3D In-Vivo Optical Skin Imaging for Topographical Quantitative ...

YOUNG INVESTIGATOR’S AWARD WINNER This study nicely quantitates the clinical effects of non-ablative laser treatments, i.e. increased skin smoothness reported by patients. However, readers should not interpret these results as validating significant visual improvement in either scars or wrinkles. Also noteworthy is that skin hydration from application of EMLA cream provided significantly more improvement in fine facial lines than following five non-ablative laser treatments. William P. Coleman, III, MD

3D In-Vivo Optical Skin Imaging for Topographical Quantitative Assessment of Non-Ablative Laser Technology Paul M. Friedman, MD,* † Greg R. Skover, PhD, ‡ Greg Payonk, PhD, ‡ Arielle N. B. Kauvar, MD,* and Roy G. Geronemus, MD* *Laser and Skin Surgery Center of New York, New York, NY, †DermSurgery Associates, Houston, TX, and ‡Skin Research Center, Johnson and Johnson Consumer Companies, Skillman, NJ

background. A new method for treating facial rhytides and acne scars with nonablative laser and light source techniques has recently been introduced. Given the inherent limitations of photographic and clinical evaluation to assess subtle changes in rhytides and surface topography, a new noninvasive objective assessment is required to accurately assess the outcomes of these procedures. objective. The purpose of this study was to measure and objectively quantify facial skin using a novel, noninvasive, Invivo method for assessing three-dimensional topography. This device was used to quantify the efficacy of five treatment sessions with the 1064 nm QS Nd:YAG laser for rhytides and acne scarring, for up to six months following laser treatment. methods. Two subjects undergoing facial rejuvenation procedures were analyzed before and after therapy using a 30-mm, three-dimensional microtopography imaging system (PRIMOS, GFM, Teltow, Germany). The imaging system projects light on to a specific surface of the skin using a Digital Micromirror Device (DMD Texas Instruments, Irving, TX) and records the image with a CCD camera. Skin Surface microtopography is reconstructed using temporal phase shift algorithms to generate three-dimensional images. Measurements were taken at baseline, at various times during the treatment protocol, and then at three and six-month follow-up visits. Silicone skin replicas (FLEXICO, Herts, England) were also made before and

after the laser treatment protocol for comparison to In-vivo acquisition. results. Skin roughness decreased by 11% from baseline after three treatment sessions in the wrinkles subject, while a 26% improvement of skin roughness was recorded by 3D Invivo assessment six months following the fifth treatment session. The subject with acne scarring demonstrated a 33% decrease in roughness analysis after three treatment sessions by 3D In-vivo assessment. A 61% improvement in surface topography was recorded 3-months following the fifth treatment session, which was maintained at the 6-month follow-up. conclusion. Three-dimensional In-vivo optical skin imaging provided a rapid and quantitative assessment of surface topography and facial fine lines following multiple treatment sessions with a 1064-nm QS Nd:YAG laser, correlating with clinical and subjective responses. This imaging technique provided objective verification and technical understanding of nonablative laser technology. Wrinkle depth and skin roughness decreased at the three and six-month follow-up evaluations by 3D In-vivo assessment, indicating ongoing dermal collagen remodeling after the laser treatment protocol. Future applications may include comparison of nonablative laser technology, optimization of treatment regimens, and objective evaluation of other aesthetic procedures performed by dermatologists.

Background

brasion, chemical peels, and ablative laser resurfacing.1–7 Nonablative laser and light source techniques have recently been introduced as a treatment that selectively heats the upper dermis, inducing a wound healing response in the papillary and upper reticular dermis without epidermal ablation.8–12 Histologic studies have shown that new collagen production and deposition results from such procedures, and that removal of the epidermis and

A VARIETY OF methods have been used to treat facial rhytides associated with photoaging including dermaAddress correspondence and reprint requests to: Paul M. Friedman, MD, Director of Laser Surgery, DermSurgery Associates, 7515 Main, Suite 240, Houston, TX 77030, or email: [email protected]

© 2002 by the American Society for Dermatologic Surgery, Inc. • Published by Blackwell Publishing, Inc. ISSN: 1076-0512/02/$15.00/0 • Dermatol Surg 2002;28:199–204

