Measurement of Transcutaneous Hemoglobin Concentration by Noninvasive White-Light Spectroscopy in Infants
http://www.100md.com
《小儿科》
Department of Neonatology, Brighton and Sussex University Hospitals NHS Trust, Brighton, United Kingdom
Department of Neonatology, Children’s University Hospital, Münster, Federal Republic of Germany
Department of Neonatology, Institute of Cancer Research, University of Witten/Herdecke, Herdecke, Federal Republic of Germany
ABSTRACT
Objective. To compare transcutaneously spectroscopically measured hemoglobin values with venous hemoglobin values in infants.
Study Design. Prospective study in healthy preterm and term infants who were breathing spontaneously.
Results. Recordings were obtained from 85 stable infants (median gestational age at measurement: 36 weeks [range: 34–43 weeks]; median body weight: 1890 g [range: 1095–4360 g]). The spectroscopic hemoglobin values were corrected for inhomogeneous distribution of hemoglobin in the tissue. The venous and spectroscopic hemoglobin values were then compared by using the Bland-Altman method, which gave an error of <5%.
Conclusions. This pilot study could illustrate a good relation between the 2 methods for measuring hemoglobin. Larger studies are required to validate the spectroscopic method in those with conditions that affect the skin microcirculation (eg, septicemia).
Key Words: skin microcirculation phlebotomy iatrogenic blood loss noninvasive method neonate
Anemia of prematurity is a common problem in the care of preterm infants that requires frequent donor packed red cell transfusions for treatment.1 Iatrogenic blood losses caused by venipuncture are a major cause of anemia in premature infants.2 Hematopoiesis cannot produce enough new erythrocytes to balance the loss. Therefore, new noninvasive techniques for estimating serum bilirubin, for example, have been introduced into neonatology to reduce iatrogenic blood loss.3,4 We report on a new noninvasive spectroscopic technique that measures hemoglobin quantitatively and transcutaneously in preterm and term infants. In this preliminary study, the spectroscopically measured hemoglobin values are compared with the results of venous sampling in the same infants.
METHODS
The Device
The visible and near-infrared spectroscopic device (Mediscan 2000; MBR Messtechnik, Herdecke, Federal Republic of Germany; patent number DE 19831424, German Patent Office, Munich, Federal Republic of Germany) uses halogen white light that is emitted through 6 silica fibers (200-μm diameter) onto the skin. The reflected light, which has traveled through the skin tissue, is then collected via a silica fiber of 200-μm diameter and analyzed for spectra in the range of 350 to 1020 nm. The spectra are calibrated before each recording. Thus, in contrast to conventional near-infrared spectroscopy, the whole frequency range of light is analyzed. The distance between the outer surface of the 6 emitting fibers and the central receiving fiber is 250 μm (SD: 4 μm). The separation of the fibers is achieved by the individual cladding of the fibers. The technical principles of the method have been published elsewhere.5 Quantitative measurements of carotenoids in the skin of adults have been performed by using this technique.6–8
Mathematical Background of the Method
In brief, the method works as follows. The anatomy of the capillaries in infants shows an inhomogeneously distributed pattern.9 The hemoglobin is inhomogeneously distributed in the capillaries and the tissue because of changes in the distribution of the capillary perfusion.10
The first step of the method is to calculate the product of concentration and path length of the light between the light emitter and the light-receiving fiber. In a second step, the concentration of hemoglobin is calculated. Both steps are performed with the help of an iterative process.
The first step is performed in the following way: If one calculates the product of concentration and path length from direct measurements, the resulting value will be low. The reason for this is the inhomogeneous distribution of hemoglobin in tissue. The inhomogeneous distribution of hemoglobin is a geometric property, which differs from 1 measurement site to another. Therefore, the degree of inhomogeneity is determined individually at each measurement site. If the degree is known, the path-length concentration factor of homogeneously distributed hemoglobin is calculated.
