ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 9, No.4, 185-190, 1995 
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 
Hiroshi ITO,* Kiyoshi ISHII ** Hrroto ATSUMI ** Yoshimasa INUKAI ** Shigeto ABE,** Masami SATO,** Toshifumi KINOSHITA,** Ryuta KAWASHIMA,* Shuichi ONO* and Hiroshi FUKUDA* 
*Department of Nuclear Medicine and Radiology, Division of Brain Sciences, lnstitute ofDevelopment, Aging and Cancer, Tohoku University **Department of Radiology, Sendai City Hospital 
N-isopropyl-p-[123I]iodoamphetamine (IMP) has been commonly used as a cerebral blood flow tracer, but, significant clearance of IMP from the brain to the blood causes underestimation of cerebral blood flow (CBF) as compared with true CBF when the conventional microsphere model method is applied. Previously, we reported an "Autoradiography method" (ARG method) for measuring CBF by using IMP in which this clearance effect was corrected. This method was based on a two-compartment model (influx: Kl, efflux: k2, Kl/k2 = distribution volume of IMP (Vd)), the Kl (corresponding to CBF) being obtained from the table which showed a correlation between CBF and the brain counts of SPECT scan with a constant Vd Value. Arterial input data used were obtained by one point blood sampling 10 min after IMP infusion against the standard input function. In the present study, the ARG method was compared with the table look-up method (TLU method) and the conventional microsphere model method (MS method) for 30 subjects. When the Vd Value in the ARG method was assumed to be 50 ml/ml, CBF values obtained by the ARG method were correlated well with those obtained by the TLU method (Y= 1.04X - 2.5; X: TLU, Y: ARG, r = 0.97) and those obtained by the MS method (Y = 0.82X + 12.1 ; X: ARG, Y: MS, r = 0.84). But, when the Vd Value was assumed to be more or less than 50 ml/ml, ARG method CBF were under-or overestimated compared with the TLU method. This indicated that the ARG method could be a reliable method for CBF measurement if the Vd Was determined properly. CBF values obtained by the MS method were actually 13.2% higher than those obtained by the ARG method against previous studies. As reasons for this, errors In the effects of gray-white matter mixture in the ARG method and in estimation of the SPECT brain counts at 8 min in the MS method were considered. 
Key words: IMP, SPECT, cerebral blood flow, autoradiography method 
INTRODUCTION 
IODINE-123 Iabeled N-isopropyl-p-iodoamphetamine (IMP) has been used as a cerebral blood flow (CBF) tracer for single photon emission computed tomography (SPECT) due to its large extraction fraction and high affinity for the brain. 1,2 But its significant clearance from the brain caused a change in IMP distribution3,4 and underestimation of CBF when a conventional microsphere model method5 was applied to prolonged data acquisition.6-9 Previously, we reported an "Autoradiography method" (ARG method) with IMP, a new simple approach to the measurement of CBF, in which this clear-ance effect was corrected. 10,11 This approach was based on a two-compartment model (influx: Kl, efflux: k2, Kl/ k2 = distribution volume of IMP (Vd)) in which the Vd value was assumed to be constant, and the Kl value (corresponding to CBF) was obtained from the table which showed a correlation between CBF and the brain counts of the SPECT scan (mid-scan time: 40 min). Arterial input function was obtained by calibrating against the standard input function from one point arterial blood sampling at 10 min after intravenous infusion of IMP. It has been considered that the CBF value obtained by the ARG method would vary according to the assumed Vd value, and the Vd Value would be dependent on the intrinsic performance of the SPECT scanner.l0 Previously, we also reported a "Table look-up method" (TLU method) with IMP. 12-14 This TLU method was also based on a two-compartment model, in which CBF and Vd Pair could be obtained from two SPECT scans and one point arterial blood sampling. The purpose of the present study was to compare the ARG method with the TLU method and with the conventional microsphere model method (MS method)15 for a major patient series. 
MATERIALS AND METHODS 
Subjects 
SPECT studies were performed in 30 subjects including 19 patients with cerebral contusion, 3 with cerebrovascular disease, 2 with hypoxic brain, 2 with carbon monoxide toxicosis and 4 normal volunteers. None of the patients had any heart or lung disease. A informed consent was obtained from each subject. 
