ORIGINAL ARTICLE Annals of Nuclear Medicine Vol. 11, No. 2, 101-108, 1997 A method for the quantification of benzodiazepine receptors by using 123I-iomazenil and SPECT with one scan and one blood sampling Hiroshi ITO, Ryuta KAWASHIMA, Masamichi KOYAMA, Ryoui GOTO, Kazunori SATO, Shuichi ONO and Hiroshi FUKUDA Depanment of Nuclear Medicine and Radiology. Division of Brain Sciences. Institute ofDevelopment. Aging and Cancer, Tohoku University Iodine-123-iomazenil (Iomazenil) is a ligand of central type benzodiazepine receptors for single photon emission computed tomography (SPECT). Previously we reported a simple, table look-up method for quantification of its binding potential (BP) by using two SPECT scans and calibrated standard input function with one blood sampling. This method is based on a two-compartment model (K1: influx rate constant; k2: efflux rate constant; Vd (= K1/K2): the total distribution volume corresponding BP), and requires two SPECT scans for calculating both Kl and Vd values. If the K1 value in the two-compartment model can be assumed to be constant, the radioactivity of one SPECT scan at 180 min after injection can be considered to tabulate as a function of Vd for a given K1 value and a given input function, and a table look-up procedure provides the corresponding Vd Value. The purpose of this study was to develop a simple, autoradiographic method for quantification of BP by using one SPECT scan and calibrated standard input function with one blood sampling. SPECT studies were performed on 14 patients. A dynamic SPECT scan was initiated following an intravenous bolus injection of Iomazenil. A static SPECT scan was performed at 180 min after the injection. Frequent blood sampling from the brachial artery was performed on all subjects to determine the arterial input function. Simulation studies revealed that errors in calculated Vd Values were around +- 10-l5% for varied K1 values. A good correlation was observed between total distribution volume values calculated by three-compartment model analysis and those calculated by the present method (r = O.90), supporting the validity of this method. The present method is simple and applicable for clinical use, and will be able to provide images of BP. Key words: iodine-123-iomazenil, SPECT, benzodiazepine receptor, brain INTRODUCTION IODINE-123-IOMAZENIL (Iomazenil), a ligand for central type benzodiazepine receptors, is suitable for single photon emission computed tomography (SPECT) measurement due to its high affinity to bcnzodiazepine receptors in the brain and small non-specific binding.1-4 Changes in central benzodiazepine receptor binding have been reported in some diseases.5-15 Quantitative assessment of neuroreceptors using positron emission tomography (PET) and SPECT has been reported for dopamine D2 receptors16-20 and central type benzodiazepine receptors.21-28 Although the binding potential (BP equal to Bmax/KD, Bmax: receptor density; 1/KD: affinity to receptors) 16 can be obtained by kinetic analysis with a single bolus injection,21,22,24,26-30 this method necessitates dynamic scan and frequent arterial blood sampling with separation of unchanged radioligand for each sample, and therefore is not applicable for routine clinical use. We previously reported a simple, table look-up method for quantification of BP of Iomazenil using two SPECT scans and calibrated standard input function with one blood sampling,31 which was originally developed as a method of measuring cerebral blood flow (CBF) and distribution volume using N-isopropyl-p-[123I]iodoam-phetamine and SPECT.32-35 This method is based on a two-compartment model (Kl: influx rate constant; k2: efflux rate constant; Vd (= K1/k2): the total distribution volume corresponding BP), and requires two SPECT scans for calculating both Kl and Vd Values, but, limited reproducibility of selecting regions-of-interest in the same location on two SPECT images (early and delayed scan images) might hamper accurate calculation of Vd Values. Since the registration between two