- Email: [email protected]
Linear polymeric agents for probing tumor vasculature
Abstracts of the AMI Annual Meeting 2004 101
tion of 26% and a coincidence timing window of 10 ns. A 0.5mm pointsource was placed at 29 transverse locations and radial and tangential resolutions were computed for filtered-back Projection (FBP) and LeastSquares by QR decomposition (LSQR). In the case of FBP, the data was histogrammed into sinograms of 24 angular and 48 linear samples. No histogramming was needed for LSQR. The results indicate that the use of the system response in iterative reconstruction produces improved and more uniform resolution across the field-of-view than with FBP. On average, radial and tangential resolutions are 1.57mm and 1.81mm, for LSQR, and 2.47mm and 2.51 mm for FBP, respectively. No. 121 IMAGING BETA-AMYLOID PLAQUES IN SILICO K. Shoghi-Jadid1, S. Huang1, V. Kepe2, J. R. Barrio2 1UCLA Department of Biomathematics, Los Angeles, CA, 2UCLA Department of Pharmacology, Los Angeles, CA. Alzheimer’s disease (AD) is defined histologically by the formation of neurofibrillary tangles (NFTs) and extracellular β-amyloid fibril (fAβ) plaques comprised of monomeric β-amyloids (mAβ). Recently, PET in conjunction with molecular probes targeting fAβ have been used to image fAβ density in the living brains of AD patients. We present a mathematical model for the formation of fAβ plaques and use this model to address challenges in imaging fAβ load. In the simulation, we setup a 30x30 lattice mimicking a segment of brain tissue. In each lattice site, we consider the polymerization, diffusion, and degradation of fAβs and mAβs. Polymerization takes place once the concentration of mAβ surpasses a critical threshold. Both fAβ and mAβ diffuse throughout the lattice with established diffusion constants. The degradation of fAβ and mAβ follows Michaelis-Menten type kinetics. As the concentration of fAβ at a given site reaches an assumed maximal concentration, fAβ are immobilized to mimic the action of astrocytes. At the boundaries of the lattice, mAβs are secreted at varying rates to reflect differences in neuronal mAβ production. Various degrees of pathology are simulated by varying model parameters. Finally, we use the typical three-compartment receptor model to characterize fAβ load in the lattice with hypothetical imaging probes. The ability to image and discriminate AD pathology is largely a function of the probe’s ability to penetrate densely packed plaques (e.g. senile versus diffuse plaques). We present these and other findings in the context of the molecular requirements of the ideal fAβ imaging probe. Our findings may also be applicable to NFT formations and underscore the notion that the ideal imaging probe should target structural components of fibrils common to both fAβs and NFTs. No. 122 BIOLOGICAL CORRELATES OF 2-DEOXY-2-[18F]FLUORO-DGLUCOSE UPTAKE IN ORAL SQUAMOUS CELL CARCINOMA MEASURED BY POSITRON EMISSION TOMOGRAPHY M. Tian1, H. Zhang2, Y. Nakasone1, K. Mogi1, K. Endo1 1Gunma University School of Medicine, Gunma, JAPAN, 2National Institute of Radiological Sciences, Chiba, JAPAN. Background: Increased expressions of glucose transporters have been reported in several human cancers. The mechanism of glucose entry into oral squamous cell carcinoma (OSCC) remains unclear. Increased but variable uptake of the glucose analog 2-deoxy-2-[18F]fluoro-D-glucose (FDG) has been noticed in positron emission tomography (PET) studies of in human OSCC patients. This initial study compared FDG uptake in vivo with biomarkers expected to be involved in the underlying biologic mechanisms. Materials and Methods: Preoperative FDG-PET scans were performed in 31 patients. FDG activity was assessed visually by two observers using a four-point score. Tumor sections were stained by immunohistochemistry for glucose transporter-1 (Glut-1); Hexokinase (HK) II; hypoxia-inducible factor-1-alfa (HIF-1alpha); vascular endothe-
