2001 ANNUAL SCIENTIFIC MEETING 61 Results and Discussion The results presented in Figure 2 show that the lipophilized flavonoid studied is able to significantly reduce the volume of bags under the eyes: 3 hours after application, the volume of the bags under the eyes had decreased on average by 0.21 nun, 5 hours after application, the volume of the bags under the eyes had decreased on average by 0.51 nun, The maximum reduction observed was 1.85 nun (5 hours after application of the formula containing 3 % of the lipophilized fiavonoid). Under the same experimental conditions, the placebo formula caused no significant reduction in volume of bags under the eyes 3 hours after its application. 5 hours after the application of the placebo formula, a significant reduction in volume of the bags under the eyes of 0.06 mm was recorded. This small decrease in volume could be attributed to a natural "deflation" of the bags during the day. 3 hours 5 hours 01 0.2 o3 04 o5 o6 - 0.21 "k Placebo Lipophilized flavonoid 3% ß 0.06 . - 0.51 Conclusion In conclusion, the lipophilized flavonoid is shown to have a statistically significant effect on the volume of bags under the eyes from the third hour after treatment. However, these results also show the relevance of the method used by COLETICA in the evaluation of the effects of this product. Only band projection could show its beneficial action on the volume of bags under the eyes The hpophilized fiavonoid developed is the only product at present for which the effectiveness on this parameter has been demonstrated and quantified in vivo. Bibliograph3 I. Gabor M In ,Plant Flavonoids m Biology and Medicine. Biochemical, Pharmacological, and Structure-Activity Relationships)) ¸ Alan R. Liss, Inc. pp. 471-480 (1986) 2. Lagarde J., Rouvrais C., Black D., Diridollou S, Gall Y. Skin Res. Technol. 7, 112-121 (2001) 3. Pittet J-C., Branka J-E., Servant J-J., Sechet B., Jacob S, Beau P. Skin relief: comparison of an interferometry technique by fringes projection (m wtro and m vtvo) and the image analysis of silicon replicas ISSI Congress, Washington (200 I) 4. Perrier E., Mariotte A., Boumendjel A., Rival D. fiavonoid esters and their use notably in cosmetics Patent (France: FR 98 06194), (United States of America' USA 6233296), (Germany, Japan, South Korea: pending)

62 JOURNAL OF COSMETIC SCIENCE THE BEAUTY IMAGING SYSTEM: FOR THE OBJECTIVE EVALUATION OF SKIN CONDITION Kukizo Miyamoto and Greg G. Hillebrand, Ph.D. The Procter & Gamble Company INTRODUCTION: Consumers want skin care products that target their personal skin care needs. Professional medical personnel need to be able to track changes in skin condmon as a result of various skin treatments. D•gital imaging offers a fast, accurate and objective means to evaluate and track changes in skin condition over time. We have developed a unique •mage analysis tool called the Beauty Imaging System (BIS) for measuring the most common facial skin problems. Results are conveyed in a meaningful way that both scientists and non-scientists can understand and use. The system employs conventional high resolution digital cameras for capturing facial •mages under both normal and long wave length UVA illumination combined with sophisticated image analysis software for identifying and quantifying facial wrinkles, texture, pores and hyperpigmented spots. Drawing on a vast store of images collected from around the world, a person's image analysis results are expressed as a percentfie for her age and ethnic background. The Beauty Imaging System has several utilities including skin consultation, clinical studies, monitoring the progress of skin treatments m a professional setting and image archiving. SYSTEM DESCRIPTION: The Beauty Imaging System is shown m Fig. 1. The •maging booth was designed to be free from ambient light and can provide two types of illuminanon for measurement of visible and invisible information. The client's head is positioned in the booth with the aid or chin and forehead rests. An image is captured of either the left or right side of the face. A 1.3 mesa pixel CCD digital camera eqmpped with a close-up lens •s mounted in the imaging booth and the collected images are digitally transformed and saved to the computer. These collected images are then analysed to quantify the skin features by the set of image analysis programs. Each measured skin feature can be then compared with the average for the person's age and ethnic group and expressed as a percentile. The captured images and image analysis data are stored in the computer hardware, for subsequent tracking of changes in skin con&tion. All the system operation •s controlled by the user friendly software program. IMAGE ACQUISITION: System illuminanon and the camera can be cahbrated by a white-balancing program to adjust the target brightness values by using a GretagMacbeth neutral 8.0 gray color board accurately positioned in front of the camera. We equipped a fixed- position chin rest and forehead rest in the booth for accurate posmomng of the face. Once the subject's head is positioned within the imaging system, two facial images ofe•ther the left or right side are captured under normal and UV-fiuorescent illumination. For measurement of changes in facial skin condition during a course of the treatment, a baseline image is taken at the first visit and •s used for subsequent images to accurately reposition the subject's head. SubJects should be evaluated at approximately the same nine of the day during the course of the study to prevent any influence of diurnal vananon affecting the skin condinon. Factal image byNormal dhtmtnatton 6500øK high frequency fluorescent white light sources were mounted on the top and on the sideof the booth providing even illumination while enhancing skin topographical reformanon (e.g., wrinkles and texture) in the facial images. This mmge •s used for the measurement of facial wrinkles, texture, pores and hypcrpigmented spots as one would observe under normal ambient lighting condmons. Facial rotage b)' UV- fittoresce,t dhmtmation Two strobe lights were mounted on the left and right s•des of the •mag•ng booth. The strobes were eqmpped with glass band pass filters whmh can allow transmission of onb, long wavelength UVA t, 350-400nm, peak range of 370nm). This UV- fluorescent •mage is used for the visuahzation and measuremcnt of 1) famal hyperp•gmentation not visible under normal white light illumination 2) bacterially-produced porphyrins m the skin pores which can not be perceived under normal white light illumination. These two parameters measured by UV-fluorescent images were compared w•th visible hyperpigmented spots and pores measured by fluorescent images. QUANTIFICATION OF SKIN INFORMATION A: Normal fluorescent •mage Image is masked for the region of eye around area and cheek and four key parameters were measured by the •mage analysis programs as following parameters: 1: Total wrinkle area/masking area = wrinkle area fraction 2: Total texture area/masking area = texture area fraction 3: Total pore count/masking area = pore count fraction 4: Total visible hyperp•gmented spot count/mask area = spot count fraction B: Diagnose of skin condition Using BIS, we have collected facial images of 3160 healthy females who are Caucasians, Asians, Indians, H•spanics and African Americans living m rather the US, Europe and Japan from age 10 to 65 years. These •mages were analyeed by BIS and statisncal models were developed for the age dependent changes •n each of the skin features of interest (wrinkles, texture, pores and spots) for each racial group. Using these models, a individual's BIS results can be compared to the average for her