This study introduces an experimental approach to estimate percentage of hemoglobin content and volume sampled by near infrared spectroscopy (NIRS). Carbogen (5% CO2, 95% O2) respiratory intervention was used to induce physiological changes in a group of six Fisher rat breast tumors. Changes in total hemoglobin concentration, Δ[Hb]total, and in total blood volume, ΔVT-blood, of the tumors were measured by NIRS and 19F magnetic resonance spectroscopy of perflubron, respectively. The ratio of Δ[Hb]total/ΔVT-blood was used to calculate the fraction of hemoglobin contents sampled by NIRS. The results showed that the mean value of estimated fractions is within a range of 15~30% of total hemoglobin content in the tumor tissues. Based on the results, we suggest that NIRS does not sample the entire hemoglobin volume of the tissue vasculature, but is more sensitive to microvasculature. This study helps to understand the blood vascular volume sampled by NIRS, and demonstrates that the low cost, portable NIRS system may be a reliable, non-invasive, real-time, monitoring tool for changes in tumor blood contents.
©2005 Optical Society of America
In biomedical research, optical spectroscopy and imaging can potentially provide rapid, economical, and non-invasive diagnostic links between crucial tissue characteristics and cancer detection or diagnosis [1–7]. Near infrared (NIR) light in the range of 700 nm to 900 nm has an optimal penetrating depth in tissue (up to 15 centimeters) since photon transport in tissue is dominated by light scattering rather than absorption in the NIR region. Therefore, the use of NIR light enables us to sample large tissue volumes, such as the breast, the brain, and skeletal muscles as well as tumors deep in the tissue. In the NIR spectral range, two important endogenous tissue chromospheres, i.e., oxygenated and deoxygenated hemoglobin (HbO2 and Hb, respectively), display oxygen-dependent absorption features [8,9]. Experimental determination of absorption quantities of tissue can provide quantification of several physiological parameters, such as hemoglobin/myoglobin concentrations, vascular oxygenation, and blood volume changes. Such quantification could be important for cancer prognosis and treatment monitoring, as we have demonstrated in our previous reports [10–12].
Because of large scattering of NIR light in tissue, however, it is not clear how much of the total vascular volume within a tissue is sampled by the NIR signals. It has been reported that the NIR detection is most sensitive to microvessel density, i.e., arterioles, capillaries, and venules [13–14]. The opaque microstructures in tissue also have contributions to the absorption . The measured absorption is actually a volume-weighted average of absorption of the sampled vessels and the surrounding tissue. Thus, the measured (apparent) chromophore concentration may be less than the actual concentration. Moreover, the blood flow measured by NIR spectroscopy (NIRS) was reported 2~3 times lower than the plethysmographic flow [16–17]. These reports motivated us to investigate what fraction of blood volume in tissue is sampled by NIR. In particular, tumor tissue is recognized as highly heterogeneous. Better understanding and accurate determination of tumor vascular blood volume sensed by the NIRS could provide crucial information for assessing tumor progress and guiding tumor therapy and treatment.
Several investigators have shown that vascular volume can be estimated using 19F Magnetic Resonance spectroscopy (MRS) following infusion of a perfluorocarbon blood substitute emulsion [18–19]. Our previous studies have demonstrated that NIRS can be used to monitor changes in total (i.e., oxygenated and deoxygenated) hemoglobin concentrations within tumor vasculature. Thus, correlations between MRS and NIRS measurements may allow us to investigate the relationship between the blood volumes sampled by these two respective methods. Specifically, in this paper, we will report our measurements of changes in total hemoglobin concentration, Δ[Hb]total, and changes in tumor blood volume, ΔVT-blood, induced by hyperoxic gas intervention in rat breast tumors using NIRS and 19F MRS of perflubron (formerly called perfluorooctyl bromide or PFOB), respectively. The ratio of rat tumor Δ[Hb]total/ΔVT-blood will be estimated to quantify the fractions of tumor blood volume sampled by NIRS and MRS in tumors.
