Characterization of a pinhole tomograph with 180° acquisition using a discrete vertex set reconstruction algorithm

Pinhole single photon emission computed tomography (SPECT) offers attractions for thyroid imaging, but acquisition must be limited to 180°. The authors have mounted a 3.2 mm pinhole collimator on a rotating gamma camera, magnification 2.1, and characterized its response in the 10 cm field of view (FOV) by imaging points, lines, and a Picker thyroid phantom, comparing both full and half circle data acquisition. Image reconstruction was done with a new Discrete Vertex Set (DVS) algorithm and the Feldkamp (FDK) algorithm. For full circle acquisition, spatial resolution is quite uniform across the FOV, averaging 5.3 mm FWHM, and the reconstructed images show an intensity variation less than 5% in all directions. For this level of noise, the FDK and DVS algorithms perform similarly. For half circle reconstruction, spatial resolution shows more variation across the FOV, averaging 5.4 mm FWHM, and the maximum intensity variation is 20%. Differences between the FDK and DVS algorithms are more apparent. The thyroid phantom and its cold spots can be visualized in both full and half circle reconstructions, allowing size determination