mathit{o}\mathit{r}\mathit{r}}^{\mathit{R}\mathbf{0}}$ = 13% and ${\mathit{R}\mathit{M}\mathit{S}}_{\mathit{c}\mathit{o}\mathit{r}\mathit{r}}^{\mathit{c}\mathit{u}\mathit{t}\mathit{o}}$ = 53%).

Fig 11. Mean and standard deviation of Offset values in mm (measured by the Cutometer software).

Left: Mean Offset over all locations for each observer (O1, O2 and O3). Middle/Right: Mean Offset over all observers for each location (VF, FH, BH and LB) separated for two subjects (VO1 and VO2).