lial growth factor (VEGF), P53 and Ki-67. Mitotic activity index (MAI) and amount of necrosis were assessed. Results: There were positive correlations between FDG uptake and Glut-1 expression (P< 0.01), MAI (P< 0.04), Ki-67 (P< 0.05), P53 (P< 0.02), amount of necrosis (P< 0.03), number of tumor cells/volume (P< 0.01), expression of microvessel density (P =0.05). HIF-1alpha, VEGF, HK II showed no univariate correlation with FDG. In logistic regression, however, HIF-1alpha and HK II added value to Glut-1. Conclusion: FDG uptake in OSCC is a function of microvasculature for delivering nutrients, Glut-1 for transportation of FDG into the cell, HK for entering FDG into glycolysis, number of tumor cells/volume, proliferation rate (also reflected in necrosis) and HIF-1alpha for upregulating Glut-1. Together, these features explain why OSCC vary in FDG uptake and elucidate the low uptake in some cases. No. 123 PRELIMINARY RESULTS FROM A COMPACT MINI GAMMA CAMERA FOR IN VIVO MOLECULAR IMAGING T. Tumer1, A. Chatziioannou2, S. Gambhir3, S. Yin1 1NOVA R&D, Inc., Riverside, CA, 2University of California, Crump Institute for Molecular Imaging, Los Angeles, CA, 3Stanford University, Stanford, CA. We are developing a highly sensitive, direct conversion detector system for studies of in vivo molecular imaging in small animals. This detector system can be applied to either gamma cameras, single photon emission computed tomography (SPECT) or positron emission tomography (PET) systems. Currently, most gamma cameras, SPECT or PET systems employ scintillator-based detectors, where the gamma rays are converted to visible lights, and then into charge signals. These detectors, compared to the direct conversion detectors we propose, are less efficient. They have higher intrinsic noise, poor energy resolution and poor image qualities. The direct conversion detectors can directly convert the incident gamma rays into charge signals with high efficiency and better energy resolution. Also due to the direct conversion of the gamma ray photon into electron-hole pairs will allow high spatial resolution and at the same time high quantum efficiency. This new technology is expected to lead to molecular imaging systems with higher efficiency, high spatial resolution and better performance compared to the current scintillator based systems. We will report preliminary results obtained from a first prototype detector system. No. 124 LINEAR POLYMERIC AGENTS FOR PROBING TUMOR VASCULATURE E. Uzgiris, L. Brogan, B. Grimmond, J. F. Smith GE Global Research Center, Niskayuna, NY. Linear polymeric contrast agents were constructed as possible carriers of specific ligands for magnetic resonance imaging (MRI) detection of specific binding events in tissue. Such constructs may be useful in this regard since a large number of Gd ions can be associated with the polymers depending on the polymerization number of the backbone polylysine peptide. In addition, these agents may allow better determination of heomodynamic parameters of tissue vasculature than with other agents currently being investigated. This is of particular importance in assessing tumor angiogenesis for example. For this purpose, macromolecular agents have been used in dynamic signal acquisition protocols. In particular, Gd-albumin agents have been used to show that the permeability and vascular volume fraction in aggressive and non-aggressive tumor models, and benign vs malignant tumors differ. Of the two parameters studied, the permeability differences are the more significant between the differing tumor types. We hypothesized that a repeating polymeric agent, that is a linear, uncoiled, unbranched polymer with a small cross section relative to length will give perhaps a higher dynamic range of permeability values between different tumor types than the albumin agents used previously. This
102 Molecular Imaging and Biology, Volume 6, Number 2
hypothesis is based on a different mechanism of trans-endothelial transport for the polymeric agents. We have confirmed this hypothesis with an MCF7 human breast cancer tumor model and with a human prostate cancer tumor model. Higher permeability values were observed than with the globular albumin agents and we were able to clearly distinguish between slow and fast growing tumors.
