Standard versus low-dose rubidium-82 dynamic positron emission tomography imaging with scanner-dependent bias correction for myocardial perfusion imaging and blood flow quantification

Introduction: 3D PET-CT is evolving as the standard for myocardial perfusion imaging (MPI) due to its increased sensitivity and high image quality with lower patient doses. Accurate corrections for increased random and scattered counts are required for accurate quantification of myocardial blood flow (MBF), particularly during the first-pass tracer transit when detector saturation and dead-times are highest. Administered doses could be further reduced to minimize dead-time losses; however, this can compromise the MPI image quality, and effects on MBF are not well-established. The objective of this study was to assess the quality of MPI images and accuracy of MBF quantification with standard-dose and low-dose Rb-82 PET using a post-reconstruction activity bias correction

Methods:

Methods: N = 50 patients underwent standard-dose Rb-82 rest dynamic imaging on a LBS 3D PET-CT camera with injected activity of 9 MBq/kg, followed immediately by a second, low-dose, rest acquisition with half of the initial injected activity (4.5 MBq/kg). Static MPI (2-4 min data) and dynamic images (0-6 min data) were reconstructed separately for STD and LOW list-mode data. Image quality of the MPI scans was assessed via signal-to-noise ratio (SNR). Dynamic images were analyzed using a one-tissue compartment model to measure MBF. Previously established post-reconstruction bias correction functions were applied to all datasets to correct for underestimation of reconstructed image activity at high dead-time values. Image quality and MBF values were compared between standard and low-dose scans, with (STD BC, LOW BC) and without (STD, LOW) bias correction.

Results: MPI image quality was found to be significantly higher for the STD vs. LOW scans: SNR STD = 9.5 ± 2.0 vs. SNR LOW = 8.6 ± 1.6, p = 0.002. Rest MBF estimates were determined to be 0.81 ± 0.24 mL/min/g for STD scans and 0.80 ± 0.26 mL/min/g for LOW scans. Following application of the previously established bias correction function (BCF = 0.17×DTF² + 0.038×DTF + 1), MBF estimates were decreased to 0.77 ± 0.23 for STD BC scans and 0.79 ± 0.25 for LOW BC scans. This represents a significant difference in the MBF estimates after bias correction for STD vs LOW scans (p< 0.001 and p = 0.03, respectively). The BCF had a bigger effect on the STD vs LOW dose scans as expected, since the dead-time effects are higher using the standard dose. After bias correction of the standard-dose scans (STD BC), there was no residual bias vs. the LOW or LOW BC scan results (p = NS). These results suggest that low-dose imaging is sufficient to obtain accurate quantitative MBF estimates, whereas standard-dose imaging allows for both diagnostic quality MPI images and accurate MBF quantification following bias correction. Conclusion: The accuracy of dynamic 3D PET myocardial blood flow measurements can be improved using a non-linear calibration of the reconstructed image activity as a function of scanner-recorded dead-time factors. Standard-dose imaging provides diagnostic quality MPI images and accurate MBF quantification following bias correction.