SPECT images is difficult, it will also be difficult to calculate an image of Vd by the table look-up method. If the Kl value can be assumed to be constant, the Vd value can be obtained from one SPECT scan, which will remove the error caused by limited reproducibility in selecting regions-of-interest and allow Vd images to be calculated. In addition, it will improve the throughput of clinical SPECT examinations. The purpose of this study was to develop a simple, autoradiographic method for quantification of BP by using one SPECT scan and calibrated standard input function with one blood sampling. MATERIALS AND METHODS Subjects SPECT studies were performed on 14 patients with cerebrovascular diseases, dementia or brain tumors (age range, 33-71 yr; mean age +- SD, 56.0 +- 12.2 yr). None of the patients had any heart, renal or liver diseases. All patients received potassium iodide to minimize thyroid uptake of free radioactive iodine. Informed consent was obtained from each subject. The project was approved by the Clinical Test Committee of the Institute of Development, Aging and Cancer, Tohoku University. No side effects were observed in any of the subjects after administration of lomazenil. SPECT study Iomazenil was supplied by Nihon Medi-Physics Co., Ltd. The specific activity of lomazenil was 92.5 GBq/umol. The amount of the ligand administered was 0.5 /rg (1.2 nmol)/111 MBq. Radiochemical purity was more than 95%. A dynamic SPECT scan was initiated following an intravenous bolus injection of 111 (one subject) or 167 MBq (13 subjects) of Iomazenil. The dynamic scan sequence consisted of sixteen 200 second scans with 360deg continuous rotation of the camera. A static SPECT scan was performed at 180 min after the injection. Frequent blood sampling from the brachial artery was performed on all subjects to determine the arterial input function, and the radioactivity concentrations of the octanol-extracted component from whole blood were measured. In seven subjects, the radioactivity of the octanol-extracted component from plasma were also measured. In this study, the radioactivity concentrations of the lipophilic component extracted by octanol from plasma were considered as representative of true arterial input function.28.31.36 The arterial input function for each subject was calculated by using mean ratios of the radioactivity of the octanol-extracted component from whole blood to that from plasma in these seven subjects, same as our previous report.31 One SPECT scanner (SPECT-2000H, Hitachi Medico Corp., Tokyo, Japan),37 a four-head rotating gamma camera with in-plane and axial resolutions of 14 mm full width at half maximum (FWHM), was used for all measurements. Image reconstruction was performed by filtered backprojection using a Butterworth filter and attenuation correction was made numerically by assuming the object shape to be an ellipse for each slice and the attenuation coefficient to be uniform (0.08 cm-1).38,39 Image matrix size was 64 x 64. Correction for scattered photons was not performed. Image slices were set up parallel to the orbitomeatal (OM) line and obtained at 8 mm intervals through the whole brain. A cross-calibration scan was performed using a cylindrical uniform phantom 16 cm in inner diameter to calibrate sensitivity between the SPECT scanner and the well counter system. X-ray CT scans were also obtained with the same slices as those of SPECT images in all subjects. Image analysis Data analysis was performed on a UNIX work station (TITAN-750, Kubota computer Corp.. Tokyo, Japan). Regions-of-interest in the cerebellum, brain stem, thalamus, basal ganglia, centrum semiovale and cerebral cortex regions including the frontal, temporal, parietal and occipital lobes were outlined on SPECT images, by referring to X-ray computed tomographic images of the patients. The shapes of regions-of-interest were circular with a 20 mm diameter for the brain stem, and elliptic with a short axis of 16-20 mm and a long axis of 28-40 mm for other regions. Theory (Autoradiographic method) In the present method, a two-compartment model was employed in line with our previous report.31 In this model, the following model equation can be expressed: where Cb(t) = radioactivity in the brain and Ca(t) = arterial input function.