SPECT study 
Two SPECT scans were performed, at 40 min and 180 min of mid-scan time, after intravenous infusion of 222 MBq IMP for 1 min. Planar brain images for 50 sec were obtained at 8 min after IMP infusion for the conventional microsphere model method (MS method) (Fig. l). The SPECT scanner used was a Neurocam (Yokogawa Medical Systems Corp., Tokyo, Japan),16 equipped with a three-head rotating gamma camera. In-plane resolution was 9 mm full width at half maximum (FWHM), and axial resolution was 1O mm FWHM with low energy high resolution (LEHR) collimators. The SPECT scan protocol acquired 64 projections at 50 sec per projection with 120' rotation of the camera. Reconstruction was per-formed by filtered backprojection with a Butterworth 17 filter (cutoff frequency 0.45 cycle/cm, power factor 1O). Attenuation correction was made numerically by as-suming the object shape to be circular or elliptical and the attenuation coefficient to be uniform (0.12 cm-1) (Sorenson's methodl8). The scattered photons were not corrected. Image slices were set up parallel to the orblto-meatal (OM) line and obtained at 8 mm intervals through the whole brain. 
One point arterial blood sampling from the brachial artery was performed at 1O min after IMP infusion. Radio-activity of the whole blood was measured with a well counter and was used for calibration against the standard input function to provide an arterial input function for the autoradiography method (ARG method) and the table look-up method (TLU method). Continuous arterial blood sampling at a constant rate from the brachial artery was also performed during the first 8 min after IMP infusion, and octanol extracted radioactivity was measured for the MS method (Fig. l). 
A cross calibration scan was performed with an elliptic cylindrical uniform phantom (long axis: 1 9 cm, short axis: 14 cm inner diameter) for calibrating the relative sensitivities of the SPECT scanner and the well counter system . 
Image analysis 
Regions-of-interest were placed in the cerebellum, pons, thalamus, putamen, centrum semiovale and cerebral cortex including frontal, temporal, parietal and occipital lobes on the 40 and 180 min SPECT images. The shape of regions-of-interest was circular: 35 mm diameter for the cerebellum, and elliptic with short axes of 16-25 mm and long axes of 25-50 mm for other regions. 
Theory Autoradiography method (ARG method)10,ll: In this method, a two-compartment model was employed in line with previous reports.6-8,19,20 
In this study, we assumed the first-pass extraction fraction of IMP to be equal to 11¥2,21 and therefore, Kl equals CBF. The ratio of Kl to k2 is called the distribution volume of IMP (Vd (ml/ml)). Solving Eq. I provides: 
For a given Vd Value (= Kl/k2) and a given arterial input function, C.(t), Eq. 2 provides a unique relation between Cb(t) and Kl; then Kl values (corresponding to CBF) are obtained. The arterial input function, C.(t) is obtained by calibration against the standard input function by using the arterial blood radioactivity gained from the one point sampling. 
Table look-up method ( TLU method) 12-14: In this method, a two-compartment model was also employed. For this method, two SPECT scans were performed. The model equation (Eq. 2) can therefore be expressed for each scan. 
where tc and td are mid-scan times at first and second scans, respectively. Calculating the ratio of Eq. 3a to Eq. 3b gives: 
For a given arterial input function, C,(t), the radioactiv-ity ratio of the first to second scans (the right side of Eq. 4) can be considered to tabulate as a function of k2. Then, for a given radioactivity ratio of first to second scans, the table look-up procedure provides a corresponding k2 value, and the K1 value can be calculated by inserting this k2Value into Eq. 3a. The arterial input function is obtained by one point arterial blood sampling as in the ARG method. 
Microsphere mod el method (MS method ) 1 5 . 
CBF values were also calculated by microsphere model analysis as follows: 
In this study, SPECT data at 8 min (Cb) were 40 min SPECT data calibrated by the count ratio of 8 min/40 min whole brain planar images. 
Simulation of the effects of gray-white matter mixture The limited spatial resolution of the SPECT scanner causes a gray-white matter mixture in regions-of-interest. The effects of gray-white matter mixing on CBF values calculated by the ARG method were evaluated.22 Brain counts in heterogeneous tissue on a 40 min SPECT scan were generated as mixtures of gray and white matter. CBF values for the gray and white matter were assumed to be 80 and 20 ml/1OO ml/min, respectively. The Vd Valut for gray and white matter was assumed to be the same as 40, 50 or 60 ml/ml. The difference between true CBF values (= 80 ml/100 ml/min x gray matter fraction + 20 ml/1OO ml/min x white matter fraction) and CBF values calculated by the ARG method by using the generated hetero-geneous tissue radioactivity were estimated where the fraction of gray matter per given region-of-interest varied from O to 100%. In this simulation, the arterial input function was the standard input function used for the ARG method. 
