Evaluation of dark circles of the lower eyelid: comparison between reflectance meters and image processing and involvement of dermal thickness in appearance

• Published in: Skin Research and Technology Vol. 14; no. 2; pp. 135 - 141
• Main Authors: Ohshima, Hiroshi, Takiwaki, Hirotsugu
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.05.2008; Wiley
• Subjects: Adult; Colorimetry; Colorimetry - methods; Computer Simulation; dark circles of the lower eyelid; dermal thickness; Dermoscopy; Dermoscopy - methods; Eyelids; Eyelids - anatomy & histology; Eyelids - physiology; Humans; Image Interpretation, Computer-Assisted; Image Interpretation, Computer-Assisted - methods; image processing; MI&EI; Models, Biological; Photometry; Photometry - methods; reflectance meters; Reproducibility of Results; Sensitivity and Specificity; Skin; Skin - anatomy & histology; Skin Pigmentation; Skin Pigmentation - physiology
• ISSN: 0909-752X; 1600-0846; 1600-0846
• DOI: 10.1111/j.1600-0846.2007.00277.x

Background/aims: Dark circles of the lower eyelid (DCLE) represent a well‐known beauty problem. The pathogenesis of DCLE is obscure, although stasis and hyperpigmentation of the eyelids have been considered to be involved. One reason for the small number of studies on DCLE may be the difficulty in measuring such soft and curved skin as the eyelids using bulky reflectance meters. The purpose of this study was to quantitatively analyze DCLE using various bioengineering methods. Methods: The lower eyelid and cheek areas of 14 subjects with DCLE and 28 without DCLE were examined using two kinds of reflectance meters and image analysis to measure erythema index (EI), melanin index (MI), and oxygenation index (OX). Ultrasound echo images were also recorded to evaluate the thickness and echo density of the dermis. An in vitro model using collagen gel and hemoglobin solution was also examined as a phantom of eyelid skin. Results: When contact‐type reflectance meters were used, no significant differences in EI, MI, and OX were found between groups with and without DCLE. However, mean values of both MI and EI at eyelids were significantly higher in subjects with DCLE on image analysis, paralleling the results of inspection. Mean dermal thickness was significantly smaller in subjects with DCLE. Conclusion: Evaluation of EI and MI by image processing methods seems suitable for quantitative evaluation of DCLE, since inadequate contact of the measuring head with the skin is avoided. Whether stasis or hyperpigmentation is more responsible for DCLE remains uncertain. Dermal thickness of eyelid skin may be involved in the appearance of DCLE.