* denotes the convolution integral. For a given K1 value and a given arterial input function, Ca(t), the radioactivity in the brain. Cb(t), can be considered to tabulate as a function of Vd. For a given radioactivity in the brain, the table look-up procedure then provides a corresponding Vd Value (Fig. 1). The Vd value approximates the BP value in case of Iomazenil.31 In this study, the midscan time of SPECT scan was set at 180 min after the injection of Iomazenil. The arterial input function, Ca(t), is obtained by calibration of the standard input function using the arterial whole blood radioactivity gained from the single sampling at 30 min after the injection of Iomazenil, same as our previous report.31 The standard input function was obtained by averaging arterial input functions of all subjects.31 Simulation stndy of K1 range of Iomazenil In order to investigate the range of the K1 value of Iomazenil, a simulation study was performed. The firstpass extraction fraction of Iomazenil (E) is probably about 0.5.26,28 First, the first-pass extraction fraction of Iomazenil was assumed to be 0.5 for a CBF of 0.500 ml/ml/min, which corresponds to a capillary permeability-surface area product (PS) value of 0.347 ml/ml/min, and then K1 values of Iomazenil were calculated by using this PS value according to the Renkin-Crone model (E = 1 - e^-(PS/CBF)),40,41 where CBF was varied from O.150 to 1.000 ml/ml/min. Next, tissue time-activity curves (0-180 min) were generated based on the three-compartment model using these calculated Kl values and assumed rate constants, where total and nondisplaceable distribution volume values were assumed to be 30 and 3 ml/ml, respectively (k3 = 0.180 min-1 and k4 = 0.020 min-1).26,28,31 Then, the table look-up method previously reported31 was applied to these tissue data for calculating K1 and Vd values, and these K1 values were compared with CBF and K1 values calculated by the Renkin-Crone model. In the table look-up method, the combination of scan times (early/delayed SPECT scan) used was 30/180 min. In this simulation, the standard input function was used as an input function. Simulation study of the error of Vd calculated by the present method In order to estimate the error of Vd Values calculated by the present method, another simulation study was also per-formed. First, tissue time-activity curves (0-180 min) were generated based on the three-compartment model using four assumed rate constants, where K1 was varied 0.135-0.293 ml/ml/min corresponding to a CBF of 0.150-1.000 ml/ml/min (PS = 0.347 ml/ml/min), total distribution volume values were assumed to be 10,20,30 and 40 ml/ml (k4 = 0.020 min-1), and nondisplaceable distribution volume value was assumed to be 3 ml/ml.26,28,31 Then, the present method was applied to these tissue data for calculating Vd Values, and these Vd Values were com-pared with assumed Vd Values. In this simulation, the standard input function was used as an input function. Non-linear least squares ftting analysis (NLLSF) In order to confirm the validity of the present method, NLLSF based on the three-compartment model was performed to determine four rate constants,42 and then total distribution volume (Vd (= (K1/k2)*(1 + k3/k4))) values were calculated.26,28,31 In these analyses, dynamic SPECT data and static SPECT scan data at 180 min were combined and used. However, first frame data (0-200 sec) of dynamic SPECT were not used for removing the effects of cerebral blood volume.43 The individually measured arterial input function was used in these analyses. RESULS Figure 2 shows the range of Iomazenil K1 calculated by the Renkin-Crone model (K1(true)) and that calculated by the table look-up method (K1(TLU)) for a CBF range of 0.150-1.000 ml/ml/min. A CBF range of 0.150-1.000 ml/ml/min corresponded to a