system shielding. Slots machined into the system shielding allow the easy interchange of pinhole apertures (three plates are pictured, surrounding the object space). Four modular gamma cameras are used with the system with one abutting each of the ports seen in the shielding. A lid is affixed and the entire system is elevated so that a rotation stage may be used for tomographic data acquisition.
Example signal images (generated from a signal) used in the study. (a) Aperture 0 had a magnification of 1.6 and five pinholes. (b) Aperture 1 had a magnification of 3.33 and one pinhole. (c) Aperture 2 had a magnification of 2.47 and nine pinholes. (d) Aperture 3 had a magnification of 3.33 and four pinholes.
Bead phantom used to generate random background images. The phantom is filled with nonuniform polystyrene beads which are then stirred after each image to simulate object variability. Radioactive liquid occupies the space surrounding the beads.
Projection images of an example bead background dominated by (a) counting noise, and (b) structured bead noise. Projection images of an example lumpy background dominated by (c) counting noise, and (d) structured lump noise.
Example 2D projections of 3D lumpy backgrounds for lump widths of (a) , (b) , and (c) .
SNR for detection in a lumpy background with increasing exposure time. (a) Signal size is . (b) Signal size is (c) Signal size is . Please refer to Table I for aperture information. Note: Scales have been fixed across the three plots in this figure.
SNR for detection in a bead background with increasing exposure time. (a) Signal size is . (b) Signal size is . (c) Signal size is . Please refer to Table I for aperture information. Note: Scales have been fixed across the three plots in this figure.
SNR for different backgrounds as a function of exposure time. (a) Signal diameter , Aperture 1. (b) Signal diameter , Aperture 3. (c) Signal diameter , Aperture 0. Please refer to Table I for aperture information. Note: Scales have been fixed across the three plots in this figure.
(a) AUC as a function of channel width for three different aperture and signal combinations. Channel widths producing the minimum MSE between the signal and the 0th order Laguerre–Gauss polynomial are indicated by arrows. (b) AUC as a function of number of channels for the same three aperture and signal combinations as were used in (a). The arrow indicates the number of channels for the signal-detection study in lumpy backgrounds. Channel widths used for this study are the same as indicated by the arrows in (a).
(a) Maximum AUC values (out of 20 possible aperture combinations) as a function of signal diameter and lump size (a total of 77 combinations). (b) The total number of occurrences of each aperture in the top-five performing aperture combinations for all signal and lump combinations.
A map of the frequency of occurrence of Apertures 0, 1, and 3 among the top-five performing aperture combinations as a function of signal size and lump size. Green corresponds to Aperture 0, blue corresponds to Aperture 1, and red corresponds to Aperture 3. The plot is dominated by Aperture 1 (blue and purple sections) but there is a large region in which Aperture 3 appears most critical to generating high AUC and a smaller region where Aperture 0 is most critical.
Example system configuration. Aperture number, magnification, number of pinholes, pinhole diameter, sensitivity (normalized to the most sensitive aperture), and full opening angle are provided.
The first six columns of the table provide aperture combinations, AUC, and standard deviation values for two example signal and lump combinations. Note the differences in rank order for the aperture combinations. The final three columns provide ranked aperture combinations, AUC and standard deviation values for all aperture combinations averaged over the 77 signal and lump pair combinations. represents signal diameter while refers to lump width.
Aperture combinations returning maximum AUC values for all signal and lump combinations.
Article metrics loading...
Full text loading...