RESULTS 
Figure 2a-2c show correlations between CBF values by TLU method and those by the ARG method in which the Vd Values were assumed to be 40, 50 and 60, respectively. When the Vd Value in the ARG method was assumed to be 50 ml/ml, the best correlation and good linearity between the TLU and ARG methods were observed (Y = 1.04X -2.5; X: TLU, Y: ARG, r = 0.97) (Fig. 2b) and these values for the two methods were consistent. However, when the Vd Value was assumed to be 40 or 60 ml/ml, the ARG method CBF was over- (Fig. 2a) and under- (Fig. 2c) estimated compared with the TLU method CBF, respectively, in particular in the hyperperfusion region. The mean CBF values obtained by the ARG method for all regions-of-interest data were 39.1 +- l0.3, 36.3 +- 8.6 and 34.8 +- 7.8 ml/100 ml/min (+- S.D.) when the Vd Values were assumed to be 40, 50 and 60 ml/ml, respectively. The mean CBF value by the TLU method was 37.4 +- 8.1 ml/ 1OO ml/min (+- S.D.). 
The mean Vd Value evaluated by the TLU method in X-ray CT normal density regions was 48.7 +- 9.15 ml/ml (+- S.D.). There was no significant difference between gray and white matter in Vd Values. 
A good correlation was obtained between CBF values evaluated by the ARG method in which the Vd Value was assumed to be 50 ml/ml and those by the MS method (Y= 0.82X+ 12.1; X: ARG, Y: MS, r =0.84) (Fig. 3). However, overestimation of CBF values was observed with the MS method as compared with the ARG method. Mean CBF values evaluated by the ARG method were 13.2% lower than those by the MS method (mean MS method CBF: 41.8 +- 8.5 ml/1OO ml/min (+- S.D.)). 
Figure 4 shows the effects of gray-white matter mixing for CBF values calculated by the ARG method. For each Vd value, CBF values calculated were systematically underestimated when the fraction of gray matter was varied from O to 100% (-7.6% for a gray matter fraction of 50% and a Vd of 50 ml/ml). 
DISCUSSION 
The best correlation between the TLU and ARG methods was observed (Fig. 2b) when the Vd Value in the ARG method was assumed to be 50 ml/ml which was almost the same as the mean Vd Value evaluated by the TLU method. These ARG method values also closely agreed with the MS method values (Fig. 3). However, this Vd Value was larger than those in previous studies, i.e., 22 in man,5 22 in rat23 and 29.6 in man8 (ml/ml). As reasons for this discrepancy, the differences in the reconstruction algo-rithm, the attenuation correction method and the scattered photons correction method for each SPECT scanner sys-tem were considered.lo These indicated that the Vd Value in the ARG method should be determined for each SPECT scanner system by using an other method, e.g., the TLU method, since the reconstruction algorithm, the attenua-tion correction method and the scattered photons correc-tion method were different for each SPECT scanner system. 
Vd Values in the normal brain tissue might be almost constant, but it was reported that Vd Values in the patho-logical reglons, e.g., cerebral infarction were lower than those in normal regions.14'19'20 In the present study, all regions-of-interest in the pathological regions, e.g., cere-bral contusion and cerebral infarction showed low Vd values and hypoperfusion. Since in hypoperfusion re-gions accurate CBF values could be calculated by the ARG method in any assumed Vd Values,lo a good correla-tion was observed between the ARG and TLU methods (Fig. 2b). However, if there were hyperperfusion lesions with low or high Vd Values, the error in measuring CBF by the ARG method would be significant. 
While the MS method was routine]y used as a method for measuring CBF by means of IMP, the underestimation of CBF was caused due to significant clearance of IMP from the brain.6~9,24 In this study, a good correlation was obtained between CBF values evaluated by the ARG method and those by the MS method (Fig. 3), suggesting equivalent applicability. However, CBF values obtained with the MS method were actually 13.2% higher than those obtained with the ARG method against previous studies.6~9,24 
One possibility is error in the effects of gray-white matter mixture. Simulation of the effects of gray-white matter mixture indicated differences between true CBF values and CBF values calculated by the ARG method. In this simulation, CBF values calculated were systemati-cally underestimated (-7.6% for a gray matter fraction of 50% and a Vd of 50 ml/ml) (Fig. 4). On the other hand, there were no effects of gray-white matter mlxing on CBF values calculated by the MS method, because the correlation between the brain radioactivity and CBF value is linear in the microsphere model (Eq. 5),22 
Another potential source of error is in estimation of the SPECT brain counts at 8 min in the MS method with count ratios of 8 min/40 min whole brain planar images and 40 min SPECT data.24~26 This error would be caused by non-linearity of 40 min SPECT counts due to a significant clearance of IMP which had been described in our re-port.24 An additional potential source of error is the determination of the arterial input function, i.e., correc-tions for time delay and dispersion of input for two methods27~29 and unknown errors due to differences in the arterial input curve shape for each subject for the ARG method.24 
In conclusion, the best correlation between the TLU and ARG methods was observed when the Vd Value in the ARG method was assumed to be 50 ml/ml which was a]most the same as the mean Vd Value evaluated by the TLU method. These ARG method values closely agreed with conventional MS method values, suggesting equiva-lent applicability. This indicates that the Vd Value in the ARG method should be determined properly for each SPECT scanner system by using an other method, e.g. , the TLU method. CBF values obtained by the MS method were actually 13.2% higher than those obtained by the ARG method against previous studies. As reasons for this, errors in the effects of gray-white matter mixture in the ARG method and in estimation of the SPECT brain counts at 8 min in the MS method were considered. 
