RadiAnt DICOM Viewer - Discussion Forum
Pixelated image while browsing in MIP (MPR)

First of all, i wanted to thank you for this amazing piece of software. It really helps in my daily practise.
I have a little problem though. While browsing a serie in MIP/MinIP mode, the image shown is pixelated and i can't get the right information.
When i stop browsing, the image come back in full resolution.

I can send you a video capture if you want more explanation :)


Jan 22, 2017 01:17 PM
by montasser
Thank you for your kind words!
Rendering MIP images requires substantial computing power. So we have to balance between fluidity and quality when interacting with volumes in MIP mode. This is called "progressive rendering" and cannot be avoided with slower machines.


Jan 22, 2017 08:18 PM
by RadiAnt
I have an i7 6th generation processor with 12 gigs of RAM. Can't be that avoided with not so slow machines?


Feb 10, 2017 08:58 AM
by montasser
I have the same problems with a little older machine, but powerful anyway (I think) i7 3770 with 16Gb of RAM, 256 Gb SSD and an 8Gb memory video card.
is there a way to do that, or I must buy a workstation to achieve that.
thank you very much


Feb 13, 2017 09:37 AM
by cezar matau
How I understand it, the problem is that the rendering is CPU based. Any CPU, even a fast one, can only run 2-8 threads parallel. The GPUs can compute much more in parallel.

We have to wait till they release the GPU based version.


Feb 13, 2017 01:54 PM
by aaaaaaa
Increasing thickness of a MIP slab quickly "saturates" even the fastest modern CPUs (if you want to browse real-time, full screen).
GPU rendering will alleviate this problem.
Currently what you can do is to make the 3D MPR window as small as possible to reduce the number of calculations. We will also add the option to sacrifice some fluidity and get better intermediate resolution instead.


Feb 13, 2017 08:16 PM
by RadiAnt
Has the MIP pixelation issue been fixed in the new beta version (4.5.1) that uses Nvidia GPU for rendering? Asking in order to know if I should buy a new graphics card....


Jan 27, 2018 02:05 PM
by Alex
I dont think MIP is using gpu acceleration.

I tried it with an HP Worsktation Laptop and nvidia quadro M1000M.
MIP pixelation is still present.


Jan 29, 2018 05:05 PM
by marcelvn
Please check our latest beta (4.5.5), there is a MIP preset in the 3D Viewer window. It is also GPU accellerated.


Feb 3, 2018 03:48 PM
by RadiAnt
@Radiant

Nice addition.

But the MIP preset is from the whole slice volume.

I believe MIP is most useful when used with an adjustable thick slice, just like the you implemented in the 3D MPR.


Feb 6, 2018 08:09 PM
by marcelvn
@Radiant,
Just as @marcelvn has stated, the issue is MIP pixelation in the 3D MPR with adjustable slice thickness. This specially critical for chest CTs.

Keep up the fantastic work!!


Feb 9, 2018 10:36 AM
by Alex
Yes, we will also add the GPU acceleration for MIPs in the 3D MPR window in one of the next beta versions.


Feb 20, 2018 05:25 PM
by RadiAnt
@Radiant

Great to hear that GPU acc is coming to the 3D VR.

MIP and also Average thick slices could benefit a lot from that optimization.


Feb 21, 2018 12:17 AM
by marcelvn
@Radiant,
Does the new beta fix de pixelation issue in MIP 3D MPR?


Mar 8, 2018 01:51 PM
by Alex
No, not yet. It will be implemented most probably in the next beta version.


Mar 9, 2018 05:35 PM
by RadiAnt
Post reply

Author (first name or pseudonym):


Message:


You have to accept cookies to be able to post messages.
Visitors can post messages on the forum without having to register. Please input your first name or pseudonym in the "Author" field. Do not publish any personally identifiable information (PII).

Please ensure you write your posts in English so that our international audience can understand and reply to your message.

The Administrator reserves the right to remove any messages that violate our forum rules; for example, those that are not related to the RadiAnt DICOM Viewer software, published in a language other than English, contravene regulations, etc.