The degree of inhomogeneous distribution is calculated in the following way: In the first step, the mapping from homogeneously distributed hemoglobin to the measured inhomogeneously distributed hemoglobin is calculated. Then the nonlinear fitting algorithm of Levenberg-Marquardt, a polynomial of third order, is fitted to this mapping. Then the homogeneous distribution of hemoglobin is changed step by step to an inhomogeneous one. At each step the mapping is calculated. If the mapping becomes linear, it is then a polynomial of first order; the 2 distributions are equal, and the degree of inhomogeneity is known. Now the true product of concentration and path length is known.
In the second step, the true concentration of hemoglobin is calculated. Because the degree of inhomogeneity of the hemoglobin is now known, then a 2-component system can be used. One component is the hemoglobin, and the other component is the hemoglobin-free tissue.
It was shown by Wodick and Lübbers11 that the spectra are disturbed only by the distribution of hemoglobin and not by the special arrangement of the distribution. All different arrangements of hemoglobin with the same degree of inhomogeneity affect the spectra in the same way. From the hemoglobin compartment all optical properties are known: absorption, scattering, and the anisotropy coefficient (optical properties for the whole blood). The path length of light in both compartments is the same. The degree of inhomogeneity of tissue distribution is known (which is 1 minus the degree of inhomogeneity of hemoglobin). With the fit of Levenberg-Marquardt, it is possible to determine the optical properties of tissue from the measured spectrum for a given path length. This measurement is performed in the following way: in the capillary bed we found oscillating of hemoglobin. It is a well-known phenomenon described by many methods that hemoglobin in the capillary bed is detected with an oscillating pattern.
With the method described above, 2 different hemoglobin concentrations and path lengths are measured at the same measurement site (higher and lower hemoglobin content in tissue). Then the difference of hemoglobin is calculated. The difference affects the path length in tissue and, therefore, the path length in hemoglobin-free tissue. From the hemoglobin differences the path-length changes in tissue are calculated, and these processes are performed repeatedly on the path lengths until the calculated differences and the error of the measured differences in the hemoglobin-free tissue component becomes minimal (iterative process). Finally, the best path-length difference is found. From the ratio of tissue path length/tissue path length of the difference, the tissue path length of the light is calculated. Then the concentration is calculated from the well-known product of path length and concentration.
The variability of repeated transcutaneous hemoglobin measurements has been reported as <5%.12,13
Measurements
The present pilot study was aimed at quantitative measurement of the skin hemoglobin in preterm and term infants and a comparison with venous values. The venous hemoglobin was taken weekly according to unit protocol. The spectroscopic measurements were performed as part of a larger study on skin microcirculation in preterm and term infants. No blood was taken for study purposes alone. Therefore, only 85 comparative sets of data were obtained. The spectroscopic measurements were performed by a handheld probe on the forearm of the infants. All infants were in a stable condition (with normal blood pressure for age and a capillary refilling time of < 3 seconds at the finger tip) and were measured at the same time in the late afternoon or evening after having been feed. This setting depended on the availability of the research staff and did not disturb the routine work of the ward. Exclusion criteria were clinical or laboratory signs of infection, congenital malformations, and requirement of oxygen or artificial ventilation. The spectroscopic baseline values of the total hemoglobin in the skin were obtained for 1 minute, and the mean values were calculated. The venous samples for measuring the hemoglobin were processed by a standard Coulter counter (K-1000; Sysmex GmbH, Düsseldorf, Federal Republic of Germany). The hemoglobin values then were compared by using the Bland-Altman method.14
The study was approved by the hospital ethics committee. Informed consent was obtained from the parents before the study.
RESULTS
Eighty-five infants were enrolled into this study. The median birth weight of the infants was 1040 g (range: 600–4160 g), and the median gestational age was 33 weeks (range: 24–41 weeks). The study was performed at a mean corrected age of 36 weeks (range: 34–43 weeks). The median body weight of the infants was 1890 g (range: 1095–4360 g). The median time between the blood sampling and spectroscopic measurements was 11 hours (range: 9–14 hours). All infants were studied only once.
The comparison according to Bland and Altman14 of venous hemoglobin and the spectroscopically obtained hemoglobin with correction for inhomogeneous distribution is shown in Fig 1. The results show an error of <5%, which is well within the standard range for comparing 2 laboratory tests.