ACKNOWLEDGMENTS 
We are greatly indebted to the staff of Sendai City Hospital and the Institute of Development, Aging and Cancer, Tohoku Uni-versity. 
This study was supported by a Grant-in-Aid for Scientific Research (05454297) from the Japanese Ministry of Education, Science and Culture. 
REFERENCES 
1. Winchel] H S, Baldwin RM , Lin TH. Development of I-123- labeled amines for brain studies: Iocalization of I-123 iodophenylalkyl amines in rat brain  J Nucl Med 21 : 940-946, 1980. 
2. Winche]1 HS, Horst WD, Braun L, OldendorfWH, Hattner R, Parker H. N-isopropyl-[123I] p-iodoamphetamine: single-pass brain uptake and washout; binding to brain synapto-somes; and localization in dog and monkey brain. J Nucl Med 21 : 947-952, 1980. 
3. Creutzig H, Schober O. Gielow P, Friedrich R, Becker H, Dietz H, et al. Cerebral dynamics of N-isopropyl-(1231)p-iodoam hetamine. JNucl Med 27: 178-183, 1986. 
4. Nishizawa S. Tanada S, Yonekura Y, Fujita T, Mukai T, Saji H, et al. Regional dynamics of N-isopropyl-(i231)p-iodoamphetamine in human brain. JNucl Med 30: 1 50= 1 56, 1989. 
5. Kuhl DE, Barrio JR, Huang SC, Selin C, Ackermann RF, Lear JL, et al. Quantifying local cerebral blood fiow by N-isopropyl-p-[123l]iodoamphetamine (IMP) tomography. J Nucl Med 23: 196-203, 1982. 
6. Greenberg JH. Kushner M, Rango M, Alavi A, Reivich M. Validation studies of iodine-123-iodoamphetamine as a cerebral blood flow tracer using emission tomography. J Nucl Med 31 : 1364-1369, 1990. 
7. Murase K, Tanada S, Mogami H, Kawamura M, Miyagawa M. Yamada M, et al. Validation of microsphere model in cerebral blood flow measurement using N-isopropyl-p-(1 231) iodoamphetamine. Med Phys 1 7: 79-83, 1 990. 
8. Yokoi T, Iida H, Itoh H, Kanno I, A new graphic plot analysis for cerebral blood flow and partition coefficient with iodine-123-Iodoamphetamine and dynamic SPECT validation studies using oxygen- 1 5-water and PET. J Nucl Med 34: 498-505, 1993. 
9. Yonekura Y, Nishizawa S, Mukai T, Iwasaki Y, Fukuyama H, Ishikawa M, et al. Functional mapping of flow and back-diffusion rate of N-isopropyl-p-iodoamphetamine in hu-man brain. J Nucl Med 34: 839-844, 1993. 
10. Iida H, Itoh H. Nakazawa M, Hatazawa J, Nishimura H, Onishi Y, et al. Quantitative mapping of regional cerebral blood flow using lodine- 1 23-IMP and SPECT. J Nucl Med 35: 2019-2030, 1994. 
ll. Iida H, Itoh H, Nakazawa M, Nishimura H. Onishi Y, Uemura K. Validation of quantitative mapping of rCBF using I- 1 23 IMP from a single SPECT scan with a standard in ut function. JNucl Med 35: 191P, 1994. 
12. Iida H, Itoh H, Bloomfield PM, Munaka M, Higano S, Murakami M, 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-l084, 1994. 
13. Iida H, Itoh H, MunakaM, Murakami M, Higano S, Uemura K: A clinical method to quantitate CBF using a rotating gamma camera and 1-123-amphetamine (IMP) with one blood sampling. KAKU lGAKU (J Nucl Med) 33: 963P, 1992. 
