Modifications of the coherence radar for in vivo profilometry in dermatology

Transcription

1 Modifications of the coherence radar for in vivo profilometry in dermatology P. Andretzky, M. W. Lindner, G. Bohn, J. Neumann, M. Schmidt, G. Ammon, and G. Häusler Physikalisches Institut, Lehrstuhl für Optik, University Erlangen-Nürnberg, Erlangen, Germany ABSTRACT Important aims in dermatology are the measurement of pathological alterations of human skin and on the other hand the quantification of the influences caused by pharmaceutic and cosmetic products. We present modifications of the well established coherence radar that allow in vivo measurements of human skin in spite of involuntary body movements and bloodflow. The measuring field can be varied from 100 x100 µm² to 5 x5mm². The measuring time is 5 to 15 s and the longitudinal measuring uncertainty is about 2 µm. A fiberoptical implementation allows the separation of the sensor head from the mechanical scan. The mobile and compact sensor head can now be freely positioned and adjusted to each part of the patients skin. Disturbances caused by unavoidable movement of the patient can be compensated by modified setups of the coherence radar. We show measurements of clinical and cosmetical relevance. Keywords: optical coherence profilometry, short coherence interferometry, speckle interferometry, human skin, time-domain, dermatology, in vivo measurements 1. INTRODUCTION In 1992 the coherence radar 4 as a tool to acquire the surface topology of technical objects was introduced. Here we will present modifications of white light interferometry on rough surfaces 3 to measure human skin. Clinical relevant data about alterations of the skin can be acquired by measurements of the surface skin topology 1. Alterations of human skin may have pathological reasons, or may be caused by pharmaceutics and cosmetics. Those changes of the skin surface topology 2 have to be detected and quantitatively measured in an early state. In addition, temporal changes have to be observed and documented. In opposition to technical surfaces, in vivo measurements of the skin topology exhibit additional difficulties. During a measurement small movements caused by pulse, respiration, muscle contraction and trembling of the patient may occur. These disturbances lead to errors of the measured topology. To reduce this effect many methods are based on the measurements of silicone replicas of the skin 6. The making of the replica itself causes artifacts in the topology. Furthermore, the making of the replica affects the skin. Therefore comparative studies are not reliable. Optical coherence profilometry by coherence radar enables in vivo 3D shape acquisition of the human skin without any preparation of the surface. The measuring uncertainty is mainly limited by the roughness of the object itself and not by parameters of the sensor (e.g., observation aperture). 2. MEASURING PRINCIPLE The coherence radar is basically a Michelson interferometer (Fig. 1a). The skin and the reference mirror are illuminated by a plane wave. The scan in depth is done by moving the surface of the skin along the optical axis. In the image plane, interference can be observed for those parts of the object surface that are virtually located close to the reference plane, where the object s optical path length is equal to the reference path length. A typical correlogram is shown in Figure 1b. The correlogram has to be individually evaluated in each speckle. Each correlogram has its

2 individual (random) phase and a statistical offset from the true position 5. Hence, we can only acquire information from the contrast or from the maximum. The physical measuring uncertainty (offset) is essentially equal to the micro-roughness of the surface. It does not depend on sensor properties, such as aperture and object distance. This is completely different from classical interferometry. That is why we gave the system a new name coherence radar. Fig. 1: a) Experimental setup of the coherence radar; b) correlogram of interference intensity versus z position 3. THE FIBEROPTICAL COHERENCE RADAR The conventional setup of the coherence radar, developed for technical objects, is fixed on a bulky translation stage. The human body to be measured has to be brought to the sensor. This limits the area available for measurements. In addition, the patient has to be motionless to avoid movements during a measurement. Macroscopic movements of the patient can destroy a measurement by introducing a pseudo contrast. Therefore, in vivo measurements were possible up to now only in the labratory. We implemented a fiber optical implementation of the coherence radar with a translation stage separated from the sensor head. Now only the ssmall sensor head has to be brought to the patient and not vice versa. Single mode fibers are used in the interferometer. The surface of the skin is imaged onto the CCD camera and within one single scan the whole surface gets measured. For the separation of the translation stage from the sensor head a fiber has to be implemented in the reference arm of the interferometer (Fig. 2). To ensure equal optical path length in object and reference arm a second fiber has to be placed into the object arm. This is done by inserting an additional beam splitter (BS1 in Fig. 2) which splits the incoming light up into the reference and object beam before the main beam splitter of the coherence radar (BS2). That setup is adequate to a combination of a Mach-Zehnder interferometer with a Michelson interferometer. Since the depth scan is realised by a translation stage in the reference arm of the Mach-Zehnder interferometer the sensor head gets mobile and can be miniaturized.

