Figure 1.
A typical slice of the FORBILD head phantom.
Figure 2.
Images reconstructed by SART (first row), TDM-STF (second row) and WTDM-STF (third row) algorithms after 50, 100, 200, 400 iterations using noise-free projections, respectively.
The gray scale window is set to [1.03, 1.08].
Figure 3.
Zoom-in views of images reconstructed by SART (first row), TDM-STF (second row) and WTDM-STF (third row) algorithms after 50, 100, 200, 400 iterations, respectively.
The gray scale window is set to [1.03, 1.08].
Figure 4.
1D profiles of the images reconstructed by different algorithms using noise-free projections.
(a) Horizontal profiles (240th row, 200th column to the 300th column); (b) Vertical profiles (258th column, 180th row to the 260th row).
Figure 5.
Different performance evaluations as a function of iteration numbers on the reconstructions by TDM-STF and WTDM-STF.
Table 1.
Evaluations of the results reconstructed by different algorithms (with 400 iterations from noise-free projections for a typical slice of the FORBILD head phantom).
Figure 6.
Images reconstructed by SART (first row), TDM-STF (second row) and WTDM-STF (third row) algorithms after 50, 100, 200, 400 iterations using noisy projections, respectively.
The gray scale window is set to [1.03, 1.08].
Figure 7.
Different performance evaluations as a function of iteration numbers on the reconstructions by TDM-STF and WTDM-STF from noisy projection data.
Table 2.
Evaluations of the results reconstructed by different algorithms (with 400 iterations from noisy projections for a typical slice of the FORBILD head phantom).
Table 3.
Summary of Welch’s t test analysis results of performance evaluations of image quality between different algorithms (with 100 iterations from noise-free projections for 500 slices of the FORBILD head phantom).
Table 4.
Summary of Welch’s t test analysis results of performance evaluations of image quality between different algorithms (with 100 iterations from noisy projections for 500 slices of the FORBILD head phantom).