2. Experimental materials and methods
2.1 Near Infrared spectroscopy for the measurement of Δ[Hb]total
A homodyne, frequency-domain NIRS system (NIM, Philadelphia, PA) was used in our experiments, as described in detail previously [10–12]. Briefly, the amplitude-modulated light at 140 MHz from two NIR laser diodes at 758 nm and 785 nm was projected on one side of the tumor through a delivery fiber bundle. The diffused light through the tumor was collected and propagated to a photomultiplier tube (PMT) by a second fiber bundle. The signal from the PMT was demodulated through an In-phase and Quadrature-phase circuit, and the amplitude and phase were recorded, as shown in Fig. 1. Based on modified Beer-Lambert’s law [10, 20–21] and the recorded amplitude, changes in oxygenated, deoxygenated, and total hemoglobin concentrations, Δ[HbO2], Δ[Hb] and Δ[Hb]total, due to respiratory intervention were estimated, as described below. In this NIRS system, only amplitude signal was used for the calculation since phase signal was very noisy.
2.2 19F MRS of perflubron for the measurement of tumor blood volume VT-blood
An Omega CSI 4.7 T superconducting magnet system (AcustarTM, Bruker Instrument, Inc., Fremont, CA) was used for the measurement of tumor blood volume. An emulsion of perflubron (OxygentTM, Alliance Pharmaceutical Corp., San Diego, CA) (2 ml) was infused into tumor-bearing rats i.v. as a blood volume indicator. The tumors were placed within a frequency-tunable (1H/19F), single-turn, solenoid coil, together with a sealed capillary containing sodium trifluoroacetate (TFA), which was used as an external standard reference for calibrating tumor blood volume.
Under fully relaxed conditions, integration of 19F signal from the tumor was linearly proportional to the total number of 19F nuclear spins of perflubron in the tumor, which, in turn, was linearly proportional to the total blood volume in the tumor, assuming that the perflubron emulsion had reached an equilibrium state with the blood throughout the tumor.. To ensure full relaxation a repetition time (TR) of 30 seconds was used . The spectral peaks of perflubron were integrated in the data post-processing. Following the tumor measurement, the rat was removed, leaving the reference TFA capillary in the original position within the RF coil. A small amount of blood (0.5 ml) was drawn from the rat tail vein and placed in the RF coil without disturbing the reference TFA capillary. Another quantitative 19F spectrum was then acquired. Thus, the tumor blood volume can be calculated based on the following Eq.:
where VT_blood and VS_blood were the tumor blood volume and blood sample volume (ml), respectively, IT_blood and IS_blood were the integrated MRS signal of 19F taken from the rattumor and blood sample, respectively, and Iblood_TFA and ITumor_TFA were the integrated 19F MRS signals from the TFA capillary for the two respective measurements.
2.3 Animal model and protocols
Mammary adenocarcinomas 13762NF were implanted in skin pedicles  on the forebacks of female Fisher 344 rats (~150 g). Once the tumors reached 1~2 cm diameter, rats were anesthetized with ketamine hydrochloride (100 mg/ml, i.p.) and maintained under general gaseous anesthesia with 1.3% isoflurane in air (1 dm3/min). Tumors were shaved to improve optical contact for NIR light transmission. Tumor volume, V, was estimated using an ellipsoid approximation as V=(π/6).a.b.c from the three orthogonal diameters, i.e., a, b, c, measured with a caliper.
In this study, breast tumor bearing Fisher rats were challenged with carbogen (95% O2+5% CO2) using a sequence of air-carbogen-air. The Δ[Hb]total values in response to the carbogen inhalation were monitored by NIRS. For the rats that exhibited a significant change in [Hb]total induced by carbogen (n=6), they were re-anesthetized the next day after the NIRS measurement and infused with 2 ml of perflubron emulsion for the tumor blood volume determination. The 19F MRS. experiment was performed on the rat 30 min after the emulsion injection allowing sufficient time for the perflubron emulsion to reach an equilibrium state within the tumor blood vasculature. The same inhaled gas sequence was repeated during the MRS measurements.