3 BS1 SLD Object arm C O1 C O2 C R1 Sensor head K R2 mounted on a translation stage C R2 C R3 C O3 BS2 CCD - camera Reference arm Scan of reference plane Object Fig. 2: Setup of the fiber optical coherence radar Figure 3 illustrates the fiber optical implementation of the coherence radar. The sensors consists of two spatially separated components: the sensor head, and a mudule containing the translation stage. Both parts are connected by fibers. The light source is a SLD. An additional pilot laser in the visible range is used for aiming. SLD aiming laser diode x- coupler single mode fiber polarisationcontroller 80% fibercoupler OBJECT ARM 20% collimator linear translation stage mobile sensor head object Fig. 3: Fiber optical implementation of the coherence radar The scan in z-direction is done in the reference arm of the Mach-Zehnder interferometer by varying the optical path length between the two collimators C R1 and C R2. Iris diaphragms between the collimators in the reference arm and the object arm control the intensity of the reference beam and the intensity on the objects surface. The sensor head contains the main part of the coherence radar (Fig. 4) for in vivo measurements. A beam splitter

4 superimposes the parallel reference beam with the light scattered back from the surface of the skin. The measurable area of skin surface depends on the output power of the light source and on the reflectivity of the surface. Presently the field of view is 5 x5mm². During the whole measurement the object under test has a constant distance to the imaging system since the scan is done in the reference arm of the Mach-Zehnder interferometer. Hence the full depth of the object has to be in the range of focus of the imaging system. Fig. 4: Sensor head Figure 5 displays gray-level encoded height maps of in vivo measurements with the fiber optical coherence radar. The field of view is 2,5 x2,5 mm². Both measurements were performed without any preparation of the skin and a z-speed of 4µm/s. At the bottom of Figure 5 the profile of the surface along the line in the picture is displayed. The picture to the left shows the wrinkles at the skin of a finger tip which have a depth of about 80 µm. The measurement of the underarm on the right (Fig. 5) displays the topology of deep wrinkles about 80 µm in depth and shallow wrinkles about 30 µm in depth. Fig. 5: In vivo-ocp measurements with the fiber optical coherence radar

5 Figure 6 shows further measurements of the human skin. These preliminary measurements are the first made by holding the sensor head in the hand and pressing it against the surface. To reduce the measuring time and to avoid movements during the measurements the light source was focused on the surface of the skin to increase the intensity on the surface. This leads to a small field of view of just 1,4 x1,4 mm² but on the other hand we are able to use the shutter method and scan with a speed of 14µm/s. By this we can neglect involuntary body movements during the scan. The picture on the left shows a measurement on the nose. In the profile below (along the line in the grey-level encoded picture) you can see a smooth surface interrupted by a peak, a 50µm small pimple. The picture on the right shows the surface of the human skin at the belly. The typical net like structure of the skin with deep wrinkles connected by shallow wrinkles can be recognized. The quality of the data will be improved in the near future by better light sources. Fig. 6: In vivo OCP measurements: sensor head held in the hand 4. LATERAL MOVEMENT COMPENSATION One way to reduce the influence of involuntary body movements during a measurement, is to increase the speed of the scan in z-direction. However, any trembling might afflict the measurement. Hence, we developed two movement compensations one for longitudinal movements and a second for lateral motions. We introduced a modified setup of the coherence radar for longitudinal movement compensation 1. It is basically a Michelson interferometer with the reference mirror fixed on the patients skin next to the field of view. Consequently each longitudinal movement changes the path length of object arm and reference arm in the same way and the measurement is not disturbed.

