Current Clinical Concerns in CT: Results : 3D Rendering
3D Algorithms
1. I work on a volume zoom in a small hospital. We do very few CTA's of renals, carotids, etc. I have the protocols for scanning, but need more information on MIPS, volume rendering, etc. This is all foreign to me. We had applications come in twice, but it was only touched on. I need printed material that I can follow when I get these cases.
Answer:
|
2. RE: settings for volume rendered images. Your volume rendered images are really nice. We are just getting up and running with our new GE Lightspeed 8 and 16 scanners, and TeraRecon workstations. I'm still learning all the capabilities of our workstations, but haven't been able to figure out how to reproduce the images you are publishing. I've asked TeraRecon to look into this, but I also wonder if there are any Mask settings that you may have that would be transferable to other workstations, that you might be able to post? I've learned a great deal from this website, and I applaud the service your are providing to the entire radiologic community.
Answer: Templates are very specific to a workstation and are not simply transferred. I just spoke with Robert Taylor of TeraRecon and he said it might just be best to contact him, and he will get someone to update your system. For the brave, he did list this memo as well: (dated 3/16/03) Dear All, |
3. RE: VR for estimation of vascular stenoses. One of my colleagues has reservations using 3D vascular mapping to estimate the % stenosis on CTA. In particular, she is concerned about the effect of changing the windowing or trapezoid parameters on the apparent vessel caliber. We have all seen stenoses made to look like occlusions by changing the display. How do we avoid spuriously over- or under- estimating the degree of stenosis on arteries?
Answer: There is a learning process, but once you master the technique it works well. An article referenced below evaluated the accuracy of volume rendering. |
4. We have a Leonardo workstation, but do not have a MINIP tool button. We have an extra effort to do MINIP applications. Is it a general problem for other workstations, or a local problem?
Answer: You must go to www.InsideInspace.com and it will show you where on Inspace the MINIP is. If this doesn't help, let me know. |
References
BS Kuszyk, DG Heath, PT Johnson, J Eng, and EK Fishman CT angiography with volume rendering for quantifying vascular stenoses: in vitro validation of accuracy. AJR 1999; 173: 449 - 455.
- Summary: In this study, 3 phantoms of vascular stenoses (33%, 67% and 83%) were scanned with single detector helical CT using various acquisition and reconstruction protocols. Volume rendering was conducted to assess diameter and percentage of stenosis. Results showed that vessel orientation significantly impacted the accuracy, with phantoms coursing parallel to the axial plane resulting in a less accurate measured percentage of stenosis. Furthermore, results showed that for 33% and 67% stenosis, the optimal opacity to maximize accuracy at VR was 50%. In the setting of 83% stenosis, an opacity setting of 60% provided the most accurate rendering.
Calhoun PS, Kuszyk BS, Heath DG, Carley JC, and Fishman EK.
Three-dimensional Volume Rendering of Spiral CT Data: Theory and Method.
RadioGraphics, May 1999; 19: 745 - 764.
- Summary: This review provides an excellent explanation of different rendering algorithms, including MIP, shaded surface rendering and volume rendering. Such information is essential to those using these algorithms for clinical imaging, so that the strengths and weaknesses of each rendering technique are understood.