3. Algorithms for calculations of Δ[Hb]total and fraction of sampled blood volume
3.1 Calculations for tumor Δ[Hb]total
Based on Beer-Lambert’s law, optical attenuation, presented by Optical Density (OD), can be expressed as a function of concentration of chromospheres (24–28),
where A0 and A are light intensities of the incident and transmitted light, respectively, ε is the extinction coefficient of chromophore, c is the concentration of chromophore, and l is the optical path-length through the measured sample. When the measured sample has a mixture of chromophores, e.g., oxygenated and deoxygenated hemoglobin in tumor tissue, the changes in oxygenated and deoxygenated hemoglobin concentrations (i.e., Δ[HbO 2] and Δ[Hb]) can cause a change in optical density, which can be presented as [26–28]:
where ΔOD(λ) represents a change in optical density at the specific wavelength, λ. AB and AT correspond to light intensities measured under the baseline and transient conditions. εHb(λ) and εHbO2(λ) are extinction coefficients at wavelength λ for molar concentrations of deoxygenated hemoglobin and oxygenated hemoglobin, respectively. By employing two wavelengths in Eq. (3), both Δ[HbO2] and Δ[Hb] can be determined by measuring the ΔOD values at the two specific wavelengths, provided that the values for εHb(λ) and εHbO2(λ) are known:
Note that in principle, l represents the optical path length between the source and detector. While l is simply the physical separation, d, between the source and detector through a non-scattering medium, exact quantification of l for an intact tissue or organ is complex because of light scattering in tissue. Since l is in proportion to the separation, d, we can associate l to d as l=DPF*d, where DPF is a differential path length factor to account for light scattering . It has been well accepted [9, 10, 20] that together with DPF, Eq. (3) can be treated as modified Beer-Lambert’s law; and consequently, Eqs. (4) and (5) can be used to quantify changes in [Hb] and [HbO2] in highly scattering media, such as in intact tissue or organs.
To be consistent with our previous work, we adopt in this paper the ε values published by Zijlstra et al. . We had to interpolate the ε values at the two wavelengths employed in our study, i.e., ε Hb(758 nm)=1.418, ε HbO2(758 nm)=0.6372, ε Hb(785 nm)=1.111, and ε HbO2(785 nm)=0.766, all in mM-1cm-1. Note that a factor of 4 has been multiplied for each of the ε’s at the respective wavelengths to account for light absorption from 4 hemes per hemoglobin molecule  since the extinction coefficients published in the field of biochemistry were expressed on a per heme basis [26–30].
Because of the interpolation of ε values, necessary calibration is needed to assure the accuracy of Δ[HbO2], Δ[Hb], and Δ[Hb]total calculation as given in Eqs. (4)–(5). We conducted a set of phantom experiments, as previously described in ref. 11. The empirical calibration has resulted in two correction factors β1 and β2, where β1=1.103 and β2=0.9035, as expressed below:
where l has been replaced by d×DPF. Notice that the difference in the coefficients between these equations and those published earlier [10–12] results from the factor of 4 in ε values. Δ[Hb]total can also be obtained by adding Eqs. (8) and (9):
In principle, units of Δ[HbO2], Δ[Hb], and Δ[Hb]total are in mM with ε in cm-1.mM-1. Since DPF is so far an unknown parameter for tumors, we include it within the unit as mM/DPF for the relative measurements. Thus, Eq. (10) becomes Eq. (11), and it permits us to obtain Δ[Hb]total with a unit of mM/DPF for breast tumors under carbogen intervention.
3.2 Calculations for the fraction of vascular blood volume sampled by NIRS
The determination of Δ[Hb]total can further lead to estimation of changes in total hemoglobin content within a rat tumor, Δ[CHb]tumor, since total hemoglobin content of a tumor, [CHb]tumor, is directly associated with hemoglobin concentration and tumor physical volume. Namely, [CHb]tumor determined by NIRS should be a product of tumor vascular hemoglobin concentration, [Hb]total in mole/liter=103 mM, hemoglobin molecular weight, M in gram/mole, and the tumor physical volume, VT-physical in cm3, and can be expressed as
where [CHb]tumor_NIRS and Δ[CHb]tumor_NIRS are in gram, and Δ[CHb]tumor_NIRS is induced by carbogen intervention. On the other hand, the total tumor blood volume, VT-blood in liters, can be quantified from the 19F MRS measurement with and without carbogen inhalation. The corresponding hemoglobin content determined by the MRS can be written as:
where K is hemoglobin concentration in gram/liter for liquid blood, VT-blood can be determined using Eq. (1) based on the MRS measurements, and ΔVT-blood results from respiratory intervention.
To quantify the fraction of tumor blood volume sampled by NIRS and MRS in tumors, we introduce a variable, γ, to associate [CHb]tumor_NIRS with [CHb]tumor_MRS as:
Since γ is expected to be a constant for a given tumor, it follows that
where K is often given as 150 g/l , M is taken as 68000 g/mol , Δ[Hb]total and ΔVT-blood are obtained from our NIRS and MRS experimental measurements, respectively, and VT-physical is measured from the tumor physical volume. With Eq. (15), we are able to compare and estimate the fraction of hemoglobin content or blood volume of the tumor, γ, sampled by near infrared spectroscopy and 19F magnetic resonance spectroscopy.