Conclusion: One- or two-compartment model gives slightly different results, confirming that CBF values measured using either model do not represent the same quantities. In fact, two-compartment model gives an extracted water washout rate from the brain tissue (excluding vascular space), which will be less dependent on the vascular distribution. No. 127
No. 125 PRELIMINARY RESULTS OF THE SMALL ANIMAL ROTATIONAL POSITRON EMISSION TOMOGRAPHY SCANNER J. Vaquero1, A. Molins1, J. Ortuno2, J. Pascau1, M. Desco1 1Hospital GU Gregorio Maranon, Madrid, SPAIN, 2Universidad Politécnica de Madrid, Madrid, SPAIN. This contribution reports preliminary results of a high-resolution small animal positron emission tomography (PET) based on pairs of opposed scintillation rotating detectors working in time coincidence. Each detector is comprised of a 35 x 30 array of LYSO crystals coupled to a position-sensitive photomultiplier tube; the assembly is self-shielded. Detectors separation can be adjusted between 150 and 200 mm. The data acquisition computer controls the gantry rotation as well as the bed displacements. Tomographic images are reconstructed from lines of response falling within a user-specified angle of each detector normal to limit the depth of interaction effect. These reconstructions are done with filtered back projection and statistical iterative methods that includes a model for the depth of interaction to minimize the parallax error. The system field-of-view is 48 mm x 48 mm, with a resolution of 1.7 mm FWHM (FBP), 56 slices 0.7 mm thick, and a sensitivity of 350 cps/uCi for a 250-650 keV energy window. Dynamic resolution for tomographic imaging is limited to the minimum rotation time, although planar imaging projection can be done with maximum time resolution. This low cost design has probed to be a valid solution when high performance tomography and temporal resolution requirements are not simultaneously required.
OXIDATIVE METABOLIC RATE MEASUREMENT OF THE RAT BRAIN USING O15O MICROPET IMAGING S. Yee, P. Jerabek, P. T. Fox University of Texas Health Science Center, San Antonio, TX. Introduction: Cerebral metabolic rate of oxygen (CMRO2) is an important factor to study neurological disorders in terms of metabolic process. CMRO2 measurement technique without arterial blood sampling for rat using microPET is being proposed and applied to measure CMRO2 at two different body temperature levels. Methods: Animal Preparation: Eight rats (Fischer 344, 262 ± 17 g) have been anesthetized for surgical preparation using isofluorane. Tracheotomy was performed to inject O-15 labeled oxygen directly into the lung. Left femoral vein has been catheterized for administering anesthetic during imaging procedure (alpha-chloralose, 40 mg/kg i.v. initially, 30 mg/kg/hr i.v. infusion for maintenance). Water flowing thermal jacket was used to change and maintain body temperature. Cerebral Blood Flow Measurement: One-step method with three-weighted integral was used to measure CMRO2 of the whole brain. O15O (5 ml) was instantaneously injected through trachea into the lung using 10 ml syringe. Arterial input function was obtained from time activity curve of the heart without arterial blood sampling. Data acquisition was done, first at low body temperature (32 °C) and later at 37 °C, in list mode for three minutes after O15O injection, using microPET R4(Concorde Microsystems). Results: CMRO2 (ml/100g/min) measured for the whole brain was 4.98 ± 0.39 for 37 °C, whereas it was 3.52 ± 0.46 for 32 °C. For all rats, hypothermia gives about 30% less CMRO2 values compared to 37 °C. Conclusion: CMRO2 measurement technique for small animals using O15O microPET imaging was successfully demonstrated for the first time.
No. 126 No. 128 COMPARISON OF ONE- AND TWO-COMPARTMENT MODELS IN DETERMINING CEREBRAL BLOOD FLOW OF THE RAT BRAIN USING MICROPET S. Yee, P. Jerabek, P. T. Fox University of Texas Health Science Center, San Antonio, TX. Introduction: For measuring cerebral blood flow (CBF) using microPET, either one-compartment model which does not consider separate vascular space in the brain tissue or two-compartment model which does discriminate vascular space from extra-vascular space can be applied. We report here how both models are different based on the same data set obtained for rats using microPET. Methods: Animal Preparation: Eight rats (Fischer 344, 262 ± 17 g) have been anesthetized for surgical preparation using isofluorane. Two femoral veins have been catheterized for injecting labeled water and for administering anesthetic during imaging procedure (alpha-chloralose, 40 mg/kg i.v. initially, 30 mg/kg/hr i.v. infusion for maintenance). CBF measurement: One- and two-compartment models were used to measure CBF of the whole brain. H215O was injected intravenously through femoral vein. Arterial input function was obtained from time activity curve of the heart without arterial blood sampling. Data acquisition was done in list mode for three minutes after H215O injection, using microPET R4(Concorde Microsystems). Results: CBF (ml/100g/min) measured for the whole brain was 55.1 ± 5.59 for one-compartment model whereas it was 52.2 ± 5.09 for two-compartment model. For all rats, two-compartment model gives about 5% less CBF values than one-compartment model.
THE CLEARPET™ NEURO SCANNER: A DEDICATED HIGH PERFORMANCE LSO/LUYAP PHOSWICH SMALL ANIMAL PET SCANNER K. Ziemons, A. Bauer, M. Holschbach, U. Pietrzyk, M. Streun, H. Coenen, H. Halling, K. Zilles Forschungszentrum Juelich GmbH, Juelich, GERMANY. The PrimatePET project is proposed by working groups of the Research Center Jülich. The aim of this project is to apply the non-invasive positron emission tomography (PET) technique to in vivo investigations of signal transduction in non-human primates under physiological conditions. While in recently developed dedicated small animal PET systems a high spatial resolution of about ~1.5mm was the main research interest, it has become clear that it is equally important not to sacrifice the sensitivity of the scanners since the specific activity of the radiotracers used may be limited. In connection with the Crystal Clear Collaboration (CCC) we developed a dedicated second generation high performance PET scanner, called ClearPET™ Neuro. High sensitivity and high spatial resolution for the ClearPET camera is achieved by using a phoswich arrangement combining two types of lutetium-based scintillator materials: LSO and LuYAP:Ce. The dual layer created by 8x8 crystals of 2x2x10mm3 are coupled to multichannel photomultiplier tubes (PMT). A unit of four PMTs arranged inline represents one of 20 sectors of the ring design. The opening diameter of the ring can be varied between 130mm and 300mm, the axial detector length is 110mm. The PMT pulses are digitized by free-running ADCs which allows the determination of the gamma energy, the depth of interac-
tion of 26% and a coincidence timing window of 10 ns. A 0.5mm pointsource was placed at 29 transverse locations and radial and tangential resolutions were computed for filtered-back Projection (FBP) and LeastSquares by QR decomposition (LSQR). In the case of FBP, the data was histogrammed into sinograms of 24 angular and 48 linear samples. No histogramming was needed for LSQR. The results indicate that the use of the system response in iterative reconstruction produces improved and more uniform resolution across the field-of-view than with FBP. On average, radial and tangential resolutions are 1.57mm and 1.81mm, for LSQR, and 2.47mm and 2.51 mm for FBP, respectively. No. 121 IMAGING BETA-AMYLOID PLAQUES IN SILICO K. Shoghi-Jadid1, S. Huang1, V. Kepe2, J. R. Barrio2 1UCLA Department of Biomathematics, Los Angeles, CA, 2UCLA Department of Pharmacology, Los Angeles, CA. Alzheimer’s disease (AD) is defined histologically by the formation of neurofibrillary tangles (NFTs) and extracellular β-amyloid fibril (fAβ) plaques comprised of monomeric β-amyloids (mAβ). Recently, PET in conjunction with molecular probes targeting fAβ have been used to image fAβ density in the living brains of AD patients. We present a mathematical model for the formation of fAβ plaques and use this model to address challenges in imaging fAβ load. In the simulation, we setup a 30x30 lattice mimicking a segment of brain tissue. In each lattice site, we consider the polymerization, diffusion, and degradation of fAβs and mAβs. Polymerization takes place once the concentration of mAβ surpasses a critical threshold. Both fAβ and mAβ diffuse throughout the lattice with established diffusion constants. The degradation of fAβ and mAβ follows Michaelis-Menten type kinetics. As the concentration of fAβ at a given site reaches an assumed maximal concentration, fAβ are immobilized to mimic the action of astrocytes. At the boundaries of the lattice, mAβs are secreted at varying rates to reflect differences in neuronal mAβ production. Various degrees of pathology are simulated by varying model parameters. Finally, we use the typical three-compartment receptor model to characterize fAβ load in the lattice with hypothetical imaging probes. The ability to image and discriminate AD pathology is largely