Patient motion during dynamic PET imaging can induce errors in myocardial blood flow (MBF) estimation. Motion correction for dynamic cardiac PET is challenging because the rapid tracer kinetics of 82Rb leads to substantial tracer distribution change across different dynamic frames over time, which can cause difficulties for image registration-based motion correction, particularly for early dynamic frames. In this paper, we developed an automatic deep learning-based motion correction (DeepMC) method for dynamic cardiac PET. In this study we focused on the detection and correction of inter-frame rigid translational motion caused by voluntary body movement and pattern change of respiratory motion. A bidirectional-3D LSTM network was developed to fully utilize both local and nonlocal temporal information in the 4D dynamic image data for motion detection. The network was trained and evaluated over motion-free patient scans with simulated motion so that the motion ground-truths are available, where one million samples based on 65 patient scans were used in training, and 600 samples based on 20 patient scans were used in evaluation. The proposed method was also evaluated using additional 10 patient datasets with real motion. We demonstrated that the proposed DeepMC obtained superior performance compared to conventional registration-based methods and other convolutional neural networks (CNN), in terms of motion estimation and MBF quantification accuracy. Once trained, DeepMC is much faster than the registration-based methods and can be easily integrated into the clinical workflow. In the future work, additional investigation is needed to evaluate this approach in a clinical context with realistic patient motion.

Accurate and reproducible SPECT quantification of myocardial molecular processes remains a challenge because of the complication of heterogeneous background and extracardiac activity adjacent to the heart, which causes errors in the estimation of myocardial focal tracer uptake. Our aim in this study was to introduce a heuristic method for the correction of extracardiac activity into SPECT quantification and validate the modified quantification method for accuracy and reproducibility using a canine model. Methods Dual-isotope–targeted 99mTc and 201Tl perfusion SPECT images were acquired using a hybrid SPECT/CT camera in 6 dogs at 2 wk after myocardial infarction. Images were reconstructed with and without CT-based attenuation correction, and the reconstructed SPECT images were filtered and quantified simultaneously with incorporation of extracardiac radioactivity correction, gaussian fitting, and total-count sampling. Absolute myocardial focal tracer uptake was quantified from SPECT images using 3 different normal limits (maximum entropy [ME], mean-squared-error minimization [MSEM], and global minimum [GM]). SPECT-quantified percentage injected dose (%ID) was calculated and compared with the well-counted radioactivity measured from the postmortem myocardial tissue. SPECT quantitative processing was performed by 2 different individuals with extensive experience in cardiac image processing, to assess reproducibility of the quantitative analysis. Results Correlations between SPECT-quantified and well-counted %IDs using 3 different normal limits were excellent (ME: r = 0.82, y = 0.932x − 0.0102; MSEM: r = 0.73, y = 1.1413x − 0.0052; and GM: r = 0.7, y = 1.2147x − 0.0002). SPECT quantification using ME normal limits resulted in an underestimation of %ID, as compared with well-counted %ID. Myocardial focal tracer uptake quantified from SPECT images without CT-based attenuation correction was significantly lower than that with the attenuation correction. The %IDs quantified from attenuation-corrected SPECT images using MSEM and GM normal limits were not significantly different from well-counted %IDs. Reproducibility of the SPECT quantitative analysis was excellent (ME: r = 0.98, y = 0.9221x + 0.0001; MSEM: r = 0.97, y = 0.9357x + 0.0004; and GM: r = 0.96, y = 0.9026x + 0.001). Conclusion Our SPECT/CT quantification algorithm for the assessment of regional radioactivity may allow for accurate and reproducible serial noninvasive evaluation of molecularly targeted tracers in the myocardium.

No abstract

No abstract

No abstract