CellProfiler help! can't open image file or extract metadata to begin to run pipleline

Hello everyone, feeling stupid here.

Trying to learn cellprofiler to set up cell tracking for my timelapse videos but the software can’t ‘recognize’ my image files.

Some background here:
I acquire my timelapse images using two separate microscope systems:
(1) The first is on metamorph software which saves each timepoint as a TIF file and the metadata for the entire series as a .nd file.
(2) The second is Nikon NIS elements software which saves all data into a .nd file with no additional or separate files.

I am able to open both image files through Imaris or ImageJ to play as videos or analysis. However, as I have quite a lot of these files at various timepoints and different field of views, I want to start using Cellprofiler to automatically track objects over time.

So far, I have performed the following with no success:
(a) drag and drop images from either acquisition softwares into cellprofiler
(b) Opened .nd image file from NIS elements in Imaris and resave it as either .oem, .tif, .tif (adjustable file series) and then dragged these into cellprofiler
© dragged image files from metamorph (all tif + .nd, only tif, only .nd) and opened .nd as metadata file.

None of these work in getting cellprofiler to recognize/open these files.

Anyone have experience and can help?
Thanks very much!

Hi Raven

I believe you need to have each frame of the video as individual images with the time point in the file name. You can then extract the time point in the Metadata module. If you open your stack in ImageJ and choose Save As… Image Sequence… then you should get your individual images with correctly named filenames (eg xxx_t001_z001_c001) with t=time, z=z-stack position, c=channel.

/Esben

I don’t think CellProfiler can handle spaces in filenames, could you try to rename it so it contains underscores instead of spaces? It is probably easier to rename the parent file and then save as a sequence.

/Esben

Hi,

Spaces in the file names aren’t an issue; are you using the ‘Metadata-> extract from image file header’ option to pull out the Metadata before you move onto NamesAndTypes and the rest of your pipeline? That should be enough to solve your issue.

Hi Esben and Beth,

Thank you both very much for your help. I figured it out by trying different combinations to get it to work.

Essentially, I open the image files in Imaris, re-save them into .tif files and open these in ImageJ (there’s a reason why I did this instead of directly opening in ImageJ). I then re-save the images from ImageJ to Image sequences while also renaming them to a format title_time_channel. This enabled me to get CellProfiler to recognize the various files, pull the metadata out of it using the regular expressions (which was a little difficult to learn) and subsequently set up NamesAndTypes to do a simple pipeline to count the number of objects in a static single field of view.

Now that I’ve got this far, in the future, I will be attempting to do some tracking of my timelapse videos. I have now downloaded ‘Tracer’ but I haven’t exactly figured out the pipeline to get CellProfiler to track my cells over time. Also, I am not sure if the software allows me to obtain velocity, meadering index, stationary cells, etc from tracking objects. I will post an enquiry in the ‘Image Analysis’ section next week as I get more familiar with the software and progress towards that.

Thanks again!