14. 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-i sopropyl-p-[ 123l]iodoamphet-amine (1231-IMP) SPECT; two scans with one point blood sampling technique. Jpn J Nucl Med 29: I 1 93-1 200, 1 992. 
15. Matsuda H, Seki H, Sumiya H, Tsuji S, Tonami N, Hisada K, et al. Quantitative cerebral blood flow measurements using N-isopropyl-(iodine 123)p-iodoamphetamine and single photon emission computed tomograph y with rotating gamma camera. Am J Physiol Imag I : 1 86-194, 1 986. 
l 6. Kouris K. Jarritt PH, Costa DC, Ell PJ. Physical assessment of the GE/CGR Neurocam and comparison with a single rotating gamma-camera. Eur J Nucl Med 19: 236-242, l 992. 
17. Budinger TF, Gullberg GT, Huesman RH. Image recon-struction from projections. Herman GT (ed.), New York, Springer-Verlag, p. 197, 1979. 
l 8. Sorenson JA. Quantitative measurement of radioactivity in vivo by whole-body counting. In Instrumentation in Nuclear 
Medicine, Hine GJ, Sorenson JA (eds.), New York, Aca-demic Press, . 31 1-348, 1974. 
19, Itoh H, Iida H, Bloomfield PM, Murakami M, Higano S, Munaka M, et al . A technique for rapid imaging of regional CBF and partition coefficient using dynamic SPECT and I-l 23-am hetamine (IMP). J Nucl Med 33: P91 1 , 1992. 
20. Ito H, Iida H, B]oomfield PM, Murakami M, Inugami A, Kanno l, et al. Rapid calculation of regional CBF and di stribution vol ume us i ng N-i sopropyl- p- [ 1 2~- I] iodoamphet-amine (i231-lMP) and d ynamic SPECT. JNuclMed 36: 53 1 -536, 1995. 
21 . Murase K, Tanada S, Inoue T, Ochi K, Fujita H, Sakaki S, et al. Measurement of the blood-brain barrier permeability of 1-123 IMP, Tc-99m HMPAO and Tc-99m ECD in the human brain using compartment model analysis and dy-namic SPECT. J Nucl Med 32: P91 1 , 199 1 . 
22. Huang SC, Mahoney DK, Phelps ME. Quantitation in positron emission tomography: 8. Effects of nonlinear pa-rameter estimation on functional images. J Comput Assist Tomo r I l: 314-325, 1987. 
23. Lear JL. Ackermann RF, Kameyama M, Kuhl DE. Evalu-ation of [123l]isopropyliodoamphetamine as a tracer for Iocal cerebral blood now using direct autoradiographic comparison. J Cereb Blood Flow Metab 2: 1 79-1 85, 1 982. 
24, Ito H, Ishii K, Atsumi H, Kinoshita T, Kawashima R, Ono S, et al . Error analysis of Table look-up method for cerebral blood now measurement by 12~. 1-lMP brain SPECT: Com-parison with conventional microsphere model method. Ann Nucl Med 9: 75-80, 1 995. 
25. Takahashi N, Ohkubo M, Odano I, Sakai K. A problem of quantitative measurement of regional cerebra] blood flow using microsphere model and N-isopropyl-p-[123l]iodoam-phetamine (IMP): Comparison with 1~_ 3Xe SPECT and se-quential dynamic 1231-lMP SPECT. KAKU lGAKU (Jpn J Nucl Med) 3 1 : 3 1 9-326, 1 994. 
26. Takahashi N. Quantifying regional cerebral blood flow with N-isopropyl-p[1231]iodoamphetamine by ring-type single-photon emi ssion computed tomography: Val idit y of a method to esti mate early reference value by means of regional brai n time-activity curve. Ann Nucl Med 8: 253-258, 1 994. 
27. Iida H, Kanno l, Miura S, Murakami M, Takahashi K, Uemura K. Error ana]ysis of a quantitative cerebral blood flow measurement using H 2150 autoradiography and positron emission tomography, with respect to the dispersion of the input function. J Cereb Blood Flow Metab 6: 536-545, l 986. 
28. Iida H, Higano S, Tomura N, Shishido F, Kanno l, Miura S, et al. Evaluation of regional differences of tracer appear-ance time in cerebral tissues using [150]water and dynamic positron emi ssion tomography. J Cereb Blood Flow Metab 8: 285-288, 1988. 
29. Iida H, Kanno I, Miura S, Murakami M. Takahashi K, Uemura K. A determination of the regional brain/blood partition coefficient of water using dynamic positron emis-sion tomography. J Cereb Blood Flow Metab 9: 87l~885, 1989.