Accuracy verification of PET-CT image fusion and its utilization in target delineation of radiotherapy
Heyi Gong
Abstract
Heyi Gong
Abstract
Objective Evaluate the accuracy o f co-registration of PET and C T (PET-CT) images on line with phantom, and utilize it on patients to provide c l inical evidence for target delineation in radiotherapy. Methods A phantom with m arkers and different volume cylinders was infused with various concentrations of 18FDG, and scanned at 4?mm by PET and CT respectively. Aft er having been transmi tted into GE eNTEGRA and treatment planning system (TPS) workstations, the image s were fused and reconstructed. The distance between the markers and the errors were monitored in PET and CT images respectively. The volume of cylinder in PET and CT images were measured and compared by certain pixel value proportion deduc tion method. The same procedure was performed on the pulmonary tumor image in te n patients. Results eNTEGRA and TPS workstations had a good length linearity, b ut the fusion error of the latter was markedly greater than the former. Tumors i n different volume filled by varying concentrations of 18FDG required different pixel deduction proportion. The cylinder volume of PET and CT images were almost the same, so were the images of pulmonary tumor of ten patients. Conclusions T he accuracy of image co-registration of PET-CT on line may fulfill the clinica l demand. Pixel value proportion deduction method can be used for target delineati on on PET image.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Objective Evaluate the accuracy o f co-registration of PET and C T (PET-CT) images on line with phantom, and utilize it on patients to provide c l inical evidence for target delineation in radiotherapy. Methods A phantom with m arkers and different volume cylinders was infused with various concentrations of 18FDG, and scanned at 4?mm by PET and CT respectively. Aft er having been transmi tted into GE eNTEGRA and treatment planning system (TPS) workstations, the image s were fused and reconstructed. The distance between the markers and the errors were monitored in PET and CT images respectively. The volume of cylinder in PET and CT images were measured and compared by certain pixel value proportion deduc tion method. The same procedure was performed on the pulmonary tumor image in te n patients. Results eNTEGRA and TPS workstations had a good length linearity, b ut the fusion error of the latter was markedly greater than the former. Tumors i n different volume filled by varying concentrations of 18FDG required different pixel deduction proportion. The cylinder volume of PET and CT images were almost the same, so were the images of pulmonary tumor of ten patients. Conclusions T he accuracy of image co-registration of PET-CT on line may fulfill the clinica l demand. Pixel value proportion deduction method can be used for target delineati on on PET image.
Key concepts: Nuclear medicine, Imaging phantom, Volume (thermodynamics), Image fusion, Pixel, Medicine, Image registration, Radiation therapy