DISCUSSION
The last decade has seen several attempts with either near-infrared spectroscopy15 or other spectroscopic methods16–20 to measure the total hemoglobin in the skin quantitatively and to use the results as a measure of skin circulation. Conventional near-infrared spectroscopy can measure only relative changes of the parameter in the specific individual and therefore cannot be used for quantitative measurements.20,21 To our knowledge, it has not been possible thus far to achieve this goal by the above-cited methods.
A recent study by Van Woerkom et al22 using a diffuse optical spectroscopy instrument (Oxyplex TS; ISS, Champaign, IL) demonstrated a reliable correlation between tissue hemoglobin and venous hemoglobin before and after a red blood cell transfusion in preterm infants. The fiber-optic probe used with this method has a distance between the emitter and receiver of 1 cm, which is much shorter than in conventional near-infrared spectroscopy.
In our method, using the Mediscan 2000, the distance between emitter and receiver is even shorter, which is part of the physics measurement principle that enables quantitative measurements of concentration. In addition it is noninvasive, easy to handle, and not harmful to the vulnerable skin of preterm and term infants.
To our knowledge this is the first report of a noninvasive estimation of hemoglobin in the skin of preterm infants with such a good agreement with the venous hemoglobin values. This agreement is achieved by using iterative processes that take the hemoglobin-free tissue into account. This could be the first step toward reducing iatrogenic blood loss by venipuncture to obtain a blood count in the future. The next step will incorporate the mathematical model for inhomogeneous distribution of hemoglobin into the online measurement and confirm the preliminary results in a larger sample size and with a shorter time between the venous and spectroscopic measurements. This will include infants with conditions affecting the microcirculation (eg, septicemia). Additional studies to establish whether the use of this method will potentially reduce iatrogenic blood loss are required.
ACKNOWLEDGMENTS
This project was supported by a research grant from the Children’s University Hospital of Münster.
FOOTNOTES
Accepted Jan 6, 2005.
No conflict of interest declared.
REFERENCES
Maier RF, Obladen M, Messinger D, Wardrop CA. Factors related to transfusion in very low birth weight infants treated with erythropoietin. Arch Dis Child Fetal Neonatal Ed. 1996;74 :F182 –F186
Strauss RG. Neonatal anemia: pathophysiology and treatment. Immunol Invest. 1995;24 :341 –351
Jungmann H, Niedorf F, Kietzmann M. Noninvasive reflection spectra provide quantitative information about spatial distribution of skin chromophores. Med Phys. 2005;32 :1297 –1307
Postaire E, Jungmann H, Bejot M, Heinrich U, Tronnier H. Evidence for antioxidant nutrients-induced pigmentation in skin: results of a clinical trial. Biochem Mol Biol Int. 1997;42 :1023 –1033
Stahl W, Heinrich U, Jungmann H, Tronnier H, Sies H. Carotenoids in human skin: Carotenoids in human skin: noninvasive measurement and identification of dermal carotenoids and carotenol esters. Methods Enzymol. 2000;319 :494 –502
Stahl W, Heinrich U, Jungmann H, Sies H, Tronnier H. Carotenoids and carotenoids plus vitamin E protect against ultraviolet light-induced erythema in humans. Am J Clin Nutr. 2000;71 :795 –798
Ryan TJ. The blood vessels of the skin. In: Jarret A, ed. The Physiology and Pathophysiology of the Skin. New York, NY: Academic Press; 1973
Christ F, Genzel-Borovizceny O, Schaudig S, et al. Monitoring of the microcirculation in cardiac surgery and neonates using orthogonal polarization spectral imaging. In: Messmer K, ed. Orthogonal Polarization Spectral Imaging. Vol 24 (Progress in Applied Microcirculation). Basel, Switzerland: Karger; 2000:82–93
Wodick R, Lübbers W. Methods for the determination of light pathway in the photometry of turbid solutions or tissue using penetrating or incident light [in German]. Pflugers Arch. 1973;342 :29 –40
