The influence of zooming on centre-of-rotation (COR) of the rotating camera-based SPECT system and on the resolution of tomographic images was investigated. COR values were measured as a function of zoom factor varying from xl.O to x3.0. Measurements were performed using two SPECT systems collimated with three collimators. In this study, the COR was observed to shift linearly with zoom factor. The influence of zooming on COR depends on the location and the deviation of the COR from the axis-of-rotation (AOR). We recommend that the deviation of the COR from the AOR be kept to within 0.4 pixels corresponding to 2.4 mm. Unlike zooming, the magnitude of COR shift is uncorrelated to the type of collimator used. Contrast and spatial resolution variations of the tomographic images were quantitated as a function of COR shift for clinical data acquired using no zoom factor and a zoom factor of 2.50. The resolution indices were observed to be more sensitive to COR shift for clinical data acquired using no zoom factor. The sensitivity appeared to be inversely dependent on the zoom factor utilized. However, zooming was found not to preserve the same resolution degradation. Optimum resolution images can be reconstructed by requiring the COR be determined and clinical study be acquired using the same zoom factor and collimator. For qualitative evaluation of tomographic images, the zoom factor requirement can be removed provided that the deviation of the COR from the AOR is within the recommended limit. This limit allows for a maximum spatial resolution degradation of 5%. Based on the information extracted, three alternative methods have been suggested to alleviate the influence of zooming on COR. These methods can be easily incorporated in the SPECT quality control procedures.
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging