IMAGE ANALYSIS AND HAIR REMOVAL EFFICACY 353 Hair length, thus, is not a useful parameter for evaluation of depilation techniques because it represents the mean length of non-depilated hairs shortly after hair removal and gives a false impression, especially in the first days after hair removal. Using the projection area, all visible hair material contributes to the assessed values, and so the data reflect the same for wax-depilated and shaved test areas, namely the completeness of hair removal. The projection area also correlates with the view of an observer who recognizes the visible hair parts as a clearly contrasting dark matrix of hairs, often unable to discrimi­ nate between single hairs. Even very short, but visible, stubble contributes to this observation. Therefore, no detectable hair should be removed to obtain a valid regrowth parameter. The projection area of hairs also enables a valuation of the closeness of hair removal. Hair length, hair width, and hair count each could contribute one interesting part to the result, while the projection area combines the information of all three parameters in one. However, in the case of studies measuring hair growth actives for increases or decreases in growth rate, hair length still should be regarded as the parameter from which to calculate growth velocity because, in these studies, the more relevant parameter is the length of grown hairs in a certain time than the overall amount of hairs. The efficacy of a depilation method can be expressed as E r = t m / t s . E r is the relative time expansion compared to shaving, t m the detected time for a method to obtain hair regrowth after its use, and t s the time for hair regrowth after shaving. These parameters can be determined in a study design such as we used. As an example, in our study, E r was evaluated to approximately 2.3, meaning wax depilation extended the reappearance of hairs by a factor of 2.3 compared to usual shaving. Since plucking methods vary in their outcome (influence of subjects and the closeness of the methods), a new plucking method should be tested in a paired test against a usual method (for example, wax depilation or plucking with tweezers). Treatments claiming to reduce hair growth should be compared to regrowth after shaving without the active. Measurement of the active should be performed in the same test panel and on the same test area as the assessment without treatment with the active. For female leg hairs, a growing period of at least one week is needed to accurately measure the regrowth of hairs after shaving. In our experiments, wax-depilated hairs broke in different depths inside the follicles (see Figure 4). The mean follicle depth on the legs can be estimated to be approximately 2.7 mm (5 ). In case all hairs break at the bottom of the follicle and assuming a growth speed of 140 µm (data derived from our measurements), it would take approx. 20 days for the hairs to return to the surface. This correlates well with our overall finding that the original status was reached approximately 18 days after hair removal. The beginning increase of the projection area values approximately one week after hair removal and the findings on single hairs (Figure 4) give an indication that hairs break at different depths in the follicle, with a mean of approximately 1.3 mm beneath the surface. Improvement of the depilation method theoretically could increase the delay up to twofold without interfering with the hair growth cycle. Depilation with different methods, especially electrical energy-based depilation (14, 15 ), may result in different skin irritation. Image analysis of the red fraction of the image enables one to quantify spotty irritation, which is typical for depilation techniques due to irritation of the hair follicles. Image processing of the same pictures as for hair

354 JOURNAL OF COSMETIC SCIENCE evaluation can be used (see Figure 5 ). Thus, the determined data is from the same subject, test area, and measurement time as for the assessment of hair reappearance. For a final classification of a depilation method, three parameters, (i) relative regrowth velocity, (ii) depilation closeness, and (iii) irritation level are most important. In this proposed method, all these parameters can be derived from the same set of images in the same study. CONCLUSION Established methods like shaving, plucking, and chemical depilation, and even laser technologies, attain the goal of keeping the skin free of hairs only for a period of time before the hairs reappear. For the quantification of how good a hair removal method is, the determination of at least three parameters is necessary. These parameters are re­ growth velocity, depilation closeness, and skin irritation. While the time that elapses until hairs reappear on the skin is the most important parameter, the closeness of hair removal as assessed by determining non-removed or incompletely removed hairs is also of great importance to benchmark the quality of hair removal methods. The additional evaluation of the technique's unwanted effects like itch, pain, and erythema, or even injury to the skin, completes the list of the crucial parameters. All three parameters can be assessed in one study close to the daily live situation by using the described study procedure and image-analysis method. REFERENCES (1) E. A. Olsen, Methods of hair removal,]. Am. Acad. Dermatol., 40, 143-155 (1999). (2) J.B. Hamilton, Patterned loss of hair in man types and incidence, Ann. N. Y. Acad. Sci., 53, 708-728 (1951). (3) 0. T. Norwood, Male pattern baldness: Classification and incidence, South. Med.]., 68, 1359-1365 (1975). (4) 0. Braun-Falco and G. P. Heilgemeir, The trichogram. Structural and functional basis, performance, and interpretation, Semin. Dermatol., 4, 40-52 (1985). (5) R. D. Gibbons, V. C. Fiedler-Weiss, D. P. West, and G. Lapin, Quantification of scalp hair-A computer-aided methodology,]. Invest. Dermatol., 86, 78-82 (1986). (6) J. D. Peereboom-Wynia, C.H.Beek, P. G. Mulder, and E. Stolz, The trichogram as a prognostic tool in alopecia areata, Acta Derm. Venereal., 73, 280-282 (1993). (7) M. Saitoh, M. Uzuka, and M. Sakamoto, Human hair cycle, J. Invest. Dermatol., 54, 65-81 (1970). (8) D. J. Van Neste, B. de Brouwer, and W. De Coster, The phototrichogram: Analysis of some technical factors of variation, Skin Pharmacol., 7, 67-72 (1994). (9) R. Hoffmann, TrichoScan: A novel tool for the analysis of hair growth in vivo,]. Invest. Dermatol. Symp. Proc., 8, 109-115 (2003). (10) M. D. Van Neste, Assessment of hair loss: Clinical relevance of hair growth evaluation methods, Clin. Exp. Dermatol., 27, 358-365 (2002). (11) T. Leroy and D. Van Neste, Contrast enhanced phototrichogram pinpoints scalp hair changes in androgen sensitive areas of male adrogenetic alopecia, Skin Res. Technol., 8, 106-111 (2002). (12) D. Van Neste, T. Leroy, and E. Sandraps, Validation and clinical relevance of a novel scalp coverage scoring method, Skin Res. Technol ... 9, 64-72 (2003). (13) M. E. Roersma and G. J. Veldhuis, Proposal and evaluation of a Monte Carlo model for hair regrowth following plucking, Skin Res. Technol., 7, 176-183 (2001). (14) P. Bjerring, H. Egekvist, and T. Blake, Comparison of the efficacy and safety of three different depilatory methods, Skin Res. Technol., 4, 196-199 (1998). (15) R. F. Wagner, Jr., C. A. Flores, and L. F. Argo, A double-blind placebo controlled study of a 5% lidocaine/prilocaine cream (EMLA) for topical anesthesia during thermolysis,j. Dermatol. Surg. Oneal., 20, 148-150 (1994).