High-resolution imaging of pulmonary ventilation and perfusion with 68Ga-VQ respiratory gated (4-D) PET/CT.

PURPOSE: Our group has previously reported on the use of (68)Ga-ventilation/perfusion (VQ) PET/CT scanning for the diagnosis of pulmonary embolism. We describe here the acquisition methodology for (68)Ga-VQ respiratory gated (4-D) PET/CT and the effects of respiratory motion on image coregistration in VQ scanning.
METHODS: A prospective study was performed in 15 patients with non-small-cell lung cancer. 4-D PET and 4-D CT images were acquired using an infrared marker on the patient's abdomen as a surrogate for breathing motion following inhalation of Galligas and intravenous administration of (68)Ga-macroaggregated albumin. Images were reconstructed with phase-matched attenuation correction. The lungs were contoured on CT and PET VQ images during free-breathing (FB) and at maximum inspiration (Insp) and expiration (Exp). The similarity between PET and CT volumes was measured using the Dice coefficient (DC) comparing the following groups; (1) FB-PET/CT, (2) InspPET/InspCT, (3) ExpPET/Exp CT, and (4) FB-PET/AveCT. A repeated measures one-way ANOVA with multiple comparison Tukey tests were performed to evaluate any difference between the groups. Diaphragmatic motion in the superior-inferior direction on the 4-D CT scan was also measured.
RESULTS: 4-D VQ scanning was successful in all patients without additional acquisition time compared to the nongated technique. The highest volume overlap was between ExpPET and ExpCT and between FB-PET and AveCT with a DC of 0.82 and 0.80 for ventilation and perfusion, respectively. This was significantly better than the DC comparing the other groups (0.78-0.79, p < 0.05). These values agreed with a visual inspection of the images with improved image coregistration around the lung bases. The diaphragmatic motion during the 4-D CT scan was highly variable with a range of 0.4-3.4 cm (SD 0.81 cm) in the right lung and 0-2.8 cm (SD 0.83 cm) in the left lung. Right-sided diaphragmatic nerve palsy was observed in 3 of 15 patients.
CONCLUSION: (68)Ga-VQ 4-D PET/CT is feasible and the blurring caused by respiratory motion is well corrected with 4-D acquisition, which principally reduces artefact at the lung bases. The images with the highest spatial overlap were the combined expiration phase or FB PET and average CT. With higher resolution than SPECT/CT, the PET/CT technique has a broad range of potential clinical applications including diagnostic algorithms for patients with suspected pulmonary embolism, preoperative evaluation of regional lung function and improving assessment or understanding of pulmonary physiology in the vast range of pulmonary diseases.