Stücker M, Steinbrügge J, Ihrig C, et al. Rhythmical variations of haemoglobin oxygenation in cutane capillaries. Acta Derm Venerol. 1998;78 :408 –411
Steinbrügge J. Rumliche und Zeitliche Heterogenitt der Kutanen Sauerstoffsttigung und Hmoglobinkonzentration bei Probanden und Patienten mit CVI [doctoral thesis]. Bochum, Germany: Ruhr University: 1998
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476) :307 –310
Nagashima Y, Yada Y, Hattori M, Sakai A. Development of a new instrument to measure oxygen saturation and total hemoglobin volume in local skin by near-infrared spectroscopy and its clinical application. Int J Biometeorol. 2000;44 :11 –19
Berardesca E, Andersen PH, Bjerring P, Maibach HI. Erythema induced by organic solvents: in vivo evaluation of oxygenized and deoxygenized haemoglobin by reflectance spectroscopy. Contact Dermatitis. 1992;27 :8 –11
Harrison DK, Evans SD, Abbot NC, Beck JS, McCollum PT. Spectrophotometric measurements of haemoglobin saturation and concentration in skin during the tuberculin reaction in normal human subjects. Clin Phys Physiol Meas. 1992;13 :349 –363
Newton DJ, Harrison DK, Delaney CJ, Beck JS, McCollum PT. Comparison of macro- and micro-lightguide spectrophotometric measurements of microvascular hemoglobin oxygenation in the tuberculin reaction in normal human skin. Physiol Meas. 1994;15 :115 –128
Wolff KD, Marks C, Uekermann B, Specht M, Frank KH. Monitoring of flaps by measurement of intracapillary haemoglobin oxygenation with EMPHO II: experimental and clinical study. Br J Oral Maxillofac Surg. 1996;34 :524 –529
Dullenkopf A, Lohmeyer U, Salgo B, Gerber AC, Weiss M. Non-invasive monitoring of haemoglobin concentration in paediatric surgical patients using near-infrared spectroscopy. Anaesthesia. 2004;59 :453 –458
Edwards AD. The clinical role of near infrared spectroscopy. J Perinat Med. 1994;22 :535 –539
Van Woerkom R, El-Rify ES, Cerussi A, et al. Non-invasive diffuse optical spectroscopy measures tissue response to red blood cell transfusion in low birth weight infants . Pediatr Res. 2003;53 :361A(Heike Rabe, MD, Natascha )
Department of Neonatology, Children’s University Hospital, Münster, Federal Republic of Germany
Department of Neonatology, Institute of Cancer Research, University of Witten/Herdecke, Herdecke, Federal Republic of Germany
ABSTRACT
Objective. To compare transcutaneously spectroscopically measured hemoglobin values with venous hemoglobin values in infants.
Study Design. Prospective study in healthy preterm and term infants who were breathing spontaneously.
Results. Recordings were obtained from 85 stable infants (median gestational age at measurement: 36 weeks [range: 34–43 weeks]; median body weight: 1890 g [range: 1095–4360 g]). The spectroscopic hemoglobin values were corrected for inhomogeneous distribution of hemoglobin in the tissue. The venous and spectroscopic hemoglobin values were then compared by using the Bland-Altman method, which gave an error of <5%.
Conclusions. This pilot study could illustrate a good relation between the 2 methods for measuring hemoglobin. Larger studies are required to validate the spectroscopic method in those with conditions that affect the skin microcirculation (eg, septicemia).
Key Words: skin microcirculation phlebotomy iatrogenic blood loss noninvasive method neonate
Anemia of prematurity is a common problem in the care of preterm infants that requires frequent donor packed red cell transfusions for treatment.1 Iatrogenic blood losses caused by venipuncture are a major cause of anemia in premature infants.2 Hematopoiesis cannot produce enough new erythrocytes to balance the loss. Therefore, new noninvasive techniques for estimating serum bilirubin, for example, have been introduced into neonatology to reduce iatrogenic blood loss.3,4 We report on a new noninvasive spectroscopic technique that measures hemoglobin quantitatively and transcutaneously in preterm and term infants. In this preliminary study, the spectroscopically measured hemoglobin values are compared with the results of venous sampling in the same infants.