K1(true) range of 0.135-0.293 ml/ml/min, and this K1(true) range corresponded to a K1(TLU) range of 0.103-0.183 ml/ml/min. A CBF of 0.500 ml/ml/ min corresponded to a K1(true) of 0.250 ml/ml/min and to a K1(TLU) of 0.164 ml/ml/min. Mean Kl value obtained by NLLSF in cerebral cortex regions, which showed no abnormalities on X-ray CT, was 0.270 d: 0.053 ml/ml/min (+-SD). Mean Kl value obtained by the table look-up method31 with individually measured input function and calibrated standard input function (scan times combination: 30/180 min) in cerebral cortex regions were 0.166+-0.030 and 0.166+-0.039 ml/ml/min (+-SD), respectively. Figure 3 shows the error of Vd values calculated by the present method in which the K1 value was assumed to be 0.166 ml/ml/min. When CBF values were 0.250-1.000 ml/ml/min corresponding to a K1(true) range of 0.188-0.293 ml/ml/min, the error in calculated Vd Values were around +-10% and +-15% for assumed Vd of 10-30 ml/ml and 40 ml/ml, respectively. However, when CBF value was 0.150 ml/ml/min corresponding to a K1(true) of 0.135 ml/ml/min, the error in calculated Vd Values were about - 20% for assumed Vd of 30 and 40 ml/ml, while they were around +-10-15% for assumed Vd Of 10 and 20 ml/ml. Mean Vd Value obtained by NLLSF in cerebral cortex regions, which showed no abnormalities on X-ray CT, was 22.09+-3.95 ml/ml (+-SD). Figure 4 shows the correlation between Vd Values obtained by NLLSF and those obtained by the autoradiographic method using individually measured input function, in which the K1 value was assumed to be 0.166 ml/ml/min (y =0.91x + 1.24; r = 0.99; x, NLLSF; y, autoradiographic method). A good correlation was observed between the two methods. Figure 5 shows the correlation between Vd values obtained by the autoradiographic method using individually measured input function and those obtained by the autoradiographic method using calibrated standard input function, that is, the present method in which the Kl value was assumed to be 0.166 ml/ml/min (y = 1.10x-1.01; r = 0.91 ; x, using individual input function; y, using standard input function). A good correlation was observed between the two methods. Figure 6 shows the correlation between Vd Values obtained by NLLSF and those obtained by the autoradiographic method using calibrated standard input function, that is, the present method in which the K1 value was assumed to be 0.166 ml/ml/min ty = 1.01x + 0.20; r = 0.90; x, NLLSF; y, autoradiographic method). A good correlation was observed between the two methods. Figure 7 shows the correlations between Vd Values obtained by the table look-up method previously reported31 and those obtained by the autoradiographic method, in which the Kl values were assumed to be 0.130, 0.166 or 0.200 ml/ml/min. In the table look-up method, the combination of scan times (early/delayed) used was 30/180 min. In each method, calibrated standard input function was used. Good correlations were observed between the two methods for all assumed K1 values (K1 = 0.130: y = 1.25x - 2.94, r = 0.94; K1 = 0.166: y = 0.97x + 1.25, r = 0.98; K1 = 0.200: y = 0.87x + 3.08, r = 0.98; x, table look-up method; y, autoradiographic method). However, the best correlation between the two methods was ob-served when the Kl value was assumed to be 0.166 ml/ml/min. DISCUSSION Validation o the resent method It has been shown that the two-compartment model analysis, which is employed in the table look-up method31 and the present method, caused underestimation of K1 values compared with the three-compartment model analysis in case of Iomazenil.28,31 In the simulation study, the K1(TLU) values were smaller than K1(true), and the range of K1(TLU) was narrow compared with K1(true) (Fig. 2). This indicates that assuming K1 values in the two-compartment model analysis might not reveal significant errors in the calculation of Vd. In this study, mean Kl value obtained by the table look-up method was 0.166 ml/ml/min which was smaller than that obtained by NLLSF (0.270 ml/ml/min). Therefore the K1 value was assumed