a function of the probe’s ability to penetrate densely packed plaques (e.g. senile versus diffuse plaques). We present these and other findings in the context of the molecular requirements of the ideal fAβ imaging probe. Our findings may also be applicable to NFT formations and underscore the notion that the ideal imaging probe should target structural components of fibrils common to both fAβs and NFTs. No. 122 BIOLOGICAL CORRELATES OF 2-DEOXY-2-[18F]FLUORO-DGLUCOSE UPTAKE IN ORAL SQUAMOUS CELL CARCINOMA MEASURED BY POSITRON EMISSION TOMOGRAPHY M. Tian1, H. Zhang2, Y. Nakasone1, K. Mogi1, K. Endo1 1Gunma University School of Medicine, Gunma, JAPAN, 2National Institute of Radiological Sciences, Chiba, JAPAN. Background: Increased expressions of glucose transporters have been reported in several human cancers. The mechanism of glucose entry into oral squamous cell carcinoma (OSCC) remains unclear. Increased but variable uptake of the glucose analog 2-deoxy-2-[18F]fluoro-D-glucose (FDG) has been noticed in positron emission tomography (PET) studies of in human OSCC patients. This initial study compared FDG uptake in vivo with biomarkers expected to be involved in the underlying biologic mechanisms. Materials and Methods: Preoperative FDG-PET scans were performed in 31 patients. FDG activity was assessed visually by two observers using a four-point score. Tumor sections were stained by immunohistochemistry for glucose transporter-1 (Glut-1); Hexokinase (HK) II; hypoxia-inducible factor-1-alfa (HIF-1alpha); vascular endothe-
lial growth factor (VEGF), P53 and Ki-67. Mitotic activity index (MAI) and amount of necrosis were assessed. Results: There were positive correlations between FDG uptake and Glut-1 expression (P< 0.01), MAI (P< 0.04), Ki-67 (P< 0.05), P53 (P< 0.02), amount of necrosis (P< 0.03), number of tumor cells/volume (P< 0.01), expression of microvessel density (P =0.05). HIF-1alpha, VEGF, HK II showed no univariate correlation with FDG. In logistic regression, however, HIF-1alpha and HK II added value to Glut-1. Conclusion: FDG uptake in OSCC is a function of microvasculature for delivering nutrients, Glut-1 for transportation of FDG into the cell, HK for entering FDG into glycolysis, number of tumor cells/volume, proliferation rate (also reflected in necrosis) and HIF-1alpha for upregulating Glut-1. Together, these features explain why OSCC vary in FDG uptake and elucidate the low uptake in some cases. No. 123 PRELIMINARY RESULTS FROM A COMPACT MINI GAMMA CAMERA FOR IN VIVO MOLECULAR IMAGING T. Tumer1, A. Chatziioannou2, S. Gambhir3, S. Yin1 1NOVA R&D, Inc., Riverside, CA, 2University of California, Crump Institute for Molecular Imaging, Los Angeles, CA, 3Stanford University, Stanford, CA. We are developing a highly sensitive, direct conversion detector system for studies of in vivo molecular imaging in small animals. This detector system can be applied to either gamma cameras, single photon emission computed tomography (SPECT) or positron emission tomography (PET) systems. Currently, most gamma cameras, SPECT or PET systems employ scintillator-based detectors, where the gamma rays are converted to visible lights, and then into charge signals. These detectors, compared to the direct conversion detectors we propose, are less efficient. They have higher intrinsic noise, poor energy resolution and poor image qualities. The direct conversion detectors can directly convert the incident gamma rays into charge signals with high efficiency and better energy resolution. Also due to the direct conversion of the gamma ray photon into electron-hole pairs will allow high spatial resolution and at the same time high quantum efficiency. This new technology is expected to lead to molecular imaging systems with higher efficiency, high spatial resolution and better performance compared to the current scintillator based systems. We will report