In vivo skin anisotropy dataset from annular suction test

To characterize the anisotropic and viscoelastic behaviors of the skin, we conducted an experimental campaign of in-vivo suction tests using the CutiScan®CS100 device from Courage and Khazaka electronics. In this data paper, we present the raw acquired data of the tests and their respective treated data. The tests were performed 30 times on the anterior forearm of a 28-year-old Caucasian male at different pressure set-points, ranging from 100 to 500 mbar with an increment of 20 mbar, at ambient temperature in a windowless room. The primary dataset consists of videos recorded by a probe camera associated with the CutiScan® device during the tests. After data treatment with DIC (Digital Image Correlation) technique and based on a homemade Python program, we have obtained secondary data tables and 2D displacement for all mapped grid nodes.


Specifications
Material Characterization Specific subject area Anisotropic viscoelastic characterization of human skin by digital image correlation technique from multiaxial ring suction test Type of data Video Table  Image How data were acquired Multiaxial ring suction test with CutiScan ® probe-camera. Software CS100 Homemade Python program based on PyDIC suite [1] Data format Raw Analyzed Parameters for data collection The CutiScan ® probe was perpendicular to the forearm without any initial press. The observable part of the skin did not contain any hair. The tests were done at ambient temperature and as often as possible at the same time slot. Description of data collection

Value of the Data
• The data will help analyze human skin's mechanical response when subjected to multi-axial ring suction load. Thus, both anisotropic and viscoelastic phenomena could be explored as the response is temporal and covers all directions, from 0 °to 360 °. • Biomechanics researchers can benefit from these data for a complete skin characterization study. Computer scientists could also be interested in using this dataset to improve DIC (Digital Image Correlation) technique. • This dataset can be combined with data issued from other mechanical tests to identify more independent material parameters. • The reproducibility of the 30 data series may be either used in the context of Machine Learning training to predict the mechanical behavior efficiently and fast.

Data Description
There are two folders in the shared public repository, 'DATA_SUCTION' and 'PYTHON_SOURCES.' The first contains primary and secondary data with a singular 'README' text file. This text file provides additional details about the 30-measurement series, such as experiment time slots and data quality, based on a qualitative evaluation of anisotropy distribution stability. While in the second folder, one would find Python codes used to generate secondary data by treating the primary ones.
The primary data, video files with '.avi' format, are retrieved for each recorded ring suction test from the commercial software CS100 supplied with the CutiScan ® . Fig. 1 represents a snapshot from a video recording, where the dark zone corresponds to the observed skin area. The experiment is duly explained in Section 'Experimental Design, Materials and Methods'. These video files were subjected to a DIC (Digital Image Correlation) process to track the motions of all Cartesian mapped grid points. Once new positions of the letters are captured for each timestep (frame), we computed their related displacements relative to the initial frame. In addition, but not necessarily, we computed strains and polar coordinates. These secondary data are stored in '.csv' format with a size of 2401 rows and 23 columns ( Tables 1-3 ) using the scale '950 pixel / 5 mm'. The 2401 lines represent the data on the 49 × 49 nodes (rectangular grid with x and y coordinates), while the 23 columns represent respectively: Table 1 Partial view of a '.csv' table with data on every node resulting from DIC process: from column 1 to column 9. The table file path is '2020_12_15 \ p_30 0 \  Table 2 Partial view of a '.csv' table with data on every node resulting from DIC process: from column 10 to 17.  Finally, the content of columns 4,5,8, and 9 have been used to map 2D-displacement vectors on nodes with a scale factor of 5, using the reference image as background ( Fig. 2 ). Displacement vectors are plotted in red on every node. 'PYTHON_SOURCES' repository contains Python programs based on PyDIC suite methods [1] , which were used to generate CSV tables from AVI videos. The code sources are detailed in Section 'Experimental Design, Materials and Methods'.

Experimental Design, Materials and Methods
The primary measurement videos have been captured for 30 ring suction test series performed on the anterior forearm of a 28-years-old Caucasian male using CutiScan ® probe-camera [2] , as often as possible at the same time slot. The forearm was positioned perpendicular to the probe ( Fig. 3 ). Laiacona et al. [3] used a similar homemade device with a 30 mm-diameter measurement zone. Fig. 4 illustrates the working principle of the ring suction test. Negative pressure was applied within an annular section, around the measured central zone. The inner and outer diameters of the ring are, respectively, 5 mm and 14 mm. The central zone is subjected to in-plane multi-axial extension. The pink medium represents the skin layer, and black arrows display the direction of the ring area displacement.
A double-sided adhesive ring tape was placed around the measurement area to ensure a good seal between the probe and the skin ( Fig. 4 ). The tape is stuck to the probe oriented along the zero-angle of the camera ( Fig. 5 ).
Load application phases were carried as follows ( Fig. 6 ): -Primary cycle: apply the set aspiration pressure instantaneously, hold it for 1 s, and then release it for 3 s. -Standby phase: wait for the skin to recover its properties for 10 s -Measurement cycle: apply instantaneously the set aspiration pressure and hold it for 3 s, then release it for 3 s.
This routine was repeated for 21 negative pressure set-points varying from 100 to 500 mbar with a step of 20 mbar. A delay of 2 min between each measurement is required to allow the skin to recover its viscoelastic properties. Hence, a daily data series lasts around 1 h. radius r = { 0 . 5 ; 1 ; 1 . 5 ; 2 ; 2 . 5 } mm has been plotted with respect to the angle ˜ θ. Also, the evolution of the displacement norm ˜ a over time has been represented for different angles ˜ θ = { 0 • , 45 • , 90 • , 135 • } along the circle r = 1 mm. The Python codes used to generate the graphs are available in https://dmarc.femto-st.fr/fichiers/691 .     Fig. 7. A preview of derived graphs from temporal-spatial data analysis.

Ethics Statement
The described work has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. All data have been collected with the patient's consent, and the patient's privacy is fully preserved. Recently, we have obtained approval from the Committee for the Protection of Persons (CPP) to conduct a similar experimental campaign on 30 subjects.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships which have or could be perceived to have influenced the work reported in this article. gratitude to Élodie Veyrat-Durebex, from Biomedical Engineering School at the University of Franche-Comté, for her contributions in exploring the presented suction test outputs.