PURPOSE: Positron emission tomography (PET) is considered the gold standard for measuring myocardial blood flow in vivo but it is known that respiratory motion can lead to misalignment of the PET and computed tomography (CT) data sets and introduce artifacts in the CT-based attenuation correction (AC) of images. In addition, respiratory motion blurs the PET image and degrades spatial resolution. The purpose of this study is to evaluate the combined effect of respiratory motion compensation (MC) and accurate attenuation correction on relative and absolute blood flow imaging of the heart. METHODS: Dynamic (82)Rb-PET acquisitions were generated for a homogeneous tracer distribution in the heart using an anthropomorphic computer phantom and a Monte Carlo simulator. Attenuation correction was done using three different approaches in which the PET data were corrected by: (1) a respiratory-gated CT map with each respiratory phase of the PET scan corrected by its corresponding CT phase (matched); (2) a time-averaged attenuation map (avg); or (3) an attenuation map generated from the maximum CT-number of every voxel over the respiratory cycle (max). Motion compensated was done using an automated rigid-body registration algorithm that aligned all of the phases of the respiratory-gated PET data after AC. The corrected dynamic PET data were then processed by inhouse kinetic analysis software to generate 3D maps of blood flow. Polar maps of the blood-flow for each CT-AC method with and without MC were compared to the truth using a 17-segment model. The same comparison was performed on data from a pig study. RESULTS: Motion compensation significantly reduced the segmental mean percentage error (sMPE) in all cases (p < 0.01 for matched CTAC and avg CTAC and p = 0.03 for max CTAC). MC significantly increased image uniformity in the case of matched and avg CTAC (p < 0.01, p = 0.04, respectively) with the best improvement coming for matched CTAC. Without MC, there were no significant differences between the three CTAC approaches. With MC, matched CTAC had significantly smaller mean absolute sMPE (p < 0.01 vs avg CTAC; p < 0.01 vs max CTAC) and improved uniformity (p = 0.05 vs avg CTAC; p < 0.01 vs max CTAC). The results were supported with a pig study. CONCLUSIONS: Without MC, there was no significant difference between the three CTAC methods for measuring blood flow. With MC, the matched CTAC approach was significantly better, reducing the mean difference from truth by 6% in the simulated data and improving uniformity by 5%.
PURPOSE: Positron emission tomography (PET) is considered the gold standard for measuring myocardial blood flow in vivo but it is known that respiratory motion can lead to misalignment of the PET and computed tomography (CT) data sets and introduce artifacts in the CT-based attenuation correction (AC) of images. In addition, respiratory motion blurs the PET image and degrades spatial resolution. The purpose of this study is to evaluate the combined effect of respiratory motion compensation (MC) and accurate attenuation correction on relative and absolute blood flow imaging of the heart. METHODS: Dynamic (82)Rb-PET acquisitions were generated for a homogeneous tracer distribution in the heart using an anthropomorphic computer phantom and a Monte Carlo simulator. Attenuation correction was done using three different approaches in which the PET data were corrected by: (1) a respiratory-gated CT map with each respiratory phase of the PET scan corrected by its corresponding CT phase (matched); (2) a time-averaged attenuation map (avg); or (3) an attenuation map generated from the maximum CT-number of every voxel over the respiratory cycle (max). Motion compensated was done using an automated rigid-body registration algorithm that aligned all of the phases of the respiratory-gated PET data after AC. The corrected dynamic PET data were then processed by inhouse kinetic analysis software to generate 3D maps of blood flow. Polar maps of the blood-flow for each CT-AC method with and without MC were compared to the truth using a 17-segment model. The same comparison was performed on data from a pig study. RESULTS: Motion compensation significantly reduced the segmental mean percentage error (sMPE) in all cases (p < 0.01 for matched CTAC and avg CTAC and p = 0.03 for max CTAC). MC significantly increased image uniformity in the case of matched and avg CTAC (p < 0.01, p = 0.04, respectively) with the best improvement coming for matched CTAC. Without MC, there were no significant differences between the three CTAC approaches. With MC, matched CTAC had significantly smaller mean absolute sMPE (p < 0.01 vs avg CTAC; p < 0.01 vs max CTAC) and improved uniformity (p = 0.05 vs avg CTAC; p < 0.01 vs max CTAC). The results were supported with a pig study. CONCLUSIONS: Without MC, there was no significant difference between the three CTAC methods for measuring blood flow. With MC, the matched CTAC approach was significantly better, reducing the mean difference from truth by 6% in the simulated data and improving uniformity by 5%.
Authors: Jonathan B Moody; Benjamin C Lee; James R Corbett; Edward P Ficaro; Venkatesh L Murthy Journal: J Nucl Cardiol Date: 2015-04-14 Impact factor: 5.952
Authors: Chung Chan; Mark Harris; Max Le; James Biondi; Yariv Grobshtein; Yi-Hwa Liu; Albert J Sinusas; Chi Liu Journal: Phys Med Biol Date: 2014-09-26 Impact factor: 3.609
Authors: Paul K R Dasari; Arda Könik; P Hendrik Pretorius; Karen L Johnson; William P Segars; Mohammed S Shazeeb; Michael A King Journal: Med Phys Date: 2017-02 Impact factor: 4.071