It's not an alternative to volume rendering as the title suggests. It's a kind of a brand name the manufacturer wants to establish for their admittedly very impressive volume rendering implementation
It seems like these guys added lighting to the visualization of CT volumes and decided to call it "Cinematic Rendering" or "CR" for short.
I'm not sure what possessed these people to take 40 year old techniques and decide that they could come up with a hyperbolic name and a new acronym then not have everyone just roll their eyes.
Are there any physicians on here that find these renders useful? As a radiology resident, I often prefer reading from the source images, as the whole point is to see inside a patient, rather than provide some partial surface rendering.
Non-radiologists seem to love these sorts of images, but I don't think they are all that helpful.
They are where form and contour matters, which you can't always deduce from a orthographic 3 Panel view.
In my current case the form and thickness variation of nerves in Parkinson's patients is looked at and analyzed. Since nerve curve through space like worms, looking at curvature in raw image data is unpractical, especially since it's not always apparent whether you look at the cross section at an angle.
I’m not sure if it’s a language issue or a research vs clinical practice issue, but this sounds like something that diffusion tensor imaging would used for with special analysis packages for analyzing the volume and spatial relationships of neurons.
The software shown is for CT data, which is abysmal for evaluating white matter bundles. Any links to your research? Always useful to learn about other applications of imaging.
This renderer wouldn't really help with that problem though would it? It sounds like what you could use is software to transform that curved path into a straight line so you can look at it end on.
Most radiologists don't seem interested. I don't know if it is due to work flow, being used to thousands of sliced orthogonal CT readings or both, but the idea alone doesn't seem to get anyone excited.
Also on the workflow aspect I think no one wants to wait for something that takes longer to generate, longer to load, and possibly can't be shared as easily through epic or some other software.
I mean, our entire specialty day-in and day-out is looking through thinly sliced images of patients to diagnose and monitor disease. We are absolutely used to triplane images. I also agree that waiting for 3D renders is a huge time suck and often why we will just read from the source data.
They can be shared just as easily since they are usually exported back into the VNA (archive) with the original DICOM source data. However, I’m curious who this product is for, because most radiologists don’t need or want it.
Is it for:
PCPs?
Presurgical planning?
Patients?
Research?
This is very interesting, the amount of rendering complexity they pull from volume data is nothing short of impressive. Having never ventured beyond Orthogonal projections for pseudo 3D in my work, this is some fresh air.
But it's rather useless as a standalone technology. Much of the appeal comes from pulling color and using it to render from segmented volume data, which is rarely the case in the field.
"Scattering effects are modelled using a local gradient shading model"
This can mess with perception, if done incorrectly, especially where precision is spares like Volumetric data pulled from Ultrasound, not to mention such a rendering model in non segmented data sets is questionable.
But one hell of a tool to convince a board meeting of whatever.