METHODS
The Device
The visible and near-infrared spectroscopic device (Mediscan 2000; MBR Messtechnik, Herdecke, Federal Republic of Germany; patent number DE 19831424, German Patent Office, Munich, Federal Republic of Germany) uses halogen white light that is emitted through 6 silica fibers (200-μm diameter) onto the skin. The reflected light, which has traveled through the skin tissue, is then collected via a silica fiber of 200-μm diameter and analyzed for spectra in the range of 350 to 1020 nm. The spectra are calibrated before each recording. Thus, in contrast to conventional near-infrared spectroscopy, the whole frequency range of light is analyzed. The distance between the outer surface of the 6 emitting fibers and the central receiving fiber is 250 μm (SD: 4 μm). The separation of the fibers is achieved by the individual cladding of the fibers. The technical principles of the method have been published elsewhere.5 Quantitative measurements of carotenoids in the skin of adults have been performed by using this technique.6–8
Mathematical Background of the Method
In brief, the method works as follows. The anatomy of the capillaries in infants shows an inhomogeneously distributed pattern.9 The hemoglobin is inhomogeneously distributed in the capillaries and the tissue because of changes in the distribution of the capillary perfusion.10
The first step of the method is to calculate the product of concentration and path length of the light between the light emitter and the light-receiving fiber. In a second step, the concentration of hemoglobin is calculated. Both steps are performed with the help of an iterative process.
The first step is performed in the following way: If one calculates the product of concentration and path length from direct measurements, the resulting value will be low. The reason for this is the inhomogeneous distribution of hemoglobin in tissue. The inhomogeneous distribution of hemoglobin is a geometric property, which differs from 1 measurement site to another. Therefore, the degree of inhomogeneity is determined individually at each measurement site. If the degree is known, the path-length concentration factor of homogeneously distributed hemoglobin is calculated.
The degree of inhomogeneous distribution is calculated in the following way: In the first step, the mapping from homogeneously distributed hemoglobin to the measured inhomogeneously distributed hemoglobin is calculated. Then the nonlinear fitting algorithm of Levenberg-Marquardt, a polynomial of third order, is fitted to this mapping. Then the homogeneous distribution of hemoglobin is changed step by step to an inhomogeneous one. At each step the mapping is calculated. If the mapping becomes linear, it is then a polynomial of first order; the 2 distributions are equal, and the degree of inhomogeneity is known. Now the true product of concentration and path length is known.
In the second step, the true concentration of hemoglobin is calculated. Because the degree of inhomogeneity of the hemoglobin is now known, then a 2-component system can be used. One component is the hemoglobin, and the other component is the hemoglobin-free tissue.
It was shown by Wodick and Lübbers11 that the spectra are disturbed only by the distribution of hemoglobin and not by the special arrangement of the distribution. All different arrangements of hemoglobin with the same degree of inhomogeneity affect the spectra in the same way. From the hemoglobin compartment all optical properties are known: absorption, scattering, and the anisotropy coefficient (optical properties for the whole blood). The path length of light in both compartments is the same. The degree of inhomogeneity of tissue distribution is known (which is 1 minus the degree of inhomogeneity of hemoglobin). With the fit of Levenberg-Marquardt, it is possible to determine the optical properties of tissue from the measured spectrum for a given path length. This measurement is performed in the following way: in the capillary bed we found oscillating of hemoglobin. It is a well-known phenomenon described by many methods that hemoglobin in the capillary bed is detected with an oscillating pattern.
With the method described above, 2 different hemoglobin concentrations and path lengths are measured at the same measurement site (higher and lower hemoglobin content in tissue). Then the difference of hemoglobin is calculated. The difference affects the path length in tissue and, therefore, the path length in hemoglobin-free tissue. From the hemoglobin differences the path-length changes in tissue are calculated, and these processes are performed repeatedly on the path lengths until the calculated differences and the error of the measured differences in the hemoglobin-free tissue component becomes minimal (iterative process). Finally, the best path-length difference is found. From the ratio of tissue path length/tissue path length of the difference, the tissue path length of the light is calculated. Then the concentration is calculated from the well-known product of path length and concentration.
