Assessment of features in facial hyperpigmentation: Comparison study between VISIA and CSKIN

Hyperpigmentary disorder is one of the commonest skin concerns in dermatology clinics. The availability of noninvasive instruments provided a convenient, objective, and reproducible methodology for the evaluation of pigmentation and skin color. The aim of this study is to compare CSKIN and VISIA in measuring facial hyperpigmentation, as well as to assess the correlation between the instrumental analyzing and clinical evaluation.


INTRODUCTION
Hyperpigmentary disorder is one of the commonest skin concerns in dermatology clinics, and making an objective and precise assess- bring negative psychological impact and decrease life quality of the patient. 4,5 Visual severity scales are well recognized for clinical assessment by clinicians and evaluators and so on (the MASI scale for the evaluation of melasma). 6 However, visual assessment could be affected by subjective factors and time-consuming and, thus, are prone to significant interobserver variability. 7  DermaSpectrometer, 10 Chroma Meter, 10 VISIA, 11 Antera 3D, 9 and ImageJ with its specially designed plug-in, 12

Study population
Eighty volunteers (8 male and 72 female, 34.6 ± 11.6 years old) were recruited from January 2 to May 18, 2019 in the department of dermatology, West China Hospital, Sichuan University. All the volunteers enrolled have given written informed consent.

Measurements
All measurements were performed under the controlled dark conditions (22 ± 2 • C, 45% ± 5% relative humidity). Volunteers were asked to clean their face with clean water 20 min before the test. Pictures of the front, left, and right side of the face were taken by two instruments, respectively.

Statistical analysis
Correlations between VISIA and CSKIN were determined by calculating the Pearson correlation coefficient. Spearman's rank correlation coefficient was used to determine the correlation between visual grading provided by the dermatologists and each parameter for the brownies of imaging analysis produced by VISIA and

Instrumental differences between VISIA and CSKIN
The indices of brown spots were recorded and measured by VISIA ( Figure 1A-C) and CSKIN ( Figure 1D-F). A brief comparison of the two instruments was depicted in Table 1.

Correlations of the facial features of hyperpigmentation between VISIA and CSKIN
Pigmentation-related parameters measured by CSKIN (brown pixels and brown percent) showed statistically significant correlation with VISIA (brown spots feature counts) (Figure 2).

Evaluated scores of skin hyperpigmentation by the dermatologists and its internal consistency
Precisely, 240 times of visual evaluation have been made in terms of hyperpigmentation by three dermatologists who were double-blinded.
Significant consistency was observed among the results evaluated by the dermatologists (Table 2).

Correlations between parameters analyzed by instruments and the visual scores graded by dermatologists
Brown spots feature counts measured by VISIA showed moderate correlation with mean visual scores graded by the three dermatologists.
In contrast, brown pixels measured by CSKIN presented better consistency with the clinical grading. Brown spots absolute scores, another parameter of VISIA, revealed weak correlation with manual grading.
However, no significant correlation was found between brown percent by CSKIN and visual grading (Table 3).

DISCUSSION
"Brown spots" refers to lesions on and deeper within the skin, such as post-inflammatory hyperpigmentation, freckles, lentigines, and melasma. Skin color is predominantly determined by pigments, such as melanin, hemoglobin, bilirubin, and carotene. Skin hyperpigmentation occurs when there is a change in melanin production and/or its distribution. 17 Melanin has a broad absorption across visible wavelengths (400-700 nm) and ultraviolet (100-400 nm). 18 The VISIA system generates a series of photographs using standard, ultraviolet, is spectrally dependent on the encompassed chromophores, including both types of hemoglobin and melanin, it is difficult to optically monitor and precisely quantify the chromophores concentration alterations using only RGB channels in photography. 21 In comparison, the CIELab color system was designed based on psychophysical experiments to quantify colors and brightness in a way that is photometrically accurate. 20 The feature of brown spots is enhanced by the combination of the selected data from the yellow channel of the polarized white light, the luminance channel of the green image, and the blue channel of the blue image in the Lab color space. In addition, it is suggested that higher resolution images (Table 1) and well-evolved algorithms for image processing and feature identification attributed to the accuracy of CSKIN as well. VISIA is implemented with the RBX (red/brown/X) technology from Canfield, which allows a semiquantitative assessment of specific skin chromophores. However, CSKIN has implemented a modulated algorithm in specific of the feature of the hyperpigmentation, which usually presents as patches rather than punctiform macules in patients with pigmented skin diseases, such as melasma and Nevus of Ota. During the data processing, the color distinction is enhanced and analyzed between the brown area and the normal skin area, and further calculation is added for the parameter optimization.
To set up a reliable standard for comparison, not only the evaluation criteria were unified, but also the evaluation conditions for the patients were also strictly controlled. These factors were considered to have contributed to the fairly high consistency in the visual score graded by dermatologists in the study. However, even on such unified conditions, the internal consistency of the grading results was not satisfying.
In clinical practice, the evaluation by a doctor could be subject to a vari- to understand their initial concern and to notice their skin problems.
There are some factors that contribute to the applicability of CSKIN.
The iPad, a portable digital device, works as a control panel by connecting to the photo booth of CSKIN added convenience for operating.
Besides, "Cloud" data have made it more available for patient management and statistical analysis. As reported by Chen et al., 13 CSKIN also showed stronger capacity in the assessment of erythema compared with VISIA. However, further clinical trials are needed to verify the evaluation ability of other parameters of VISIA and CSKIN, including wrinkles, pores, UV spots, and porphyrins.

CONCLUSION
Both CSKIN and VISIA showed a fairly good assessing capability in the pigmented areas of the patients. The major parameter in measuring hyperpigmentation by CSKIN presented with a higher correlation with visual grading by the dermatologists in comparison with that of VISIA.
Therefore, it is suggested that CSKIN might perform a better role in the facial hyperpigmentation assessment and could serve as a reasonable alternative in the evaluation and follow-up management of skin disease with hyperpigmentation.