CatPhan® 600 phantom and its module specification. The module section CTP528 has high resolution multiple line pairs and was used in this work to test the spatial resolution loss during the proposed image processing.
The experimental design showing the power supply that drives the electric motor, which in turn moves the motion platform and phantom sinusoidally in the cranial–caudal direction as it moves through the CT bore.
The RPM infrared camera and illuminator system is mounted at the foot of the couch and used to track the motion of infrared reflectors placed on the top of the thorax phantom or the diaphragm of the patient.
An illustration of the respiratory motion waveform recorded by the RPM. Each diamond marker reflects the motion status when a CT volumetric image (for one couch position) is generated. The example here shows that 16 CT images were generated in at each couch position.
An illustration of the registration error. (a) A perfect match of the corresponding voxels; (b) when errors exist, the voxels transformed by registration may not match exactly their targeted locations.
Phantom study for the 4D-PWLS method with the CatPhan® 600 phantom. The CTP528 section (high-resolution line pairs) is compared in two window width/level settings for CT image before (top row) and after (bottom row) the proposed 4D-PWLS smoothing. The left column is displayed with window width 200, level 80; the right column is displayed with window width 1500, level 150.
Profiles across the strips indicated by the squares in the images of Fig. 6. The dotted line shows the profile of the original data, and solid line is the profile for the smoothed data.
Phantom study for the 4D-PWLS method with the thorax phantom. The left and middle columns are the original phases obtained from the GE Advantage Workstation, for 100 and , respectively; the right column shows the phases after 4D-PWLS processing. From top to bottom are phase 0%, 20%, 40%, 60%, 80%, respectively. The red rectangles represent the selected ROI for calculation of SNRs, each of which contains voxels.
Axial CT images of the thorax phantom. From top to bottom are the , , and 4D-PWLS smoothed images. The yellow squares show the air holes for measuring the “signal,” the red square shows the region for measuring the “background.” The relative contrast is defined as the CT number difference between the signal and the background.
The relations of noise and relative contrast with the penalty weight . The noise is characterized by the standard deviation of the CT numbers in the uniform region shown in Fig. 9. The red lines in the figure show the noise and relative contrast of the image. The blue line and diamond marks are for the smoothed images with different . When it represents the original data.
(a) Patient study for the 4D-PWLS method at the end-inspiration phase. The left column contains the original images acquired from the GE Advantage Workstation, and the right column shows the image after 4D-PWLS processing. The red rectangles represent the selected ROI for the calculation of SNRs, each of which contain voxels. (b). A comparison of the horizontal profiles across the center of the transaxial images in (a).
(a) Patient study for the 4D-PWLS method at the end-expiration phase. The left column contains the original images acquired from the GE Advantage Workstation, and the right column shows the image after 4D-PWLS processing. The red rectangles represent the selected ROI for calculation of SNRs, each of which contain voxels. (b). A comparison of the horizontal profiles across the center of the transaxial images in (a).
A comparison of PWLS with and without registered phase information. The top row is the result of 3D PWLS smoothing without using temporal information for the end-inspiration phase patient data. The middle row shows the difference images between the 3D-PWLS smoothing results (top row in this figure) and original images [the left column in Fig. 11(a)]. The bottom row contains the difference images between 4D-PWLS smoothing results [the right column in Fig. 11(a)] and original images. The fewer edges observed in the difference image indicate an improved preservation of the spatial resolution in the 4D method. The red rectangles represent the selected ROI for calculation of SNRs, each of which contain voxels.
Mean and standard deviation (SD) of the CT number in the selected ROIs before and after the proposed 4D smoothing for the scan of the CatPhan 600 phantom. ROI-1 is the square region in the center, ROI-2 is the square region above the center, ROI-3 is the square region above the center.
A comparison of SNRs in the thorax phantom study for different phases. The penalty was in the 4D-PWLS smoothing.
A comparison of relative contrast and noise in the thorax phantom study.
Article metrics loading...
Full text loading...