4. Experimental results
Changes of total hemoglobin concentration, Δ[Hb]total, and blood volume, VT-blood, of six selected rat tumors were monitored by NIRS and 19F MRS of perflubron on consecutive days, respectively. Figure 2 shows the time course profiles of Δ[Hb]total and VT-blood from a representative breast tumor (V=2.6 cm3) with respect to carbogen intervention. When the inhaled gas was switched from air to carbogen, Δ[Hb]total increased significantly (p<0.0001) from a baseline value of 0.0005±0.0015 mM/DPF to a maximal value of 0.020±0.001 mM/DPF over the period of carbogen intervention, with a maximal Δ[Hb]total of 0.0185 mM/DPF. After the gas was switched back to air, a significant drop (p<0.0001) of Δ[Hb]total occurred, followed by a plateau at 0.009±0.001 mM/DPF. A similar temporal pattern in response to carbogen intervention was found for VT-blood derived from the MRS measurement. Specifically, VT-blood increased significantly (p<0.0001) from a baseline of 0.809±0.004 cm3 to a maximum of 0.837±0.004 cm3 during carbogen intervention, having a maximal ΔVT-blood of 0.028 cm3.
By substituting Δ[Hb]total, ΔVT-blood, K, M, and the corresponding VT-Physical=2.6 cm3 into Eq. (15), we obtain
Typical values of DPF for the brain and muscle have been reported in the range of 4–6 and 3–4, respectively [32–35], independent of the source-detector separation when d is larger than 2.5cm [36–37]. However, the DPF values for tumors, DPFtumor, are not well studied. It is still reasonable to assume DPFtumor to be in the range of 2–4 because of 1) higher blood content, 2) denser blood vasculature, and 3) the finite size of solid tumors in this study. Thus, the percentage of tumor vascular content or volume sampled by NIRS over that by 19F MRS is within the range of 19%~39%, based on the result of Eq. (16), for a given DPFtumor of 4 and 2, respectively.
Data for all six rat breast tumors are shown in Table 1. The percentage of tumor vascular hemoglobin content sampled by NIRS and by MRS is within the range of 17~48% if considering a DPFtumor of 2, and 8~24% if considering a DPFtumor of 4. The mean value of the percentage, γ, is about 15±6% and 30±12%, depending on the DPF values of actual tumors. No obvious correlation was found between percentage values and tumor volume.
In this study, changes in total hemoglobin concentration, Δ[Hb]total, and in total blood volume, ΔVT-blood, of rat breast tumors in response to carbogen intervention were determined by NIRS and by 19F MRS of perflubron, respectively. The ratio of Δ[Hb]total/ΔVT-blood was used to calculate the fraction of vascular hemoglobin contents sampled by NIRS and MRS. Since 19F MRS of perflubron samples all blood volume in the tumor [18–19], the MRS measurement may be considered as a “gold standard” for the complete total blood volume within the tumor, provided that the entire rat tumor was placed within the RF coil. Then, this experimental approach provides evidence that the vascular blood content or volume sampled by the NIR system is about 15~30% of the total blood volume (Table 1). This estimation should be applicable not only to tumor tissues but also to other kinds of tissues, providing us with important information for better understanding of blood vascular volume sampled by NIRS. Moreover, the consistent trends between Δ[Hb]total and ΔVT-blood during hyperoxic gas intervention (Fig. 2) demonstrate that the low cost, portable NIRS system provides a noninvasive, real-time, monitoring tool to detect changes in tumor blood content.
Is it reasonable to believe that NIRS samples only 15~30% of the total blood content or volume in tissue vasculature? It has been reported that NIR light is most sensitive to microvessel densities, such as arterioles, capillaries, and venules [13–14] and surface areas of microstructures in tissue . The size of micro blood vessels is often in the range of ~100 µm, which is in the same order of magnitude as the average scattering length of NIR light in tissue [15, 31]. So NIR light is scattered more frequently by the small blood vessels than large vessels, whereas large vessels have larger chances to terminate the NIR light due to high absorption. Thus, multiple scattering from micro vasculature permits better sensitivity to be sampled by NIR light. Moreover, small vessels have lower hematocrit compared to big vessels because of reduction of hematocrit with the decrease of vessel diameters . Thus, based on our experimental measurements, it is reasonable to conclude that the NIR system samples 15–30% portion of total hemoglobin contents in tissue vasculature, mainly sensing the signals in microvessels.