200

friedman et al.: 3-d-in-vivo skin imaging

portions of the dermis are not required for neocollagenesis and collagen remodeling.8,10 Improved skin texture and turgor have been reported by patients and physicians, but has been difficult to quantify given the inherent limitation of photographic and clinical evaluation. Optical profilometry has been shown to measure wrinkle depth by quantifying contours of silicone rubber replicas, capturing surface topography,13,14 and may also serve as a valid and reliable method for determining an improvement in acne scarring.4 However, it is highly operator-dependent and may produce a variety of artifacts. The type of replica material used may accentuate or diminish the actual skin surface topography. Mixing of the replica material may enhance or delay the speed of curing altering the impression at the various time points. Polymerization of the material may alter the skin’s microtopography. Material curing may change altering the recovery of precise skin replicas. Furthermore, positioning of the patient in a supine position is required to provide a horizontal surface permitting the curing of the material from a liquid to a solid state producing a gravitational artifact. Direct three-dimensional In-vivo skin imaging can minimize these and other inherent artifacts associated with optical profilometry. To acquire an exact 3D impression of skin microtopography the technique used must achieve high lateral resolution with minimal vertical interference, or noise, enabling accurate reconstruction of images. Images must be acquired rapidly without altering the surface. The system must be able to 1. Adapt to highly contoured surfaces without losing resolution, 2. Calibrate with ease and ensure accurate measurement of reconstructed images and 3. Identify suitable landmarks permitting comparative image assessment. The PRIMOS optical three-dimensional In-vivo skin measurement device (PRIMOS, GFM, Tetlow, Germany) used in this study deploys a parallel stripe pattern imaging technique that is projected onto the skin surface and depicted on the CCD chip of a high resolution camera. Light patterns are created by a digital micromirror projector (Texas Instruments, Irving TX). The use of micromirror-based digital light projectors is advantageous when applied to optical 3D In-vivo skin measurement because the light intensity is high, exposure time is short and the light can be controlled point and/or pixelwise.15 The 3D effect is achieved by means of minute elevation differences on the skin surface deflecting the parallel projection stripes to produce a qualitative and quantitative measurement of the skin’s profile. Images are digitized and transferred for computer assisted quantitative evaluation and measurement. Mathematical algorithms embedded in the analytical software reconstruct the data into a highly precise 3D profile of the skin surface. The vertical and lateral resolution of a stripe projection measurement device such as the PRIMOS opti-

Dermatol Surg

28:3:March 2002

cal system are essentially determined by the field of view (FOV) used, the number of pixels of the recording camera and the accuracy of the determination of the smallest stripe deflections that can be processed. A highly precise area recovery was developed for the PRIMOS device that makes it possible to realize the measurement position on the skin surface with a precision of 1/10 pixel. Therefore, for the measurement field of 30 24 mm, using a 640 480 pixel CCD, positioning precision of 3 m can be attained.

Methods The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and approved as a prospective clinical trial by the Essex Institutional Review Board, Inc., Lebanon, NJ. Two subjects were evaluated by three-dimensional microtopography to demonstrate how quantitatively measurable changes in the skin following nonablative laser technology could be assessed with the PRIMOS device. Subject 1 had class III rhytides and skin phototype II, and was evaluated from October 2000 through August 2001. Images were acquired prior to therapy, before and after application of topical anesthetic under occlusion, immediately after one laser treatment session, and at specified follow-up visits. Furthermore, at the completion of the trial the subject was imaged in the same location one month following BOTOX (Allergan Inc., Irvine, CA) injections. Subject 2 had mild atrophic acne scarring and skin phototype II. 3D In-vivo microtopography and silicone rubber impressions were taken before treatment, at various times during treatment, and three and six months following the completion of the laser treatment protocol.

Laser Treatment The subjects were treated five times at 2–3 week intervals. A topical anesthetic (EMLA cream, Astra USA, Westborough, MA) was applied under occlusion for one hour prior to each treatment. Immediately before treatment, the EMLA was washed off the treatment area and followed by gentle cleansing with alcohol. The Q-Switched Nd:YAG laser (Medlite IV, Continuum, Santa Clara, CA) ( 1064 nm) was used with a spot size 6 mm, fluence 3–3.5 J/cm2. Multiple passes with the laser were performed to the periorbital and perioral regions until a clinical end point of erythema was obtained. The subject with acne scarring received multiple laser passes to both cheeks, extending from the nasolabial fold to the preauricular area and jawline until a clinical end point of erythema was obtained. The subjects applied a sunscreen of SPF30 or higher daily to the treatment areas.

