Importing micro-CT
In the context of my work I often use Mimics to import µ-CT datasets. These are generally large datasets of high quality, but once the images are loaded into Mimics the contrast is incorrectly displayed. This makes the segmentation very difficult or even impossible.
How can I make sure the images are displayed correctly?
Hi Akhilesh,
This does not seem correct as the file size isn't very large. A few things that you should verify when importing the images:
1. Where are the images being imported from? If they are from a network drive, it could result in a slower import if your connection is poor.
2. When selecting your images in the "New Project Wizard" dialogue, are you highlighting only the image files that you would like to import? You may be trying to import the parent folder containing all of the images as well as other files. This may cause a delay in the import depending on the total size of your files.
3. Make sure that you have have free memory on your system. If it is tied up by other running applications, then it could cause a delay in importing your images.
If this problem persists, please contact your Materialise representative and they will be able to assist.
I am trying to import micro-CT data that is in a *.bmp format. I keep getting an error when importing more then 250 slice that I am out of memory. Any suggestions?
As Geoff mentions, the image data you are trying to load, exceeds your RAM. Mimics stores all images in RAM to be able to work fast. Hence a simple calculation shows you how much free memory you need; if your images are 4MB each (which can happen with micro-CT images), loading 250 of them would take about 1GB of RAM.
A 32-bit operating system can only cope with 2GB of RAM; so if you work often with large datasets (micro-CT), you should consider a 64-bit computer and the 64-bit version of Mimics.
Also, I can advise you to keep an eye out for the next release of the Mimics Innovation Suite...! (scheduled at the end of April)
This is a common issue when importing very large datasets such as µCT in the 32-bit version of Mimics. All 32-bit applications have an inherent memory limit, which in this case appears to be lower than is needed to import the data. You may need to upgrade to the 64-bit version of Mimics or lower the complexity of your data by skipping images or decreasing image resolution.
I have more microCT data to work with.
Trying to manually import the data using signed short, the error message I keep getting is "The image value range must be between 0 and 32767." Even though I am using signed short as my input it still won't accept a negative for the "Minimum Image Value".
If I do enter 0 as the minimum value, then you can see from the histogram that the data is clipped.
Reading the help manual makes it appear that you should be able to enter a negative value.
I am running Mimics 13.1 under Windows XP Professional
Thanks,
Kirk
Indeed, it is not possible to load datasets with a negative value for minimum image value. This issue is known and will be solved in the following release of Mimics 14.1.
In your specific case, do they represent air or another structure? In any case, do not hesitate to contact your local Mimics support office and provide them with the dataset.
kind regards
I use microct as well and importing dicom images is the most effecient for me. I would suggest offering the ability when importing dicom images to either scale to HU range (default) or to scale to datatype range (option). There are also cases in clinical CT scanners where the HU range is adjusted, such as extended scale processing or other metal artifact reconstructions.
Thanks
The first thing you should check is if you used Lossless compression during the image import. When you use CT compression, 200 gray values are set to 0, which can remove some contrast from your dataset.
Frequently, µ-CT datasets are 16-bit images, which means that they have 65536 gray values (216). Because the generated images are DICOM files, Mimics can import them automatically. However, in this process, Mimics will convert these images in order to fit in the Hounsfield range - 4096 gray values (212), typical of 12-bit medical CT images. Therefore, the loaded images will present a low contrast, compared to the original images.
In order to avoid this effect, you can import the images manually. Select the Manual checkbox in the Import dialog and fill in the correct parameters (they should be available in a text file together with the images). Make sure you select Lossless compression. Your images should now present the complete range of gray values.
Hello All:
I am a new user to Mimics. I have a microCT data set of which i am trying to load only 5 slices(each slice is around 500Kb). However it takes a lot of time (around 15 minutes for only 5 slices) to read these slices. I am running a Mimics 14.12-64bit version on a Windows 7-64 bit with 4GB memory.
I have loaded the entire data set on other softwares like ITK-Snap with no problems.
Is there anything I am missing out in the process of importing DICOM images?
-Akhilesh