preliminary results obtained from a first prototype detector system. No. 124 LINEAR POLYMERIC AGENTS FOR PROBING TUMOR VASCULATURE E. Uzgiris, L. Brogan, B. Grimmond, J. F. Smith GE Global Research Center, Niskayuna, NY. Linear polymeric contrast agents were constructed as possible carriers of specific ligands for magnetic resonance imaging (MRI) detection of specific binding events in tissue. Such constructs may be useful in this regard since a large number of Gd ions can be associated with the polymers depending on the polymerization number of the backbone polylysine peptide. In addition, these agents may allow better determination of heomodynamic parameters of tissue vasculature than with other agents currently being investigated. This is of particular importance in assessing tumor angiogenesis for example. For this purpose, macromolecular agents have been used in dynamic signal acquisition protocols. In particular, Gd-albumin agents have been used to show that the permeability and vascular volume fraction in aggressive and non-aggressive tumor models, and benign vs malignant tumors differ. Of the two parameters studied, the permeability differences are the more significant between the differing tumor types. We hypothesized that a repeating polymeric agent, that is a linear, uncoiled, unbranched polymer with a small cross section relative to length will give perhaps a higher dynamic range of permeability values between different tumor types than the albumin agents used previously. This
102 Molecular Imaging and Biology, Volume 6, Number 2
hypothesis is based on a different mechanism of trans-endothelial transport for the polymeric agents. We have confirmed this hypothesis with an MCF7 human breast cancer tumor model and with a human prostate cancer tumor model. Higher permeability values were observed than with the globular albumin agents and we were able to clearly distinguish between slow and fast growing tumors.
Conclusion: One- or two-compartment model gives slightly different results, confirming that CBF values measured using either model do not represent the same quantities. In fact, two-compartment model gives an extracted water washout rate from the brain tissue (excluding vascular space), which will be less dependent on the vascular distribution. No. 127
No. 125 PRELIMINARY RESULTS OF THE SMALL ANIMAL ROTATIONAL POSITRON EMISSION TOMOGRAPHY SCANNER J. Vaquero1, A. Molins1, J. Ortuno2, J. Pascau1, M. Desco1 1Hospital GU Gregorio Maranon, Madrid, SPAIN, 2Universidad Politécnica de Madrid, Madrid, SPAIN. This contribution reports preliminary results of a high-resolution small animal positron emission tomography (PET) based on pairs of opposed scintillation rotating detectors working in time coincidence. Each detector is comprised of a 35 x 30 array of LYSO crystals coupled to a position-sensitive photomultiplier tube; the assembly is self-shielded. Detectors separation can be adjusted between 150 and 200 mm. The data acquisition computer controls the gantry rotation as well as the bed displacements. Tomographic images are reconstructed from lines of response falling within a user-specified angle of each detector normal to limit the depth of interaction effect. These reconstructions are done with filtered back projection and statistical iterative methods that includes a model for the depth of interaction to minimize the parallax error. The system field-of-view is 48 mm x 48 mm, with a resolution of 1.7 mm FWHM (FBP), 56 slices 0.7 mm thick, and a sensitivity of 350 cps/uCi for a 250-650 keV energy window. Dynamic resolution for tomographic imaging is limited to the minimum rotation time, although planar imaging projection can be done with maximum time resolution. This low cost design has probed to be a valid solution when high performance tomography and temporal resolution requirements are not simultaneously required.