Clinical Photography Photographs were taken before and after the procedure with a Nikon N6006 camera Nikkor 60 mm F2.8 lens and CSI Twin

Dermatol Surg

28:3:March 2002


201

Flash (Canfield Scientific, Fairfield, NJ) using KODAK Kodachrome film. Replicate photographs were taken from 0, 45 (right) and 45 (left) using a standardized reproduction ratio of 1 : 6 (f/16) for full facial photographs and 1 : 3 (f/22) for close-up photographs. The subject was positioned in a CSI head restraint enabling comparative photographs to be taken throughout the study (Canfield Scientific, Fairfield, NJ). Pre and post treatment photographs were assessed by two physicians and graded on a four point scale where (0% change from baseline was graded as no improvement, 1–25% change graded as mild improvement, 26–75% change graded as a moderate improvement, 75–100% a marked improvement).

Skin Surface Topography Patients were positioned in front of the light projector and the head position fixed and recorded using a head restraint. Three-dimensional microtopography was performed with the PRIMOS 30 mm imaging system (PRIMOS Imaging System, G.F.M. Teltow, Germany). A digital micro mirror device (Texas Instruments, Irving TX) creates a parallel stripe pattern imaging technique that is projected onto the skin surface and depicted on the CCD chip of a high-resolution camera. Images are digitized and transferred for computer assisted quantitative evaluation and measurement. Mathematical algorithms embedded in the analytical software reconstruct the data into a highly precise 3D profile of the skin surface. Algorithms contained in the evaluation software permit Star Roughness to be calculated from the acquired surface profile. The average of 16 profile lines arranged in a radial array were used to assess the surface. Two different roughness equations were used to analyze the data. Roughness (Ra) is the arithmetic average of the absolute values of all points of the profile. In other words, Ra is the height of the rectangle with the same length and surface as the profile encloses in the specified sector. The value yields an impression of surface smoothness therefore was used to analyze surface images taken from subjects with acne scarring. On the other hand, Roughness (Rz) is the mean peak to valley height and is the arithmetic average of the maximum peak to valley height of the roughness values Y1 to Y5 of 5 consecutive sampling sections over the filtered profile. Alternatively, this value yields an impression of the prominence of wrinkles that are the greatest contributors effecting the surface deflection in the crow’s feet area of patients with photodamage.16

Silicone Impressions Silicone skin replicas (FLEXICO, Herts, England) were made before and after the laser treatment protocol to measure surface texture and topography. Impressions were taken from representative areas on the cheek from patient 2 with acne scarring. Analysis was performed with the PRIMOS 30 mm imaging system. The light projector was rotated 90 degrees perpendicular to mounting stand stage. The replica was positioned underneath the light source and the image acquired in a similar manner to the in-vivo acquisition.17

Figure 1. Roughness Analysis (Rz) of photodamaged skin following nonablative laser resurfacing. Images from the periorbital area were captured with the PRIMOS 30 mm imaging system (PRIMOS Imaging System, G.F.M. Teltow, Germany). Images were digitized and saved for computer assisted quantitative evaluation and measurement. Algorithms contained in the evaluation software permit Roughness to be calculated from the acquired surface profile. Sixteen profile lines arranged in a radial display were used to compute the average surface roughness. Roughness (Rz) is the mean peak to valley height and is the arithmetic average of the maximum peak to valley height of the roughness values Y1 to Y5 of 5 consecutive sampling sections over the filtered profile measured in micrometers. Time points: Baseline: pretreatment, Post EMLA: measurement after one hour of topical anesthetic under occlusion before laser treatment, Post Laser: immediately following laser procedure, Mid Tx: measurement before the fourth laser treatment, 6 month FU: six months after five laser treatments, 1 month Botox: 1 month after BOTOX injections.

Results A reduction in the Rz value occurred following application of the topical anesthetic under occlusion for 1 hr, likely from moisture retention in the skin (Figure 1). An increase in Rz was observed immediately following the laser treatment consistent with mild abrasion. Rz decreased by 11% from baseline after three treatment sessions, while a 26% improvement of surface topography was recorded six months following the fifth treatment session. BOTOX injections in the periorbital and glabela areas were performed after completion of the study protocol, and In-vivo imaging a month later demonstrated a further 15% decrease in skin roughness (Figure 1). The subject with acne scarring demonstrated a decrease in Ra at each follow-up visit by In-vivo assessment (Figure 2). Ra decreased by 33% after three treatment sessions. A 61% improvement was recorded 3-months following the fifth treatment session which was maintained at the 6-months follow-up. Compari-