6 reference mirror object In the case of lateral movements of the skin during the measurement, a camera pixel may see dark and bright areas of the measured field alternately, because of different local reflectivity. As the coherence radar looks for the highest signal contrast this false contrast (pseudo contrast) may destroy the measurement. LED L 1 L 2 cam 2 cam 1 L 1 L 2 reference plane scanning of the reference plane To compensate for lateral movements of the patient, another modification of the optical setup (Fig. 7) was developed which exploits the second interferometer output too, with a second CCD camera (cam2). The system evaluates the difference of corresponding camera pixels of the two cameras in the same video cycle. As the second CCD camera sees the same image but with a π phase shift, the difference of the two camera signals is zero, except for those pixels where the object wave interferes with the reference wave. mounted on the linear translation stage Fig. 7: Coherence radar with lateral movement compensation Figure 8 shows the measurement of a laterally moved object with (on the left) an without (on the right) movement compensation. The object was laterally moved by 500µm (20 pixels). The compensated measurement displays no pseudo contrast. Fig. 8: Measurement of a german coin with (right) and without (left) movement compensation 5. MICROSCOPIC APPLICATIONS The fiber optical coherence radar was developed for measurements on human skin with low magnification. The field of view has been adjusted to a diameter of 1,4 mm to detect the typical structures of the skin. A higher magnification is necessary for object topologies with fine structures like the human hair. In Figure 9 we see a measurement of a human hair made with a setup for microscopically applications. The hair on the left side is slightly damaged. In the profile you can the a rough surface topology. The healthy hair on the right side shows a more regular height profile. Under high magnification the 3D surface structure can be examined and the influence of cosmetic treatment might be tested.

7 Fig. 9: Measurement on human hair: damaged hair (left) healthy hair (right) with a regular structure 6. CONCLUSION The fiber optical coherence radar is a mobile sensor for the acquisition of 3-D topology data of rough surfaces. The separation of a mobile sensor head from the translation stage allows the measurement of the surface topology of human skin even at areas difficult to reach. The separation is done by a combination of two interferometers. Not like the standard coherence radar, now the object depth has to be in the range of focus, since there is no depth scanning. The wrinkles in the human skin have a depth of about 100 µm and can be imaged with sufficient depth of focus. By scanning with high speed involuntary body movements of the patient can be neglected. These factors enable us to measure the human skin within a few seconds and an uncertainty in the range of 2 microns. With a coherence radar for microscopic applications, measurements of small details like the human hair are possible. Further improvements will be achieved by better light sources. 7. ACKNOWLEDGEMENT We acknowledge for suggestions and support of the project supplied by Dr. F. Kiesewetter, Dermatologische Universitätsklinik mit Poliklinik, Erlangen. This paper is funded by the BMBF, registration 13N REFERENCES 1. G. Ammon, P. Andretzky, G. Bohn, G. Häusler, J. M. Herrmann, M. W. Lindner, Optical coherence profilometry (OCP) of human skin in vivo, SPIE, Vol. 3196, G. Häusler and M. W. Lindner, Coherence radar and spectral radar - new tools for dermatological diagnosis, Journal of Biomedical Optics 3(1), 21-31, January P. Ettl, B. Schmidt, M. Schenk, I. Laszlo, g. Häusler, Roughmess parameters and surface deformation measured by coherence radar, International Conference on Applied Optical Metrology, Balatonfüred, Hungary, SPIE, Vol. 3407, Th. Dresel, G. Häusler, and H. Venzke, Three-dimensional sensing of rough surfaces by coherence radar, Appl. Opt., Vol.31, , Thomas Dresel, Gerd Häusler, and Holger Venzke, Three-dimensional sensing of rough surfaces by coherence radar, App. Opt., Vol. 31, No. 7, U. Hoppe, R. Lunderstädt, G. Sauermann, Quantitative Analyse der Hautoberfläche mit Hilfe der digitalen Signalverarbeitung, Ärztl. Kosmetol 16, 13-37,1986