OXIDATIVE METABOLIC RATE MEASUREMENT OF THE RAT BRAIN USING O15O MICROPET IMAGING S. Yee, P. Jerabek, P. T. Fox University of Texas Health Science Center, San Antonio, TX. Introduction: Cerebral metabolic rate of oxygen (CMRO2) is an important factor to study neurological disorders in terms of metabolic process. CMRO2 measurement technique without arterial blood sampling for rat using microPET is being proposed and applied to measure CMRO2 at two different body temperature levels. Methods: Animal Preparation: Eight rats (Fischer 344, 262 ± 17 g) have been anesthetized for surgical preparation using isofluorane. Tracheotomy was performed to inject O-15 labeled oxygen directly into the lung. Left femoral vein has been catheterized for administering anesthetic during imaging procedure (alpha-chloralose, 40 mg/kg i.v. initially, 30 mg/kg/hr i.v. infusion for maintenance). Water flowing thermal jacket was used to change and maintain body temperature. Cerebral Blood Flow Measurement: One-step method with three-weighted integral was used to measure CMRO2 of the whole brain. O15O (5 ml) was instantaneously injected through trachea into the lung using 10 ml syringe. Arterial input function was obtained from time activity curve of the heart without arterial blood sampling. Data acquisition was done, first at low body temperature (32 °C) and later at 37 °C, in list mode for three minutes after O15O injection, using microPET R4(Concorde Microsystems). Results: CMRO2 (ml/100g/min) measured for the whole brain was 4.98 ± 0.39 for 37 °C, whereas it was 3.52 ± 0.46 for 32 °C. For all rats, hypothermia gives about 30% less CMRO2 values compared to 37 °C. Conclusion: CMRO2 measurement technique for small animals using O15O microPET imaging was successfully demonstrated for the first time.
No. 126 No. 128 COMPARISON OF ONE- AND TWO-COMPARTMENT MODELS IN DETERMINING CEREBRAL BLOOD FLOW OF THE RAT BRAIN USING MICROPET S. Yee, P. Jerabek, P. T. Fox University of Texas Health Science Center, San Antonio, TX. Introduction: For measuring cerebral blood flow (CBF) using microPET, either one-compartment model which does not consider separate vascular space in the brain tissue or two-compartment model which does discriminate vascular space from extra-vascular space can be applied. We report here how both models are different based on the same data set obtained for rats using microPET. Methods: Animal Preparation: Eight rats (Fischer 344, 262 ± 17 g) have been anesthetized for surgical preparation using isofluorane. Two femoral veins have been catheterized for injecting labeled water and for administering anesthetic during imaging procedure (alpha-chloralose, 40 mg/kg i.v. initially, 30 mg/kg/hr i.v. infusion for maintenance). CBF measurement: One- and two-compartment models were used to measure CBF of the whole brain. H215O was injected intravenously through femoral vein. Arterial input function was obtained from time activity curve of the heart without arterial blood sampling. Data acquisition was done in list mode for three minutes after H215O injection, using microPET R4(Concorde Microsystems). Results: CBF (ml/100g/min) measured for the whole brain was 55.1 ± 5.59 for one-compartment model whereas it was 52.2 ± 5.09 for two-compartment model. For all rats, two-compartment model gives about 5% less CBF values than one-compartment model.
THE CLEARPET™ NEURO SCANNER: A DEDICATED HIGH PERFORMANCE LSO/LUYAP PHOSWICH SMALL ANIMAL PET SCANNER K. Ziemons, A. Bauer, M. Holschbach, U. Pietrzyk, M. Streun, H. Coenen, H. Halling, K. Zilles Forschungszentrum Juelich GmbH, Juelich, GERMANY. The PrimatePET project is proposed by working groups of the Research Center Jülich. The aim of this project is to apply the non-invasive positron emission tomography (PET) technique to in vivo investigations of signal transduction in non-human primates under physiological conditions. While in recently developed dedicated small animal PET systems a high spatial resolution of about ~1.5mm was the main research interest, it has become clear that it is equally important not to sacrifice the sensitivity of the scanners since the specific activity of the radiotracers used may be limited. In connection with the Crystal Clear Collaboration (CCC) we developed a dedicated second generation high performance PET scanner, called ClearPET™ Neuro. High sensitivity and high spatial resolution for the ClearPET camera is achieved by using a phoswich arrangement combining two types of lutetium-based scintillator materials: LSO and LuYAP:Ce. The dual layer created by 8x8 crystals of 2x2x10mm3 are coupled to multichannel photomultiplier tubes (PMT). A unit of four PMTs arranged inline represents one of 20 sectors of the ring design. The opening diameter of the ring can be varied between 130mm and 300mm, the axial detector length is 110mm. The PMT pulses are digitized by free-running ADCs which allows the determination of the gamma energy, the depth of interac-