DPF is a very important parameter for quantifying NIR measurements. As mentioned earlier, many investigators have conducted research on the estimation of optical path length of the brain and muscle of both human subjects and animal models [20, 32–37]. However, the NIR light propagated in tumor tissue is different from that in brain and muscle. Due to tumor angiogenesis, tumor tissues possess more blood vessels and have higher light absorption than the normal brain and muscle. Furthermore, the solid tumors in the study have a finite size, where “photon escape” from the measured volume can take place . The term of “photon escape” has been used to refer to the situation where the sample size under study is finite, and a large amount of light may escape from the sample and never be detected, leading to a shorter path length in comparison with the path length existing in a large volume of tissues. Steen et al took DPF=2.5 for their tumor study . Therefore, we believe that DPF tumor=2~4 is a reasonable range used in our estimation of hemoglobin contents for breast tumors.
The errors in the percentage measurement may be potentially reduced by conducting measurements of NIRS and 19F MRS simultaneously, eliminating experimental variations between the two kinds of measurements. One might also apply 1H MRI of vascular volume markers such as super paramagnetic iron oxide particles (SPIOs) [41–42] in place of the 19F NMR approach used here. Since agents, such as Combidex, are in clinical use, correlation studies with NIRS could allow examination of percentage of blood content/volume sampled by the NIRS system in human tumors. Also, the uncertainty in both the size measurement and volume calculation of the tumors should be reduced since they all contribute critically to the error in the estimated fraction of blood vascular contents.
In our study, not all tumors showed significant changes in [Hb]total accompanying challenge with hyperoxic gas. This is mainly due to both heterogeneity and irregularity of breast tumors, which have leaky, broken, and dis-functioning blood vasculatures. The inconsistency of changes in tumor [Hb]total induced by hyperoxic gas interventions has been observed throughout our research studies over several years. Thus, we believe it is reasonable to select rat breast tumors, which show significant changes in [Hb]total, for this study in order to obtain estimated fraction of vascular blood contents sampled by NIRS. In conclusion, we have introduced a new experimental approach to estimate the percentage of tumor blood content or volume sampled by the NIRS system by combining the measurements of NIRS with 19F MRS. The results gave rise to an estimated range of 15~30%, suggesting that NIRS does not sample the entire blood hemoglobin content/volume of the tissue, but it senses more toward microvasculature. The study helps better understand blood vascular volume sampled by NIRS, and demonstrates that the low cost, portable NIRS system can be a reliable, non-invasive, real-time, monitoring tool on changes in tumor blood contents.
This work was supported in part by the Department of Defense Breast Cancer Research grants DAMD17-01-1-0423 (YG) and DAMD17-00-1-0459 (HL), and in part by NIH RO1 CA79515/EB2762 (RPM). Tumor cells were provided by the Division of Cancer Therapeutics, NIH, and perflubron emulsion (OxygentTM) was kindly provided by Alliance Pharmaceuticals. The authors would like to acknowledge Yulin Song’s initial work and useful discussion for this publication.
References and links
1. W. Cheong, S. A. Prahl, and A. J. Welch,.“A review of the optical properties of biological tissue,” IEEE J. Quantum Electron. 26, 2166–2183 (1990). [CrossRef]
3. A. M. Siege, J. J. A. Marota, and D. A. Boas. “Design and evaluation of a continuous wave diffuse optical tomography system,” Opt. Express. 4, 287–298 (1999). [CrossRef]
4. B. Chance, E. Anday, S. Nioka, S. Zhou, L. Hong, K. Worden, C. Li, T. Murray, Y. Ovetsky, D. Pidikiti, and R. Thomas, “A novel method for fast imaging of brain function, non-invasively, with light,” Opt. Express. 2, 411–423 (1998). [CrossRef] [PubMed]
5. V. Ntaiachristors, H. Ma, and B. Chance, “Multichannel photon counting instrument for spatially resolved near infrared spectroscopy,” Rev. Sci. Instrum. 70, 1444–1451 (1999).
