Help with Array Scan file format (C01)


#1

I am trying to analyze some images acquired using Cellomics ArrayScan. The arrayscan allows you to download the images as 8-bit tiff, 16-bit tiff, or C01 (their own format). Cell profiler works very well with the 8-bit tiff, but the resolution is not good enough. The 16-bit tiff should have good resolution, but the object detection algorithm does not work with this format. At last, I tried to load in the images in the original format (C01) in CellProfiler, but got error message.

I am wondering if there is anyway to make the cell profiler work with the cellomics arrayscan images? Any advice is highly appreciated! Thank you!


#2

Hi,

Oftentimes, many microscope cameras capture 12-bit images, which contain finer detail (in terms of graylevels) than 8-bit images. However, 12-bit file formats are rare and incompatible with most software, so the two options are usually to save the image in a 16-bit format or an 8-bit format. Saving 12-bit image data in a 16-bit file format is preferable to saving it in an 8-bit format because converting to 8-bit format causes some of the fine detail in the image data to be lost. Most image-viewing software on your computer will display 12-bit images very dark when they are stored in 16-bit image formats (or they might fail to recognize/open 16-bit format at all), but all the information in the image is actually there.

To rectify this, try using the RescaleIntensity module on the original image, with the rescaling method set to “Enter min/max” and the settings of 0, 0.0625, 0, 1, 0, 1 for the next six boxes (if you’re using latest version downloaded from our website). This will convert a 16-bit image containing only 12 bits of data to the proper range.

Hope this helps!
-Mark


#3

Hi Mark,

Thanks a lot, this is very helpful! Once I do the intensity conversion, would I be able to use the primary and secondary object detection algorithm of CellProfiler to analyze the images? Thank you!

Peng


#4

Hi Peng,

Yes, definitely, you will be able to use the identification modules exactly as before.

Regards,
-Mark


#5

But rescale intensity module causes changes in original image that is visible once you see histogram of rescaled image and original image and that will change your image analysis results. I am also using arrayscan images and its till doable without rescaling the images. I would suggest do export images from Arrayscan in 16 bit tiff format and till you can do image analysis. I am sure about it.
Enjoy
A


#6

Can you tell us what options you are using in RescaleIntensity to cause this to occur? For “Enter min/max” and the parameters I specified above, it should change the scaling, but the intensity distribution should remain exactly the same…

Regards,
-Mark


#7

Hi Mark
This is what I noticed.
I took image from Orca ER II camera of hamamastu (same as arrayscan) and rescaled with the parameters as you suggested. Saved image in tiff 16 bit format. I opened both the images using NIH Image J and I found that though the image is rescaled its causing discontinuous histogram that means some kind of artifact to me (may be I am not correct) and thats the reason why I avoided rescaling function and did image analysis without that.
I was unable to attach images from this forum menu otherwise I would have done it for your reference.
Can you let me know how can I attach images?
Thanks and let me know if you have something to share on.
Regards
A


#8

Re: Posting attachments - When you post a reply, do you see two tabs below the “Post a reply” edit box, one that says “Options” and another that says “Upload attachment”?

If so, click the attachment tab, and follow the instructions from there. If not, let us know and we’ll see if it’s a problem on our end.

Regards,
-Mark


#9

Dear Mark,
Original Arrayscan format C01 is some special compression done by Cellomics so we can not open these images in cellprofiler (which is not good as we have to export always images in 16 bit tif format in order to process it by CellProfiler and creats another step and data accumulation but hopefully Cellomics will open format one day and we can process it!).
If we try to rescale images as per your suggestions what we get is following. I have put both the files in attachement and seems rescaling causes some changes in file which you can assess if you try to see histograms). I have attached pipeline also by which I did it. Please let me know if I am doing some mistake!
Best,
Anil
rescalingPIPE.mat (754 Bytes)





#10

Hi Anil,

After looking at the images (thanks for sending, BTW), it turns out that rescaling the images isn’t the problem, saving them is.

RescaleIntensity works just fine, and any object identification/measurements from those imageswill be OK. But when saving the images as a 16-bit TIF, the SaveImages code was doing so in a way that caused information to be lost. This is a bug in SaveImages which will be fixed for the next release.

Regards,
-Mark