to be 0.166 ml/ml/ min in the present method. The simulation study revealed that the error of Vd values calculated by the present method was not significant when CBF values were 0.250-1.000 ml/ml/min (Fig. 3). However, significant underestimation was observed when the CBF value was 0.150 ml/ml/min and assumed Vd Values were 30-40 ml/ml. This indicates that the present method can provide accurate Vd values except in the regions which show very low CBF and normal or high Vd. Mean Vd value obtained by NLLSF was 22.09+-3.95 ml/ml (+-SD), and those previously reported were 27-31 ml/ml in healthy subjects.27 Vd values obtained by the autoradiographic method using individually measured input function were consistent with those obtained by NLLSF (Fig. 4). This indicates that the statistical noise of the table between SPECT counts and Vd in the autoradiographic method (Fig. 1) was low. A good correlation was observed between Vd Values obtained by the autoradiographic method using individually measured input function and those obtained by the autoradiographic method using calibrated standard input function (i.e., the present method) (Fig. 5). This indicates that the individual difference in input function was small, and that calibrated standard input function can be used as an input function. A good correlation was observed between Vd Values obtained by NLLSF and those obtained by the autoradiographic method using calibrated standard input function (i.e., the present method) (Fig. 6), support-ing its validity. Since the present method requires only one SPECT scan at 180 min after injection and one blood sampling, it is applicable for routine clinical use, and will be able to provide images of BP. Furthermore, venous blood sampling can be substituted for invasive arterial sampling, thus simplifying the approach.31,34,44 In addition, the separation of lipophilic component in a blood sample for calibrating the standard input function might be able to improve the error of individual difference in input function. In this study, the blood sampling time was set at 30 min after injection, however it can be set at 20-40 min.31 In the present method, SPECT data at 180 min after injection were used. If the time of SPECT scan can be set later than 180 min, e,g., 300 or 360 min, the error of Vd values calculated by the present method should be smaller. Further studies will be required. Mean Vd Value obtained by NLLSF in the present study (22.09 ml/ml) was smaller than those previously reported in healthy subjects (27-31 ml/ml).27 The main reason for this must be that all subjects in the present study were patients, not healthy subjects. Further studies for deter-mining normal Vd Values in a major healthy subject series are required. Limitations The following factors need to be considered as sources of error in the present method. 1. Error in the regions with low CBF and normal or high Vd. 2. Assumed K1 value 3. Physical accuracy of SPECT scanner system 4. Individual differences of input function Simulation study revealed that the present method caused underestimation of Vd Values in the regions with very low CBF and normal or high Vd (Fig. 3), while a good correlation was observed between Vd values obtained by NLLSF and those obtained by the present method (Fig. 6). Hatazawa et al. reported that normal Iomazenil uptake was observed in remote areas with hypoperfusion in a patient with cerebral infarction.15 The present method might cause underestimation of Vd Values in such regions. Further studies are required. In addition, the table look-up method previously reported may also not be able to calculate accurate Vd Values in such regions, because the normal Vd regions with hypoperfusion will show very small k2 values, hampering accurate calculation.31 While good correlations were observed between Vd values obtained by the table look-up method and those obtained by the present method for all assumed K1 values, the best correlation between the two methods was observed when the assumed K1 value was 0.166 ml/ml/min (Fig. 7). This K1 value was the mean K1 value obtained by the table look-up method. Since the differences in the reconstruction algorithm, including the scattered photon correction, the attenuation correction and the filter for each SPECT scanner system, might cause the differences in calculated K1 values, the assumed K1 value in the present method should be determined for each SPECT scanner s stem by using the table look-up method.45,46 The scattered photons not removed in this study can cause errors in Vd estimation. Incomplete correction for the attenuation of gamma rays is also a problem for the calculation of accurate Vd Values. The differences in the reconstruction algorithm for each SPECT scanner system might cause errors in calculating Vd Values. The limited spatial resolution of SPECT scanners cause gray-white matter mixture in regions-of-interest, and this might give rise to non-negligible errors. Further studies are required on these subjects.31 In this study, none of the patients had any heart, renal or liver diseases. Since the shape of the arterial input function in patients with these disorders will differ from the standard input function,32 accurate Vd Values may not be obtained for such patients by the present method.31 CONCLUSION We have developed a simple method for quantification of Iomazenil binding, requiring one SPECT scan, one blood sampling and an assumed Kl value. A good correlation was observed between Vd Values obtained by NLLSF and those obtained by the present method, supporting its validity. The present method is simple and applicable for routine clinical use, and it will be able to provide images of BP. However, it does have some limitations, and further studies are required. ACKNOWLEDGMENTS We are greatly indebted to the staff of the Institute of Development, Aging and Cancer, Tohoku University and in particular to Tachio Sato for handling the SPECT scanner. We thank Yoshihiro Onishi in Nihon Medi-Physics Co., Ltd. for providing the computer program of non-linear least squares fitting. This study was supported by a Grant-in-Aid for Scientific Research (05454297) from the Japanese Ministry of Education. Science and Culture. REFERENCES 1. Beer H-F. Blauenstein PA, Hasler PH, Delaloye B, Riccabona G, Bangerl I, et al. In vitro and in vivo evaluation of iodine-123-Ro 16-0154: A new imaging agent for SPECT investigations of benzodiazepine receptors. J Nucl Med 31 : 1007-1014, 1990. 2. Innis RB, Zoghbi SS. Johnson EW, Woods SW, Al-Tikriti MS, Baldwin RM, et al. SPECT imaging of the benzodiazepine receptor in non-human primate brain with [123I]Ro 16-0154. Eur J Pharmacol 193: 249-252, 1991. 3. Innis RB, Al-Tikriti MS, Zoghbi SS, Baldwin RM, Sybirska EH, Laruelle M, et al. SPECT imaging of the benzodiazepine receptor: Feasibility of in vivo potency measurements from stepwise displacement curves. J Nucl Med 32: 1754-1761, 1991. 4. Woods SW, Seibyl JP, Goddard AW, Dey HM. Zoghbi SS, Germine M, et al. Dynamic SPECT imaging after injection of the benzodiazepine receptor ligand [1231]iomazenil in healthy human subjects. Psych Res Neuroimaging 45: 67-77, 1992. 5. Huffelen AC, Isselt JW, Veelen CWM, Rijk PP, Bentum AM, Dive D, et al. Identification of the side of the epileptic focus with 123I-Iomazenil SPECT. A comparison with rsFDG-PET and ictal EEG findings in patients with medically intractable complex partial seizures. Acta Neurochir Suppl 50: 95-99, 1990. 6. Bartenstein P. Ludolph A, Schober O. Lottes G, Scheidhauer K. Sciuk J, et al. Benzodiazepine receptors and cerebral blood flow in partial epilepsy. Eur J Nucl Med 1 8: 111-118, 1991. 7. Schubiger PA, Hasler PH, Beer-Wohlfahrt H. Bekier A, Oettli R, Cordes M, et al. Evaiuation of a multicentre study with iomazenil - a benzodiazepine receptor ligand. Nucl Med Commun 12: 569-582, 1991. 8. Cordes M, Henkes H, Ferstl F, Schmitz B, Hierholzer J, Schmidt D, et al. Evaluation of focal epilepsy: A SPECT scanning comparison of 123-I-iomazenil versus HM-PAO. AJNR 13: 249-253, 1992. 