10 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Figure 12. Octadecyl rhodamin fluorescence: a) longitudinal section from 20 Izm below the surface b) longitudinal axial section cutting the medulla. The main advantage of confocal microscopy over conventional light microscopy is im- proved resolution, as confirmed by the images presented here. This led to a remarkable similarity between the images reconstructed from optical sections acquired by the TSM and those provided by the SEM. SEM requires extensive sample preparation that can introduce artefacts. In general, exogenous deposits on the hair surface are disturbed by the coating required for SEM. In addition, the fact that electron microscopy requires samples to be observed in a vacuum rules out visualization in their natural environment. Our results show that oil- and water-based deposits can be visualized on the hair surface by means of confocal microscopy, and this opens up new possibilities for cosmetic research, particularly in the study of protective preparations such as waxes and oils. Another advantage of confocal microscopy is that it provides not only 3D images, but also exact spatial measurements. It is therefore possible to quantify changes in the organization of cuticular cells and to compare objectively the efficacy of cosmetic prod- ucts. The periodicity, contours, orientation, thickness, and lifting of the surface scales can be measured to within about 0.1 Ixm, and the rapidity and simplicity of the method make it possible to apply valid statistical methods to the results. In addition to micro- metric measurements, the illumination of the specimen by reflected white light provides information on brightness in the "reflection map." Quantitative evaluation of brightness needs a calibration of the TV camera. This was not done in our experiment, but we were able to evaluate local variation in brilliance on a single fiber. Such information is not provided by the SEM. Important modifications of the hair fiber caused by permanent waving have already been reported (8), but subtle changes of the surface aspect of the cuticle had never been assessed. Images of the TSM showed that the treatment lifted the free border of cuticular cells this effect could be objectively assessed. Bleaching is known to cause very little damage to the hair surface, as observed by conventional SEM (8,9). Zelinski (7) recorded an increase in roughness but did not

CONFOCAL 3D RECONSTRUCTION OF HAIR 11 interpret his observation. With the confocal microscope, we were able to confirm the corrosive effect of the treatment, as previously suggested (8). The simplicity of this imaging method and its non-destructive nature open the way to studies of dynamic phenomena in four dimensions. This is illustrated by our study of swelling and stretching. Although hair swelling has previously been studied by means of physical techniques, this is the first time, to our knowledge, that periodic phenomena (bulging) have been described. The same is true of the stretching studies. Although measurements of resistance to traction have classically been used to evaluate the visco- elastic properties of hair, changes in surface structure during stretching have rarely been reported because of technical difficulties inherent in conventional microscopy. The variations we observed in the brightness of the borders of cuticular cells during stretch- ing suggest their rising as previously reported by Brown and Swift (10). Hair specialists are well aware of the fastidious and time-consuming nature of methods for preparing longitudinal and transverse sections. Even the most skilled technicians rarely produce sections that are perfectly aligned with the hair axis. The optical sec- tioning property of the confocal microscope overcomes these problems, since sections can be visualized at any given level and in any given direction in a fraction of a second. The images of the internal hair structures obtained after labeling with a fluorescent marker were also of exceptional quality, compared to those obtained by observation of fiber sections in conventional fluorescence microscopy (11). Finer details of the structure of the cortical cells were observed, both in the radial and axial directions. This technique also provides a simple tool for monitoring the penetration of dyes throughout the hair thickness. In conclusion, confocal microscopy is highly suited to studies of human hair and the effects of cosmetic preparations. It is particularly adapted to observations of rounded surfaces and is thus a valuable alternative to classical microscopic methods. A particular advantage is the non-destructive nature of the technique, which means that samples can be observed both before and after treatment this opens up exciting new possibilities for cosmetic research. ACKNOWLEDGMENTS We are grateful to R. Aslund from Molecular Dynamics (Sunnyvale, CA) for providing the images obtained with the Sarastro microscope. REFERENCES (1) J. A. Swift, Fine details on the surface of human hair, lnt. J. Cosmet. Sci., 13, 143-159 (1991). (2) T. Wilson, Confocal Microscopy (Academic Press, London, 1990). (3) M. Petran, M. Hadravsky, and A. Boyde, The tandem scanning reflected light microscope, Scanning, 7, 97-108 (1985). (4) C. Zviak, The science of hair care, Dermatology, Vol. 7 (Marcel Dekker, New-York, 1986). (5) N. Aslund, K. Carlsson, A. Liljeborg, and L. MajliSf. Phoibos, a microscope scanner designed for micro-fluorometric applications, using laser induced fluorescence. Proceedings of the 3rd Scandinavian Conference on Image Analysis (Student Litteratur, Lund, Sweden, 1983), pp. 338-343. (6) C. R. Robbins, in Chemical and Physical Behavior of Human Hair (Springer-Verlag, New York, 1988).