202


Figure 2. Roughness Analysis (RA) of acne scarred skin following nonablative laser resurfacing. Images from the cheek were captured with the PRIMOS 30 mm imaging system (PRIMOS Imaging System, G.F.M. Teltow, Germany). Silicone replicas were taken from the same area. Both the In-vivo and replica images were digitized and saved for quantitative measurement. Algorithms contained in the evaluation software permit Roughness to be calculated from the acquired surface profile. Sixteen profile lines arranged in a radial display were used to compute the average surface roughness. Roughness (Ra) is the arithmetic average of the absolute values of all points of the profile. In other words, Ra is the height of the rectangle with the same length and surface as the profile encloses in the specified sector. Time points: Baseline: pretreatment, Mid Tx: measurement before the fourth laser treatment, 3 and 6 month FU: three and six months after five laser treatments.

Dermatol Surg

28:3:March 2002

Figure 4. Photograph of subject with acne scarring three months after the end of five treatments.

up. In contrast, replica analysis demonstrated an 8–18% improvement of surface topography at the follow-up visits (Figure 2). Evaluation of clinical photography did not clearly demonstrate these changes in skin topography (Figure 3,4). The high-resolution camera images obtained with the PRIMOS camera before and after treatment (Figure 5,6) clearly demonstrates the improvement of sur-

son of In-vivo vs. replica impression Ra acquisition was performed utilizing the PRIMOS imaging device. The In-vivo surface changes correlated with the clinical assessment and subjective responses of a 25–50% improvement after three treatment sessions, and a greater than 50% improvement at the 3- and 6-months follow-

Figure 3. Photograph of subject with acne scarring before treatment.

Figure 5. High resolution camera image of subject with acne scarring before treatment.

Dermatol Surg

28:3:March 2002


203

Figure 6. High resolution camera image of subject with acne scarring three months after the end of five treatments.

face topography in this patient. Color coded height display of surface topography show smoother skin texture 3-months after the five treatment sessions by the evenness in color distribution with greater amount

Figure 8. Color coded surface topography image of subject with acne scarring three months after the end of five treatments (Ra 37.0 m).

of color closer to the 0 height after the laser treatments (Figures 7 and 8).

Discussion

Figure 7. Color coded surface topography image of subject with acne scarring before treatment (Ra 90.4 m).

The study of nonablative laser technology for facial fine lines and acne scarring is a new area of clinical research for which appropriate evaluation criteria are being determined. Conventional methods of evaluation have included photographic and clinical assessment, which have inherent limitations. Computerized image analysis of silicone replicas has been shown to be a reproducible, objective technique for measuring skin topography.18 Three-dimensional In-vivo imaging provides a real time, objective analysis of patient specific characteristics potentially enabling clinicians to predict the limitations and efficacy of various aesthetic procedures. Consequently, we sought to complement the subjective reporting with an objective technique that could reproducibly measure changes in surface topography. After 3–5 treatment sessions with a QS Nd:YAG laser, we quantified improvement of surface topography using 3D In-vivo skin imaging in subjects with acne scarring and facial fine lines. Surface topography and facial fine lines were effectively captured by the PRIMOS imaging system initially, during the treatment sequence and months after therapy. Evaluation of suc-

204


cessive images suggests that the method provides rapid and reliable data that can document outcomes throughout the rejuvenation process. Subjects were assessed over 11 months, with photography, skin replicas and 3D In-vivo imaging. The subjects showed successive improvement at the end of the treatment sequence continuing through the subsequent evaluation time points, suggesting ongoing dermal collagen remodeling. The greatest percentage improvement was observed with In-vivo imaging at three months following the treatment protocol. The improvement plateaued and held constant at the final six-month evaluation. 3D In-vivo optical imaging provided higher resolution of skin topography than analysis by clinical photography or skin replicas changes, and were in agreement with clinical assessment and patient subjective responses. This data corroborate results presented by Boyes et al. demonstrating a substantial correlation between subjective clinical assessment and objective 3D In-vivo imaging following nonablative laser procedures.19 These results suggest that 3D In-vivo skin imaging provides a rapid and accurate method to quantify measurable changes in subjects with acne scarring and photodamage. This imaging technique also provided an objective verification and technical understanding of this evolving technique. The value of this imaging technique is currently being evaluated further in a larger trial being performed at our center. Future application of PRIMOS imaging of the microtopography of human skin20 may provide an objective measure for comparison of nonablative laser technology and post-treatment topical regimens. This device may also allow for the objective evaluation of other aesthetic procedures performed by dermatologists, including soft tissue augmentation, BOTOX, and topical antiaging medications. Furthermore, accurate assessment of facial topography may enable clinicians to achieve desirable affects while minimizing complications. This technique may also provide an objective modality for the cosmetic and pharmaceutical industry to evaluate the efficacy of rationings, alpha hydroxy acids, beta hydroxy acids, and antioxidants for their effects on facial wrinkling. Acknowledgments The authors would like to express our sincere appreciation to Marisol Edward, Judy Dulberg, Alexis Moreno, and Michelle Turnbull of the Laser & Skin Surgery Center of New York Research Department for their clinical assistance. Furthermore, to Dick Jackson, D-Jackson

