Hybrid CT/gamma camera PET for tumor imaging: comparison with dedicated pet scanner and potential treatment planning applications

Purpose: Hybrid imaging systems are being developed which combine x-ray CT scanning and PET imaging capabilities. The major clinical advantage of these systems is the ability to obtain anatomic and corresponding physiologic information by simultaneously examining co-registered CT and PET images. In addition, the CT scan provides low-noise transmission images, and this can result in improved attenuation-correction of the PET images, compared to conventional methods. The objective of this study was to compare imaging from a hybrid CT/camera-based imaging system with a dedicated full-ring PET scanner and investigate its suitability for radiation treatment planning. Materials and Methods: Twenty patients with suspected or proven malignancy and scheduled for a routine diagnostic PET scan were enrolled in the study. Subjects included six patients being evaluated for pulmonary nodules or staging lung cancer, 5 patients with melanoma, 4 with lymphoma, and 5 with other malignancies. Following injection of a single dose of F-18 fluorodeoxyglucose (FDG), patients were sequentially imaged on a dedicated PET scanner (GE Medical Systems Advance) and the hybrid system (GE Millenium VG with 1’ NaI detectors and x-ray CT [Hawkeye]). Images were reviewed blindly and independently by two experienced nuclear medicine physicians. All images were evaluated using interactive display on a GE workstation. Potential abnormalities were identified and scored from 1 (not abnormal) to 5 (definite abnormality). Comparison was made between: 1) dedicated PET images, 2) camera-based images without CT, and 3) camera-based images with co-registered CT images. Results: Twenty chest scans and 14 abdominal/pelvic scans were available for comparison. The dedicated PET images along with diagnostic CT and MRI studies and clinical data provided the clinical standard of comparison. Based on the dedicated PET scans, 23 abnormalities were identified in the chest and 23 in the abdomen/pelvis. When camera-based images were evaluated without CT information, there was 13/20 (65%) concordance in the chest and 13/14 (93%) concordance in the abdomen/pelvis. The concordance rate in the chest was improved to 15/20 (75%) when camera-based images were interpreted along with co-registered CT images. Camera-based PET imaging, with or without CT, was less sensitive for detecting small (<1 cm) lesions compared to dedicated PET imaging. However, in six cases the co-registered CT images allowed potential abnormalities to be more confidently interpreted as being non-malignant (focal bowel activity, subcutaneous injection site, normal thyroid activity, vascular activity). Conclusion: The current state-of-the-art PET system is more sensitive for detecting small lesions compared to the camera-based imaging system. Further study is required to determine the importance of detecting these small lesions for the purposes of radiation therapy treatment planning. The additional anatomic information provided by the CT scan on the hybrid system can improve anatomic localization of specific abnormalities such as lymph nodes. In addition, the co-registered CT images were occasionally important for confirming that other foci were related to benign processes such as physiologic muscular activity or arthritis. Using this hybrid scanner, CT images are acquired over 10 minutes with the patient quietly breathing. While the resulting images do not have the diagnostic quality of a single-breathhold spiral CT study, the relatively long CT image acquisition time provides more accurate spatial correlation with the PET image acquisition. Furthermore, CT/PET imaging with the patient quietly breathing is a more accurate reflection of the patient during radiation therapy, and the hybrid CT/camera-based imaging system may have particular advantages for functional image-guided radiation therapy treatment planning.