The variability of repeated transcutaneous hemoglobin measurements has been reported as <5%.12,13
Measurements
The present pilot study was aimed at quantitative measurement of the skin hemoglobin in preterm and term infants and a comparison with venous values. The venous hemoglobin was taken weekly according to unit protocol. The spectroscopic measurements were performed as part of a larger study on skin microcirculation in preterm and term infants. No blood was taken for study purposes alone. Therefore, only 85 comparative sets of data were obtained. The spectroscopic measurements were performed by a handheld probe on the forearm of the infants. All infants were in a stable condition (with normal blood pressure for age and a capillary refilling time of < 3 seconds at the finger tip) and were measured at the same time in the late afternoon or evening after having been feed. This setting depended on the availability of the research staff and did not disturb the routine work of the ward. Exclusion criteria were clinical or laboratory signs of infection, congenital malformations, and requirement of oxygen or artificial ventilation. The spectroscopic baseline values of the total hemoglobin in the skin were obtained for 1 minute, and the mean values were calculated. The venous samples for measuring the hemoglobin were processed by a standard Coulter counter (K-1000; Sysmex GmbH, Düsseldorf, Federal Republic of Germany). The hemoglobin values then were compared by using the Bland-Altman method.14
The study was approved by the hospital ethics committee. Informed consent was obtained from the parents before the study.
RESULTS
Eighty-five infants were enrolled into this study. The median birth weight of the infants was 1040 g (range: 600–4160 g), and the median gestational age was 33 weeks (range: 24–41 weeks). The study was performed at a mean corrected age of 36 weeks (range: 34–43 weeks). The median body weight of the infants was 1890 g (range: 1095–4360 g). The median time between the blood sampling and spectroscopic measurements was 11 hours (range: 9–14 hours). All infants were studied only once.
The comparison according to Bland and Altman14 of venous hemoglobin and the spectroscopically obtained hemoglobin with correction for inhomogeneous distribution is shown in Fig 1. The results show an error of <5%, which is well within the standard range for comparing 2 laboratory tests.
DISCUSSION
The last decade has seen several attempts with either near-infrared spectroscopy15 or other spectroscopic methods16–20 to measure the total hemoglobin in the skin quantitatively and to use the results as a measure of skin circulation. Conventional near-infrared spectroscopy can measure only relative changes of the parameter in the specific individual and therefore cannot be used for quantitative measurements.20,21 To our knowledge, it has not been possible thus far to achieve this goal by the above-cited methods.
A recent study by Van Woerkom et al22 using a diffuse optical spectroscopy instrument (Oxyplex TS; ISS, Champaign, IL) demonstrated a reliable correlation between tissue hemoglobin and venous hemoglobin before and after a red blood cell transfusion in preterm infants. The fiber-optic probe used with this method has a distance between the emitter and receiver of 1 cm, which is much shorter than in conventional near-infrared spectroscopy.
In our method, using the Mediscan 2000, the distance between emitter and receiver is even shorter, which is part of the physics measurement principle that enables quantitative measurements of concentration. In addition it is noninvasive, easy to handle, and not harmful to the vulnerable skin of preterm and term infants.
To our knowledge this is the first report of a noninvasive estimation of hemoglobin in the skin of preterm infants with such a good agreement with the venous hemoglobin values. This agreement is achieved by using iterative processes that take the hemoglobin-free tissue into account. This could be the first step toward reducing iatrogenic blood loss by venipuncture to obtain a blood count in the future. The next step will incorporate the mathematical model for inhomogeneous distribution of hemoglobin into the online measurement and confirm the preliminary results in a larger sample size and with a shorter time between the venous and spectroscopic measurements. This will include infants with conditions affecting the microcirculation (eg, septicemia). Additional studies to establish whether the use of this method will potentially reduce iatrogenic blood loss are required.
ACKNOWLEDGMENTS
This project was supported by a research grant from the Children’s University Hospital of Münster.
FOOTNOTES
Accepted Jan 6, 2005.
No conflict of interest declared.