Dermatol Surg

28:3:March 2002

Software Consulting Calgary, Canada, for his invaluable expertise in developing the software for the 3D image processing and analysis.

References 1. Waldorf HA, Kauvar AN, Geronemus RG. Skin resurfacing of fine to deep rhytides using a char-free carbon dioxide laser in 47 patients. Dermatol Surg 1995;21:940–6. 2. Fitzpatrick RE, Goldman MP, Satur NM, et al. Pulsed carbon dioxide laser resurfacing of photoaged skin. Arch Dermatol 1996;132: 395–402. 3. Hruza GJ. Skin resurfacing with lasers. Fitzpatricks J Clin Dermatol 1995;3:38–41. 4. Jordan R, Cummins C, Burls A. Laser resurfacing of the skin for the improvement of facial acne scarring: a systematic review of the evidence. Br J Dermatol 2000;142:413–23. 5. Alster TS, West TB. Resurfacing of atrophic facial acne scars with a high-energy, pulsed carbon dioxide laser. Dermatol Surg 1996;22: 151–5. 6. Laufman R, Hibst R. Pulsed erbium. YAG laser ablation in cutaneous surgery. Lasers Surg Med 1996;19:324–30. 7. Kye YC. Resurfacing of pitted facial scars with a pulsed Er: YAG laser. Dermatol Surg 1997;23:880–3. 8. Kelly KM, Nelson JS, Lask G, Geronemus RG, Bernstein LJ. Cryogen spray cooling in combination with nonablative laser treatment of facial rhytids. Arch Dermatol 1999;135:691–4. 9. Zelickson BD, Kilmer SL, Bernstein E, et al. Pulsed dye laser therapy for sun damaged skin. Lasers Surg Med 1999;25:229–36. 10. Menaker GM, Wrone DA, Williams RM, Moy RL. Treatment of facial rhytids with a nonablative laser: a clinical and histologic study. Dermatol Surg 1999;25:440–4. 11. Goldberg DJ, Metzler C. Skin resurfacing utilizing a low fluence Nd: YAG laser. J Cutan Laser Ther 1999;1:23–7. 12. Goldberg DJ, Whitworth J. Laser skin resurfacing with the Q-Switched Nd: YAG laser. Dermatol Surg 1997;23:903–7. 13. Marks R. Methods for the assessment of the effects of topical retinoic acid in photoaging and actinic keratoses. J Intern Med Res 1990;18 (Suppl. 3:29) C–34. 14. Grove GL, Grove MJ, Leyden JJ. Optical profilometry. an objective method for quantification of facial wrinkles. J Am Acad Dermatol 1989;21:631–7. 15. Jaspers S, Hopermann H, Sauermann G, Hoppe U, Lunderstädt R, Ennnen J. Rapid in vivo measurement of the topography of human skin by active image triangulation using a digital micromirror device. Skin Res Technol 1999;5:195–207. 16. Schreiner V, Sauermann G, et al. Characterization of the Skin Surface by ISO Parameters for Microtopography. In: Wilhelm, K-P, Elsner, P, Berardesca, E, Maibach, HI, eds. Bioengineering of the Skin: Skin Surface Imaging and Analysis. CRC Press, Inc, Boca Raton, FL., 1997. 17. Payonk GS, Kollias N, Poster Exhibit AAD 59th Annual Meeting, Washington, D.C., March, 2001. 18. Grove GL, et al. Skin replica analysis of photodamaged skin after therapy with tretinoin emollient cream. J Am Acad Dermatol 1991; 25:231–7. 19. Boyes LE, Goldman MP, Payonk GS, Skover GR. American Society for Laser Medicine and Surgery Twenty-First Annual Meeting, New Orleans, LA, April, 2001. 20. Hof C, Hopermann H. Comparison of Replica- and In Vivo-Measurement of the Microtopography of Human Skin, Report der Universität der Bundeswahr Hamburg, Institut für Automatsierungstechnik, 2000.