6. B. W. Pogue, M. Testorf, and T. McBride, “Instrumentation and design of a frequency domain diffuse optical tomography imager for breast cancer detection,” Opt. Express. 1, 391–403 (1999). [CrossRef]
9. E. M. Sevick, B. Chance, J. Leigh, and M. Maris, “Quantitation of time-and frequency-resolved optical spectra for the determination of tissue oxygenation,” Anal. Biochem. 195, 330–351 (1991). [CrossRef] [PubMed]
10. H. Liu, Y. Song, K. L. Worden, X. Jiang, A. Constantinescu, and R. P. Mason, “Noninvasive investigation of blood oxygenation dynamics of tumors by near-infrared spectroscopy,” Appl. Opt. 39, 5231–5243 (2000). [CrossRef]
11. J. G. Kim, Y. Song, D. Zhao, A. Constantinescu, R. P. Mason, and H. Liu, “Interplay of Tumor Vascular Oxygenation and pO2 in Tumors Using NIRS, 19F MR pO2 Mapping, and pO2 Needle Electrode,” J. Biomed. Opt. 8, 53–62 (2003). [CrossRef] [PubMed]
12. Y. Gu, V. Bourke, J. Kim, A. Constantinescu, R. P. Mason, and H. Liu, “Dynamic response of breast tumor oxygenation to hyperoxic respiratory challenge monitored with three oxygen-sensitive parameters,” Appl. Opt. 42, 2960–67 (2003). [CrossRef] [PubMed]
13. H. Liu, A. H. Hielscher, F. K. Tittel, S. L. Jacques, and B. Chance, “Influence of blood vessels on the measurement of hemoglobin oxygenation as determined by time-resolved reflectance spectroscopy,” Med. Phys. 22, 1209–1216 (1995). [CrossRef] [PubMed]
14. M. Firbank, E. Okada, and D. T. Delpy, “Investigation of the effect of discrete absorbers upon the measurement of blood volume with near-infrared spectroscopy,” Phys. Med. Biol. 42, 465–477(1997). [CrossRef] [PubMed]
15. B. W. Pogue, E. A. White, U. L. Osterberg, and K. D. Paulsen, “Absorbance of opaque microstructures in optically diffuse media,” Appl. Opt. 40, 4616–4621 (2001). [CrossRef]
16. M. C. Van Beekvelt, W. N. Colier, R. A. Wevers, and B. G. Van Engelen, “Performance of near infrared spectroscopy in measuring local O2 consumption and blood flow in skeletal muscle,” J. Appl. Physiol. 90, 511–519 (2001). [PubMed]
17. R. A. De Blasi, M. Ferrari, A. Natali, G. Conti, A. Mega, and A. Gasparetto, “Noninvasive measurement of forearm blood flow and oxygen consumption by near infrared spectroscopy,” J. Appl Physiol. 76, 1388–1393 (1994). [PubMed]
18. C. Thomas, C. Counsell, P. Wood, and G.E. Adams “Use of F-19 NMR spectroscopy and hydralazine for measuring dynamic changes in blood perfusion volume in tumors in mice,” JNCI 84,174–80(1992). [PubMed]
19. N. J. Baldwin, Y. Wang, and T. C. Ng, “In situ 19F MRS measurement of RIF-1 tumor blood volume: Corroboration by radioisotope-labeled [125I]-albumin and correlation to tumor size,” Magn. Reson. Imaging , 14, 275–280 (1996). [CrossRef] [PubMed]
20. D. T. Delpy, M. Cope, P. van der Zee, S. Arridge, S. Wray, and J. Wyatt, “Estimation of optical pathlength through tissue from direct time of flight measurement,” Phys. Med. Biol. 33, 1433–1442 (1988). [CrossRef] [PubMed]
21. Y. Gu, Z. Qian, J. Chen, D. Blessington, N. Ramanujam, and B. Chance, “High resolution three dimensional scanning optical image system for intrinsic and extrinsic contrast agents in tissue,” Rev. Sci. Instrum. 73, 172–178 (2002). [CrossRef]
22. H. P. Shukla, R. P. Mason, D. E. Woessner, and P. P. Antich, “A comparison of three commercial perfluorocarbon emulsions as high field NMR probes of oxygen tension and temperature,” J. Magn. Reson. B , 106, 131–141 (1995). [CrossRef]
23. E. W. Hahn, P. Peschke, R. P. Mason, E. E. Babcock, and P.P. Antich, “Isolated tumor growth in a surgically formed skin pedicle in the rat: A new tumor model for NMR studies,” Magn.Reson. Imaging 11, 1007–1017 (1993). [CrossRef] [PubMed]
24. B. L. Horecker, “The absorption spectra of hemoglobin and its derivatives in the visible and near infrared regions,” J. of Biol. Chem. 148, 173–183 (1943).
