JOURNAL OF COSMETIC SCIENCE 80 1) Standardized photographs of the 10 individual fi ngernails of all subjects were acquired using a Nikon D90 DSLR camera with a 28–105 mm macro NIKKOR lens under standardized lightning. The recorded photographs, blind coded, were used for paired comparison and atlas creation. 2) The three-dimensional (3D) in vivo dimensions of the nail were analyzed through the PRIMOS (Phase Shift Rapid In-Vivo Measurement of Skin) imaging system using fringe projection measurement principles. It was utilized in conjunction with a custom stereotactic fi nger-positioning device from Canfi eld Scientifi c to measure certain pa- rameters. Specifi cally, the nail profi le was constructed from measures of width, arc width, length, arc length, and radius of curvature. Figure1 illustrates the visual recon- struction of a fi ngernail using the 3D measurements. Arcs of width or length refl ect the curvatures in these two respective dimensions, where a theoretically, perfectly fl at fi ngernail would present the same value as that of width or length. Accordingly, the ratios between arc/width and arc/length illustrate the extents of curvatures of fi ngernails in both axes, the higher the ratio, the more curved the fi ngernail along the respective axis. The curvature radius is the radius of a circle that is required to cover the entire circumference of the arc width of a fi ngernail. This measurement is possible because the shape of a fi ngernail is fairly symmetrical across its width (which is not the case along the length of a fi ngernail). The wide and fl at fi ngernails would thus command larger curvature radii than narrow and curved fi ngernails. Another dimension measured is the distance between the center of the nail to the center of its base (namely, sagitta in geometry terms), which is referred here as the nail apex height. Taken together, these eight parameters describe the geometrical aspect of a fi ngernail. B) Thickness measurement of the nail plate. This was carried out at the center, center left, and center right of all nail distal edges per subject, using a digital caliper offering a 0.05-mm precision. The thickness of a fi ngernail is the average value of the three measurements. C) Grading by trained experts. Trained nail experts examined all 10 fi ngernails on every subject, one fi ngernail at a time, under controlled lighting with magnifi cation. The grading was conducted in a blinded manner toward the age and ethnicity of the sub- jects. Various aspects of fi ngernails were assessed. Nail shape and presence/absence of Figure 1. 3D reconstruction of a fi ngernail.

TYPOLOGY AND ATLASES OF HUMAN FINGERNAILS 81 certain conditions (e.g., luluna, leukonychia, etc.) were assessed by categorical choices. The severity or level of vertical ridges, nail shine, peeling, and an overall “unhealthy” appearance were assessed by using an analog scale of 0–5 (0, none, 5, extremely). D) Self-assessments of subjects about their agreements on the overall conditions and fea- tures of their fi ngernails as well as “like–dislike” feeling on some aspects of their fi ngernails. Subjects responded to a questionnaire of 25 items including aspects of fragility, hardness, thickness, and dullness. The creation of the atlases, using the typology data acquired, was dependent on the type of attribute under assessment: 1) Dimensional attribute atlases were created from statistical analysis to determine the max- imum and minimum values as well as the upper and lower quartile and median values for each attribute across the entire study population. 2) Visual attribute atlases were created from assessments by consumer perception of images on a computer using a paired comparison approach. Sixteen women of each ethnicity assessed the visual attributes for subjects of the same ethnicity. The sixteen women were equally divided across two age groups: 18–30 and 31–55 years of age. Images for the paired comparison study were selected to represent values across the full range of expert assessments. RESULTS OVERALL MEASUREMENT RANGES Table I summarizes the ranges of variations of some dimensional characteristics of all observed fi ngernails (10 nails × 280 subjects), according to age and ethnicity. All values are expressed in millimeter. Data from Table I indicate that age or ethnicity are of a rather low impact on most morpho- logical criteria with regard to the respective low ranges of averaged values. The rather large Figure 2. Illustrations of radius of curvature and nail apex height.