Yes, the Epsilon™ can measure skin hydration with greater accuracy and flexibility than conventional single-sensor probes such as the Corneometer®. Both the Epsilon™ and the Corneometer use the same capacitance measurement principle. Both the Epsilon™ and the Corneometer have a sensing depth that confines the measurement predominantly to the Stratum Corneum. However, the Epsilon™ has 76800 sensors whereas the Corneometer has one. That’s a game-changer because skin is heterogeneous and skin-sensor contact is variable. The reason why the Epsilon™ can be used for hydration measurement is its unique calibration technology that converts the highly non-linear sensor output into a linear and calibrated response. By contrast, the MoistureMap (CK Technology sprl, Belgium) can only visualise hydration and needs a Corneometer for measurement. Our research shows that Epsilon™ and Corneometer measurements correlate well. The Epsilon™ measures hydration using its calibrated dielectric permittivity (dielectric constant, ε), scale rather than some arbitrary scale such as Corneometer Units. This works because the dielectric permittivity of water is much higher (ε~80) than that of other constituents of skin.
There are (at least) four reasons why Epsilon™ hydration measurements are superior to Corneometer hydration measurements, as illustrated below.
1. Hydration Heterogeneity Assessment
Epsilon™ images provide a revealing visualisation of hydration distribution in the vicinity of the site of interest. These images can be processed to give quantitative measures of heterogeneity in terms of Standard Deviation, Coefficient of Variation (CV%), or visually as images or histograms. This is illustrated in the volar forearm image shown below.
The two images left are the same, but with the 2.6mm diameter green Region of Interest (RoI) circle displaced by a horizontal distance of 3.8mm. The mean hydration within the left RoI is close to that of the whole image, with a dielectric permittivity of ε~15 and a CV~90%.
The mean hydration within the right RoI is clearly higher than that of the whole image, with a dielectric permittivity of ε~24.4 and a CV~70%.
From these data, the hydration of the two RoIs differ by more than 60%!
The Epsilon™ software displays colour-co-ordinated histograms that indicate hydration distribution for both the whole image (red) and the RoI (green). In this case, the hydration distribution within the left RoI is similar to that of the whole image, whereas there is a distinct shift towards higher hydration for the right RoI.
2. Correction for Skin-Sensor Contact
The Epsilon™ software has a powerful ε-filter to correct for bad contact between the sensor and skin. Bad contact may be due to microrelief or wrinkles, hair or other obstructions. The example below illustrates how this works.
The image on the left is of an area of male ventral forearm, where the skin-sensor contact is impeded mainly by hair and the protruding sensor surround. The bad contact shows up as black because of the low hydration/ε of hair and air. The associated histogram shows the bad contact by a prominent peak at low values of ε. Note that the histogram uses a logarithmic scale, which makes the peak look less prominant than it is.
The ε filter controls allow you to remove both low and high ε pixels from the image.
The image and its histogram show the effect of filtering pixels with ε-values below 3.5. Pixels removed by the filter are shown in grey in both the image and the histogram. Clearly, the ~43% of pixels that remain give more accurate skin hydration information (ε = 18.1) than the unfiltered image (ε = 8.3). Note that you can also use the Region of Interest (RoI) controls to focus on specific features within the filtered image.
3. Correction for Skin Surface Water
Skin surface water can be a problem when measuring hydration (i) in the vicinity of mucous membranes, occluded or damaged skin, (ii) in the presence of water-containing topical products, or (iii) in the presence of insensible perspiration or sweat. Skin surface water is not hydration, but its presence will increase hydration readings of conventional instruments. The example below illustrates how the ε-filter described in 2. above can deal with skin surface water.
The image on the left is of the second joint of a male left thumb. Its main feature is a size mismatch between the skin contact area and the sensor. However, the non-contacting pixels (76% of the total) have been removed by the ε-filter described in 2. above, as indicated by the grey colouring. For the remaining pixels, the mean hydration is ε~21.1 with a CV~60%.
The yellow/white spots are surface water from perspiration. The 2mm diameter RoI encloses an area where surface water predominates. Within this RoI, the mean permittivity is considerably higher (ε~27.8, CV~78%) than for the whole area of contact.
In the left image, the surface water was filtered out for ε-values above 60, as indicated by the light grey colour in the image and histogram. For the whole image, this filter changes the mean ε from 21.1 to 20.1. For the RoI, the change is from 27.8 to 21.3.
The effect of the surface water alone was a ~5% measurement error for the whole contact area and a ~23% measurement error for the RoI.
4. Time-dependent Measurementss
The Epsilon™ system can be used to study time-dependent phenomena by recording short bursts or long videos. This is illustrated below with a study of occlusive surface water accumulation in a tape stripping experiment. Scotch tape was used on a volar forearm skin site, repeatedly stripping the same site. Epsilon™ bursts (60 second total duration @ 1 frame per second) were recorded on the intact site and after every second strip.
Occlusion by contact between the skin and the Epsilon™ sensor during the 60 second burst measurements causes TEWL to be trapped as skin surface water. Trapped skin surface water causes the dielectric permittivity to increase, which shows up in the images as a colour change from dark red through to yellow.
Shown here are example images from the recorded bursts. Note that significant barrier damage becomes visible after just 2 strips, as indicated by the increased rate of surface water accumulation compared with intact skin. Also, the barrier damage after 14 strips is highly heterogeneous, with the bright yellow areas indicating greater than average damage.
These occlusion plots were calculated as whole-image averages. They clearly show (i) an almost unchanged hydration at t=0, irrespective of the number of strips removed, and (ii) increased barrier damage with number of strips removed.
These examples illustrate the capabilities of the Epsilon™ for static and dynamic skin hydration measurement on previously inaccessible sites such as hairy skin, scalp and skeletal joints, and under less than ideal conditions such as when sweat glands are active.