28. O. W. Van Assendelft and W. G. Zijlstra, “Extinction coefficients for use in equations for the spectrophotometric analysis of haemoglobin mixtures,” Anal. Biochem. 69, 43–48 (1975). [CrossRef] [PubMed]
29. W. G. Zijlstra, A. Buursma, H. E. Falke, and J. F. Catsburg, “Spectrophotometry of hemoglobin: absorption spectra of rat oxyhemoglobin, deoxyhemoglobin, carboxyhemoglobin, and methemoglobin,” Comp. Biochem. Physiol. 107B, 161–166 (1994).
30. M. Cope, “The application of near infrared spectroscopy to non invasive monitoring of cerebral oxygenation in the newborn infant,” Ph.D. thesis, Appendix B, 316–323, University College London (1991).
31. S. I. Fox, Human Physiology, 6th edition, McGraw-Hill Companies, Inc. Chapter 13, Heart and Circulation, 388–391(1999).
32. S. R. Arridge, M. Cope, and D. T. Delpy, “The theoretical basis for the determination of optical pathlengths in tissue: temporal and frequency analysis,” Phys. Med. Biol. , 37, 1531–1560 (1992). [CrossRef] [PubMed]
33. M. Ferrari, Q. Wei, L. Carraresi, R. A. de Blasi, and G. Zaccanti, “Time-resolved spectroscopy of the human forearm,” J. Photochem. Photobiol. B:Biol. , 16, 141–153 (1992). [CrossRef]
34. P. van der Zee, M. Cope, S. R. Arridge, M. Essenpreis, L. A. Potter, A. D. Edwards, J. S. Wyatt, D. C. McCormick, S. C. Roth, E. O. R. Reynolds, and D. T. Delpy, “Experimentally measured optical pathlengths for the adult head, calf and forearm and the head of the newborn infants as a function of inter optode spacing,” Adv. Exp. Med. Biol. , 316, 143–153 (1992). [CrossRef] [PubMed]
35. M. Kohl, C. Nolte, H. R. Heekeren, S. Horst, U. Scholz, H. Obrig, and A. Villringer, “Determination of the wavelength dependence of the differential pathlength factor from near-infrared pulse signals,” Phys. Med. Biol. , 43, 1771–1782 (1998). [CrossRef] [PubMed]
37. P. van der Zee, S. R. Arridge, M. Cope, and D. T. Delpy, “The effect of optode positioning on optical pathlength in near infrared spectroscopy of brain,” Adv. Exp. Med. Biol. 277, 79–84 (1990). [PubMed]
38. L. A. Paunescu, “Tissue blood flow and oxygen consumption measured with near infrared frequency-domain spectroscopy,” Degree of Doctor of Philosophy (University of Illinois, Urbana, 2001).
39. H. Liu, M. Miwa, B. Beauvoit, N.G. Wang, and B. Chance, “Characterization of absorption and scattering properties of small-volume biological samples using time-resolved spectroscopy,” Anal. Biochem. 213, 378–385 (1993). [CrossRef] [PubMed]
40. R. G. Steen, K. Kitagishi, and K. Morgan, “In vivo measurement of tumor blood oxygenation by near-infrared spectroscopy: Immediate effects of pentobarbital overdose or carmustine treatment,” J. Neuro-Oncol. 22, 209–220 (1994). [CrossRef]
41. R. P. Choudhury, V. Fuster, J. J. Badimon, E. A. Fisher, and Z. A. Fayad, “MRI and Characterization of Atherosclerotic Plaque,” Arteriosclerosis, Thrombosis, and Vascular Biology 22,1065–1072 (2002). [CrossRef] [PubMed]
42. J. R. Forder and G. M. Pohost, “Cardiovascular nuclear magnetic resonance: basic and clinical applications,” J. Clin. Invest. 111, 1630–1639 (2003). [PubMed]