9. Johnson EW, Lanerolle NC, Kim JH, Sundaresan S, Spencer DD, Mattson RH, et al. "Central" and "peripheral" benzodiazepine receptors: Opposite changes in human epileptogenic tissue. Neurology 42: 811-815, 1992. 10. Owen F, Poulter M, Waddington JL, Mashal RD, Crow TJ. [3H]R05-4864 and [3H]flunitrazepam binding in kainatelesioned rat striatum and in temporal cortex of brains from patients with senile dementia of the Alzheimer type. Brain Res 278: 373-375, 1983. 11. Reisine TD, Wastek GJ, Speth RC, Bird ED, Yamamura HI. Alterations in benzodiazepine receptor of Huntington's diseased human brain. Brain Res 165: 183-187, 1979. 12. Onodera H, Sato G, Kogure K. GABA and benzodiazepine receptors in the gerbil brain after transient ischemia: Demonstration by quantitative receptor autoradiography. J Cereb Blood Flow Metab 7: 82-88, 1987. 13. Sette G, Baron JC, Young AR, Miyazawa H, Tillet I, Barre L, et al. In vivo mapping of brain benzodiazepine receptor changes by PET after focal ischemia in the anesthetized baboon. Stroke 24: 2046-2058, 1993. 14. Al-Tikriti MS, Dey HM, Zoghbi SS. Baldwin RM, Zea-Ponce Y. Innis RB. Dual-isotope autoradiographic measurement of regional blood flow and benzodiazepine receptor availability following unilateral middle cerebral artery occlusion. Eur J Nucl Med 21 : 196-202, 1994. 15. Hatazawa J. Satoh T, Shimosegawa E, Okudera T, Inugami A. Ogawa T, et al. Evaluation of cerebral infarction with iodine 123-iomazenil SPECT. J Nucl Med 36: 2151-2161, 1995. 16. Mintun MA, Raichle ME, Kilbourn MR, Wooten GF, Welch MJ. A quantitative model for the in vivo assessment of drug binding sites with positron emission tomography. Ann Neurol 15: 217-227, 1984. 17. Wong DF, Gjedde A, Wagner HN Jr. Quantification of neuroreceptors in the living human brain. I. Irreversible binding of ligands. J Cereb Blood Flow Metab 6: 137-146, 1986. 18. Wong DF, Gjedde A, Wagner HN Jr, Dannals RF. Douglass KH, Links JM, et al. Quantification of neuroreceptors in the living human brain. II. Inhibition studies of receptor density and affinity. J Cereb Blood Flow Metab 6: 147-153, 1986. 19. Farde L, Eriksson L, Blomquist G, Halldin C. Kinetic analysis of central [11C]raclopride binding to D2-dopamine receptors studied by PET - A comparison to the equilib-rium analysis. J Cereb Blood Flow Metab 9: 696-708, 1989. 20. Blomqvist G. Pauli S, Farde L, Eriksson L. Maps of receptor binding parameters in the human brain - a kinetic analysis of PET measurements. Eur J Nucl Med 16: 257-265, 1990. 21. Koeppe RA, Holthoff VA, Frey KA, Kilbourn MR, Kuhl DE. Compartmental analysis of [11C]flumazenil kinetics for the estimation of ligand transport rate and receptor distribution using positron emission tomography. J Cereb Blood Flow Metab 11 : 735-744, 1991. 22. Holthoff VA, Koeppe RA, Frey KA, Paradise AH, Kuhl DE. Differentiation of radioligand delivery and binding in the brain: Validation of a two-compartment model for [11C]flumazenil. J Cereb Blood Flow Metab 11 : 745-752, 1991. 23. Price JC, Mayberg HS, Dannals RF, Wilson AA, Ravert HT, Sadzot B, et al. Measurement of benzodiazepine receptor number and affinity in humans using tracer kinetic modeling, positron emission tomography, and [11C]flumazenil. J Cereb Blood Flow Metab 13: 656-667, 1993. 24. Laruelle M, Baldwin RM, Rattner Z, Al-Tikriti MS, Zea-Ponce Y, Zoghbi SS, et al. SPECT quantification of [123I]iomazenil binding to benzodiazepine receptors in non-human primates: I. Kinetic modeling of single bolus experiments. J Cereb Blood Flow Metab 14: 439-452, 1994. 25. Laruelle M, Abi-Dargham A, Al-Tikriti MS, Baldwin RM, Zea-Ponce Y, Zoghbi SS, et al. SPECT quantification of [123I]iomazenil binding to benzodiazepine receptors in non-human primates: 11. Equilibrium analysis of constant infusion experiments and correlation with in vitro parameters. J Cereb Blood Flow Metab 14: 453-465, 1994. 26. Abi-Dargham A. Laruelle M, Seibyl J. Rattner Z. Baldwin RM, Zoghbi SS, et al. SPECT measurement of benzodiazepine receptors in human brain with iodine- 123-iomazenil: Kinetic and equilibrium paradigms. J Nucl Med 35: 228-238, 1994. 