REFERENCES
Maier RF, Obladen M, Messinger D, Wardrop CA. Factors related to transfusion in very low birth weight infants treated with erythropoietin. Arch Dis Child Fetal Neonatal Ed. 1996;74 :F182 –F186
Strauss RG. Neonatal anemia: pathophysiology and treatment. Immunol Invest. 1995;24 :341 –351
Jungmann H, Niedorf F, Kietzmann M. Noninvasive reflection spectra provide quantitative information about spatial distribution of skin chromophores. Med Phys. 2005;32 :1297 –1307
Postaire E, Jungmann H, Bejot M, Heinrich U, Tronnier H. Evidence for antioxidant nutrients-induced pigmentation in skin: results of a clinical trial. Biochem Mol Biol Int. 1997;42 :1023 –1033
Stahl W, Heinrich U, Jungmann H, Tronnier H, Sies H. Carotenoids in human skin: Carotenoids in human skin: noninvasive measurement and identification of dermal carotenoids and carotenol esters. Methods Enzymol. 2000;319 :494 –502
Stahl W, Heinrich U, Jungmann H, Sies H, Tronnier H. Carotenoids and carotenoids plus vitamin E protect against ultraviolet light-induced erythema in humans. Am J Clin Nutr. 2000;71 :795 –798
Ryan TJ. The blood vessels of the skin. In: Jarret A, ed. The Physiology and Pathophysiology of the Skin. New York, NY: Academic Press; 1973
Christ F, Genzel-Borovizceny O, Schaudig S, et al. Monitoring of the microcirculation in cardiac surgery and neonates using orthogonal polarization spectral imaging. In: Messmer K, ed. Orthogonal Polarization Spectral Imaging. Vol 24 (Progress in Applied Microcirculation). Basel, Switzerland: Karger; 2000:82–93
Wodick R, Lübbers W. Methods for the determination of light pathway in the photometry of turbid solutions or tissue using penetrating or incident light [in German]. Pflugers Arch. 1973;342 :29 –40
Stücker M, Steinbrügge J, Ihrig C, et al. Rhythmical variations of haemoglobin oxygenation in cutane capillaries. Acta Derm Venerol. 1998;78 :408 –411
Steinbrügge J. Rumliche und Zeitliche Heterogenitt der Kutanen Sauerstoffsttigung und Hmoglobinkonzentration bei Probanden und Patienten mit CVI [doctoral thesis]. Bochum, Germany: Ruhr University: 1998
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476) :307 –310
Nagashima Y, Yada Y, Hattori M, Sakai A. Development of a new instrument to measure oxygen saturation and total hemoglobin volume in local skin by near-infrared spectroscopy and its clinical application. Int J Biometeorol. 2000;44 :11 –19
Berardesca E, Andersen PH, Bjerring P, Maibach HI. Erythema induced by organic solvents: in vivo evaluation of oxygenized and deoxygenized haemoglobin by reflectance spectroscopy. Contact Dermatitis. 1992;27 :8 –11
Harrison DK, Evans SD, Abbot NC, Beck JS, McCollum PT. Spectrophotometric measurements of haemoglobin saturation and concentration in skin during the tuberculin reaction in normal human subjects. Clin Phys Physiol Meas. 1992;13 :349 –363
Newton DJ, Harrison DK, Delaney CJ, Beck JS, McCollum PT. Comparison of macro- and micro-lightguide spectrophotometric measurements of microvascular hemoglobin oxygenation in the tuberculin reaction in normal human skin. Physiol Meas. 1994;15 :115 –128
Wolff KD, Marks C, Uekermann B, Specht M, Frank KH. Monitoring of flaps by measurement of intracapillary haemoglobin oxygenation with EMPHO II: experimental and clinical study. Br J Oral Maxillofac Surg. 1996;34 :524 –529
Dullenkopf A, Lohmeyer U, Salgo B, Gerber AC, Weiss M. Non-invasive monitoring of haemoglobin concentration in paediatric surgical patients using near-infrared spectroscopy. Anaesthesia. 2004;59 :453 –458
Edwards AD. The clinical role of near infrared spectroscopy. J Perinat Med. 1994;22 :535 –539
Van Woerkom R, El-Rify ES, Cerussi A, et al. Non-invasive diffuse optical spectroscopy measures tissue response to red blood cell transfusion in low birth weight infants . Pediatr Res. 2003;53 :361A(Heike Rabe, MD, Natascha )