PT - JOURNAL ARTICLE AU - Dasari, Paul AU - Jones, Judson AU - Casey, Michael AU - Dilsizian, Vasken AU - Smith, Mark TI - The Effects of Time-of-Flight and Point Spread Function Modeling on Regional Myocardial Blood Flow for Dynamic <sup>82</sup>Rb PET of Large Patients DP - 2017 May 01 TA - Journal of Nuclear Medicine PG - 1306--1306 VI - 58 IP - supplement 1 4099 - http://jnm.snmjournals.org/content/58/supplement_1/1306.short 4100 - http://jnm.snmjournals.org/content/58/supplement_1/1306.full SO - J Nucl Med2017 May 01; 58 AB - 1306Objectives: This study aims to investigate the effect of TOF and PSF modeling on myocardial blood flow (MBF) and myocardial flow reserve (MFR) values for large patients in each of the 3 main vascular territories (LAD, LCX and RCA) and in the AHA 17 segment model.Methods: 36 patients with BMI greater than 30 underwent rest/stress 82Rb cardiac PET (9 male, 27 female; mean weight ± SD, 140.6 ± 33.0 Kg; mean BMI ± SD, 47.7 ± 9.4). Patients were imaged using a TOF PET-CT system (Siemens Biograph mCT). A 3D listmode PET acquisition was started at the same time as the 82Rb infusion. The PET data were histogrammed into 26 dynamic frames (12 x 5 sec, 6 x 10 sec, 4 x 20 sec, 4 x 40 sec) and reconstructed using six methods: (1) Filtered Back Projection (FBP), (2) FBP with TOF (FBPTOF), (3) ordered-subsets expectation-maximization (OSEM) algorithm (OP), (4) OSEM with TOF (OPTOF), (5) OSEM with PSF (PSF), (6) OSEM with PSF and TOF (PSFTOF). OSEM algorithms used 3 iterations 8 subsets. A 3D Gaussian post-reconstruction smoothing filter with FWHM of 8 mm was applied. The MBF and MFR estimates were quantified using a 1-tissue compartmental model for 82Rb. The MBF (at stress and rest) and MFR values were averaged and evaluated within each of the 3 main vascular territories (LAD, LCX and RCA) and AHA-17 segments for each reconstruction method. For MBF and MFR data, ratios between the reconstructions incorporating TOF and PSF with their non-TOF and non-PSF counterparts were calculated and compared.Results: Results by vascular territories: Compared to non-TOF, flow values increased with TOF in all the vascular territories for rest and stress; the highest increase was for LAD with 21 to 30%, followed by RCA with 10 to 22% and LCX with 6 to 16%. With TOF, the MFR values in the three vascular territories showed little change, ranging between -3 to 7%. Compared to non-PSF, PSF modeling scarcely increased the flows values for rest and stress; with highest in the LAD by 4% followed by LCX and RCA by 2%. With PSF modelling, the change in MFR values for the three vascular territories was also small ranging between -3 to 1%. Results by AHA 17-segments: Compared to non-TOF, for both rest and stress, flow values increased with TOF in the apex and septal wall segments by 25 to 45% and in the inferior and lateral wall segments by 1 to 11%. With TOF, the MFR values increased in the apex and septal wall segments by 2 to 12% and decreased in the inferior and lateral wall segments by 4 to 7%. Compared to non-PSF, PSF modeling scarcely affected the flow values in all the segments for rest and stress; the change in flow values ranged within only 3% in all the segments. With PSF modelling, the change in MFR values across all the segments was also small ranging between -3 to 2%.Conclusion: TOF and PSF modeling influenced the quantitative values of MBF and MFR for dynamic 82Rb PET of large patients. Higher MBF values seen with TOF and PSF may be due to reduced sensitivity to inconsistencies with attenuation correction and reduced partial volume effects. The reason for increased MBF values in the septal wall for TOF based methods is under investigation. The MFR values with TOF and PSF had marginal change. Research Support: Siemens Medical Solutions