27. Abi-Dargham A, Gandelman M, Zoghbi SS. Laruelle M, Baldwin RM. Randall P, et al. Reproducibility of SPECT measurement of benzodiazepine receptors in human brain with iodine-123-iomazenil. J Nucl Med 36: 167-175, 1995. 28. Onishi Y. Yonekura Y, Mukai T. Nishizawa S. Tanaka F, Okazawa H, et al. Simple quantification of benzodiazepine receptor binding and ligand transport using iodine-123-iomazenil and two SPECT scans. J Nucl Med 36: 1201-1210, 1995. 29. Huang SC, Barrio JR, Phelps ME. Neuroreceptor assay with positron emission tomography: Equilibrium versus dynamic approaches. J Cereb Blood Flow Metab 6: 515-521, 1986. 30. Lassen NA. Neuroreceptor quantitation in vivo by the steady-state principle using constant infusion or bolus injection of radioactive tracers. J Cereb Blood Flow Metab 12: 709-716, 1992. 31. Ito H, Goto R, Koyama M, Kawashima R, Ono S, Sato K, et al. A simple method for the quantification of benzodiaz-epine receptors using iodine-123 iomazenil and single-photon emission tomography. Eur J Nucl Med 23 : 782-791, 1996. 32. Iida H. Itoh H, Bloomfield PM, Munaka M, Murakami M, Inugami A, et al. A method to quantitate cerebral blood flow using a rotating gamma camera and iodine-123 iodoamphetamine with one blood sampling. Eur J Nucl Med 21 : 1072-1084, 1994. 33. Itoh H, Iida H, Murakami M, Bloomfield PM. Miura S, Okudera T, et al. A method for measurement of regional cerebral blood flow using N-isopropyl-p-[123I]iodoamphetamine (123I-IMP) SPECT; two scans with one point blood sampling technique. KAKU IGAKU (Jpn J Nucl Med) 29: 1193-1200, 1992. 34. Ito H, Koyama M, Goto R, Kawashima R, Ono S, Atsumi H, et al. Cerebral blood flow measurement with iodine-123-IMP SPECT, calibrated standard input function and venous blood sampling. J Nucl Med 36: 2339-2342, 1995. 35. Ito H, Ishii K, Atsumi H, Kinoshita T. Kawashima R. Ono S, et al. Error analysis of table look-up method for cerebral blood flow measurement by 123I-IMP brain SPECT: Comparison with conventional microsphere model method. Ann Nucl Med 9: 75-80, 1995. 36. Zoghbi SS, Baldwin RM, Seibyl JP, Al-Tikriti MS. Zea-Ponce Y. Laruelle M, et al. Phannacokinetics of the SPECT benzodiazepine receptor radioligand [123Ijiomazenil in human and non-human primates. Nucl Med Biol 19: 881-888, 1992. 37. Kimura K, Hashikawa K. Etani H, Uehara A, Kozuka T, Moriwaki H, et al. A new apparatus for brain imaging: four-head rotating gamma camera single-photon emission computed tomography. J Nucl Med 31 : 603-609, 1990. 38. Chang LT. A method for attenuation correction in radionuclide computed tomography. IEEE Trans Nucl Sci 25: 638-643,1978. 39. Chang LT. Attenuation correction and incomplete projection in single photon emission computed tomography. IEEE Trans Nucl Sci 26: 2780-2789, 1979. 40. Renkin EM. Transport of potassium-42 from blood to tissue in isolated mammalian skeletal muscles. Am J Physiol 197: 1205-1210, 1959. 41. Crone C. Permeability of capillaries in various organs as determined by use of the indicator diffusion method. Acta Physiol Scand 58: 292-305, 1963. 42. Nelder JA, Mead R. A simplex method for function minimization. Computer J 7: 308-312, 1965. 43. Koeppe RA, Mangner T, Betz AL, Shulkin BL, AIlen R, Kollros P, et al. Use of [11C]aminocyclohexanecarboxylate for the measurement of amino acid uptake and distribution volume in human brain. J Cereb Blood Flow Metab 10: 727-739, 1990. 44. Onishi Y, Yonekura Y, Nishizawa S, Tanaka F, Okazawa H, Ishizu K, et al. Noninvasive quantification of iodine-123-iomazenil SPECT. J Nucl Med 37: 371-378, 1996. 45. Ito H, Ishii K, Atsumi H, Inukai Y. Abe S, Sato M, et al. Error analysis of autoradiography method for measurement of cerebral blood flow by 123I-IMP brain SPECT: A comparison study with table look-up method and microsphere model method. Ann Nucl Med 9: 185-190, 1995. 46. Iida H, Itoh H, Nakazawa M, Hatazawa J, Nishimura H, Onishi Y, et al. Quantitative mapping of regional cerebral blood flow using iodine-123-IMP and SPECT. J Nucl Med 35: 2019-2030, 1994.