I am interested in preserving the series meta data for each stack image.
If I import just the .nd file then OMERO correctly picks up all the .stk images in the series and collates them to a single image. However the metadata is not complete. The series metadata only appears to refer to the last of the stack images to be imported.
If I import the .stk images then OMERO picks up all the files from the experiment (i.e. the other .stk files), collates them into a single image and shows the series metadata for that individual stack. This results in redundancy within OMERO since a complete image is created for each .stk and .nd file. Only the metadata is different (because it incompletely refers to a single stack image).
In either case this works when all the files are in a unique folder. If I have images from two experiments in the same folder then importing the .stk images collates them incorrectly. It appears that the grouping performed by BioFormats is identifying non-related files. I can see this using the command-line showinf BioFormats tools. Look at the following experiment with 2 channels, 2 time points and 2 experiments (one has 3 z-sections the second has 7 z-sections):
- Code: Select all
[ah403@tallinn /home/scratch/images/metamorph] % ls -l *.nd *.STK
-rw-r--r-- 1 ah403 ah403 372 2012-02-14 15:06 LJN_Y1136_13.nd
-rw-r--r-- 1 ah403 ah403 113443 2012-02-14 15:06 LJN_Y1136_13_w1DAPI_t1.STK
-rw-r--r-- 1 ah403 ah403 113442 2012-02-14 15:06 LJN_Y1136_13_w1DAPI_t2.STK
-rw-r--r-- 1 ah403 ah403 113443 2012-02-14 15:06 LJN_Y1136_13_w2FITC_t1.STK
-rw-r--r-- 1 ah403 ah403 113443 2012-02-14 15:06 LJN_Y1136_13_w2FITC_t2.STK
-rw-r--r-- 1 ah403 ah403 372 2012-02-14 15:06 LJN_Y1136_14.nd
-rw-r--r-- 1 ah403 ah403 263559 2012-02-14 15:06 LJN_Y1136_14_w1DAPI_t1.STK
-rw-r--r-- 1 ah403 ah403 263559 2012-02-14 15:06 LJN_Y1136_14_w1DAPI_t2.STK
-rw-r--r-- 1 ah403 ah403 263558 2012-02-14 15:06 LJN_Y1136_14_w2FITC_t1.STK
-rw-r--r-- 1 ah403 ah403 263558 2012-02-14 15:06 LJN_Y1136_14_w2FITC_t2.STK
[ah403@tallinn /home/scratch/images/metamorph] % showinf LJN_Y1136_13_w2FITC_t2.STK -nopix | grep -A 6 LJN_Y1136_13_w2FITC_t2.STK
Filename = LJN_Y1136_13_w2FITC_t2.STK
Used files:
/home/scratch/images/metamorph/LJN_Y1136_14.nd
/home/scratch/images/metamorph/LJN_Y1136_14_w1DAPI_t1.STK
/home/scratch/images/metamorph/LJN_Y1136_14_w2FITC_t1.STK
/home/scratch/images/metamorph/LJN_Y1136_14_w1DAPI_t2.STK
/home/scratch/images/metamorph/LJN_Y1136_14_w2FITC_t2.STK
As you can see showinf reads the LJN_Y1136_13_ file and associates it with all the LJN_Y1136_14_* files.
It would be ideal if the importer could handle multiple experiments within the same folder. It should read only the .nd files (ignoring any .stk files that will produce duplicate imports) and correctly associate the .stk files to the corresponding .nd files. However I am prepared to have to import each experiment from a different folder if this is required.
The reason we would like the metadata intact is that each channel and potentially time point has a different exposure time. We use this exposure time within our analysis. It is also vital to know when each stack was imaged from the DateTime attribute for our long time course imaging experiments.
It is clear that BioFormats can read all the data within the original metamorph files. However I cannot see how to make it all import into OMERO in one clean image file.
Can anyone help with this? I can provide test images if necessary. I am using OMERO 4.3.3 and Metamorph 7.7.6.0.
Thanks,
Alex
