The VR were poor in branched blood vessels at 150 and 300 HU(P<0.05). however, at 450 HU, they possessed an equivalent image quality with those at 600 HU (n.s). (Fig.7)
However, the visual evaluation of the MIP images showed that the LSD distance was 0, indicating that there was a significant difference in all contrasts. (Fig.8)
We have confirmed that the relationship between image quality and dose decreases in both VR and MIP as the dose decreases. We show the results for SD6 and SD8 in Figure 9.10. As shown in the red frame in Fig. 11, in the VR image, our visual evaluation shows that the image quality equivalent to the original image (600HU, 785mAs) is 25% contrast agent reduced image (450HU, 785mAs), and the exposure dose is Half image (600HU, 350mAs). Also, as shown in the blue frame in Fig. 10, if the morphology of blood vessels is confirmed, it is acceptable to use a contrast agent reduced 50% image (300HU, 785mAs) and an exposure dose reduced 70% image (600HU, 200mAs).
We believe that it is possible to reduce the contrast medium by approximately 25 % for thin blood vessels of 1 mm or less which have an equivalent image quality with Conventional image, and approximately 50 % for branch confirmation by using VMI. We believe that it is necessary to have sufficient dose because the thin blood vessels are buried in the noise and the image quality were poor if the dose is low.However, in the VR image, it is possible to reproduce the shape of the present trunk blood vessel (3 mm) with a contrast agent of 25% and a dose of about 25%. Therefore, it has been confirmed that the effect of the VMI image changes considerably depending on the object.
According to the present study, blood vessels of 1 mm or less are near the limit of the spatial resolution of CT, and the CT value of blood vessels may be reduced due to the partial volume effect due to 512 pixels and a slice thickness of 0.8 mm.
Therefore, even if contrast sensitivity is increased in low Kev VIM images, the decrease in contrast agent in thin vessels less than 1 mm may not be equal.
However, it is considered that a contrast agent reduction of about 75% is possible for a blood vessel of about 3 mm.
We also believe that VIM images can contribute to a reduction in exposure dose. In this experiment, it was suggested that if the contrast was the same, it could be reduced by about 50%.
We believe that it is possible to reduce the contrast medium by approximately 25 % for thin blood vessels of 1 mm or less which have an equivalent image quality with FBP, and approximately 50 % for branch confirmation by using VMI.
We believe that it is necessary to have sufficient dose because the thin blood vessels are buried in the noise and the image quality were poor if the dose is low.