Residual activity correction for computing myocardial blood flow from dynamic ¹³NH₃ studies

Objectives A cardiac perfusion PET study requires at least 2 scans (rest and stress). With ¹³NH₃ (half life of 9.97min), some residual activity from the previous study could still be present in the subsequent scan, potentially introducing overestimation in the quantification of the myocardial blood flow (MBF) [1]. A method to estimate and correct the residual activity for computing MBF is presented.

Methods 24 patient rest and adenosine stress ¹³NH₃ datasets were obtained on a Siemens Biograph-64 PET/CT scanner. The acquisition of the stress scan started before the injection, so that the first frame of the dynamic sequence only contains residual activity from the rest. 16 frames (12x10s, 2x30s, 1x60s, 1x360s) were reconstructed. The residual activity is quantified using the value of the blood input function (BIF) and time activity curve (TAC) on the first frame, and corrected by subtracting it from all frames of the TACs and BIF. MBF and CFR values are computed using Siemens syngo.MBF software.

Results The time interval between the rest and stress injection is between 19 and 40 min. The mean (±SD) stress flow with and without residual correction is 2.09 ± 0.61 and 2.3 ± 0.65 ml/g/min, with correlation coefficient R² = 0.969, slope = 0.93, intercept = -0.05. The ratio of MBF with and without residual correction is 90.7% ± 4.9%. The mean first frame residual activity in the myocardium varies between 1,993 to 19,967 Bq/ml.

Conclusions Residual activities in the consecutive ¹³NH₃ studies introduce some overestimation of MBF. A first frame subtraction method is shown to correct the residual activity successfully for computing MBF