Abstract
241323
Introduction: O-15 gas PET enables quantitative measurement of cerebral blood flow and metabolism, making it valuable for diagnosis of cerebrovascular disorders and assessment of treatment effects. Additionally, it is a unique examination method that can image cerebral blood flow (CBF), cerebral blood volume (CBV), oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen (CMRO2) in a series of scan. Autoradiography method (Shidahara M et al 2002, Watabe H et al 2004). has the advantages over the steady-state method, such as high quantitative capability due to insensitivity to arterial activity variation, short scan time, and low radiation exposure. However, it requires continuous arterial blood activity measured with a dedicated device, as well as cross-calibrations among the PET scanner, well-type gamma counter, and the blood activity detector. Meanwhile, due to very high sensitivity featured in recent 3D-PET scanners, inhaled radioactivity is limited to ensure deadtime and scatter correction, which may deteriorate count statistics in the measured arterial input function. The aim of this study was to investigate the influence of the error of the cross-calibration factor for the continuous blood sampling system (CCFcbs) and statistical noise of the input function (aTACnoise) on the quantitative values of CBF and oxygen metabolism.
Methods: The devices used in this study included a 3D-PET scanner Discovery IQ-5 (GE), continuous blood sampling system with a Phoswich detector (Yamamoto S et al 2001) and a peristaltic pump, well-type gamma counter ARC-370M (ALOKA), dose calibrator CRC-15R (Capintec), and blood gas analyzer GASTAT-navi (Techno Medica). The subjects were 10 patients with cerebrovascular disorders and inhaled 15O-CO2 and 15O-O2 gas sequentially, each for 1.5 minutes while undergoing dynamic acquisition for 3.5 minutes from the start of inhalation. The delivered 15O-gas radioactivity to the inhalation mask was set at 0.8 – 1.0 GBq/min, based on the PET scanner count rate and phantom experiments for reliable deadtime and scatter correction. The continuous blood radioactivity measurements were taken during the dynamic acquisition of 15O-CO2 and 15O-O2 gases. Additionally, static acquisition for 15O-CO gas was performed for 4 minutes, starting 3 minutes after the end of inhalation, while blood was sampled manually three times during the PET acquisition. The intervals between the PET acquisition of 15O-CO2 gas, 15O-O2 gas, and 15O-CO gas were approximately 10 minutes. The image reconstruction was performed using OSEM algorithm (5i12s) with smoothing by a 4 mm FWHM Gaussian filter. As the simulated error analysis, first, CCFcbs was varied from -10% to +10%. Second, various levels of noise (aTACnoise) was added to the input function. The quantitative analysis used GE healthcare Xeleris (ver. 3.0514) and an application program O15Gas CBF X3 (GE). The different aTACs with varying statistical noises were simulated using the C programming language.
Results: The error in CCFcbs inversely affected CBF and CMRO2, while OEF was not influenced. The aTACnoise had a significant effect, more on OEF than on CBF, with a tendency of overestimating OEF and underestimating of CBF. Based on the estimated Poisson noise in the arterial blood activity curve under the present device and measurement condition, + 7% and - 17% maximum bias was expected for OEF and CBF, respectively.
Conclusions: When a currently available high-sensitivity 3D-PET scanners is used, quantitative capability of 15O-gas PET depends not only on the correction of deadtime and scatter of the PET scanner itself but also on the cross-calibration with the blood activity measurement system. Furthermore, statistical noise in the continuous arterial blood activity measurement may induce a substantially large bias in the quantified values of OEF and CBF. Therefore, the radioactivity and duration of inhaled 15O-gases should be determined carefully based on the estimated errors in the measurement system.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.