Fig 1.
Deployment of the Hoffman 3D brain phantom and an optical tracking device.
(a) Simulated rotational motion of the head. (b)Simulated tilting motion of the head. (c)Infrared reflective marker and Polaris Vicra optical tracking device.
Fig 2.
Outline of movements during PET scans.
Fig 3.
Evaluation of % contrast.
Fig 4.
Motion plots obtained from the Polaris Vicra optical tracking and DDBMC.
Fig 4 shows the motion plots used in the ultra-short (1 second) DDBMC for Polaris Vicra optical tracking and listmode data in each case. (A) shows the relationship between displacement in the X, Y, and Z directions and elapsed time, where the vertical axis is displacement (mm) and the horizontal axis is time (seconds). (B) shows the relationship between elapsed time and angular displacement in the X, Y, and Z directions, with angle (degrees) on the vertical axis and time (seconds) on the horizontal axis.
Fig 5.
Images with and without DDBMC for data containing rotational motion.
The left images are without DDBMC and the right images are with DDBMC. Without DDBMC, the brain was tilted to the right and the entire brain was blurry (➤). With DDBMC, the brain structure was clearer and the contrast between gray and white matter appeared improved.
Fig 6.
Images with and without DDBMC for data containing tilting motion.
In the left images without DDBMC, the structures of the basal ganglia were clear in the transverse sections and the contrast between gray and white matter was preserved, making it difficult to discern the effects of movement. In the sagittal sections, the influence of motion could be observed in the frontal lobe (➤). In the right images with DDBMC, the displacement observed in the sagittal slices disappeared and the effect of correction could be seen.
Table 1.
Visual assessment scores by professional doctors and radiologists.
Fig 7.
Normalized mean square error in transverse images.
The normalized mean square error (NMSE) results for reference images in the static state and images with each motion, compared with and without DDBMC, are shown. (A) depicts Case 1 with continuous rotational motion, (B) shows Case 2 with stepwise rotational motion, and (C) illustrates Case 3 with tilting motion. The closer the NMSE value on the vertical axis to 0, the closer the image is to the reference image.
Table 2.
Normalized mean square error and structural similarity index with and without DDBMC.
Table 3.
Comparison of % contrast between images with and without DDBMC and reference images.