Spindle: Versions

[tischi]

Hi All,

Since the version number of ../jars/fiji-plugins-morphometry-*.jar might also change because of reasons that are not related to the spindle morphometry measurement plugin, I decided to put a version number into the plugin itself (lower right corner):

[tischi]

Versions previous to 0.5.5

These versions do not have a version number. But we could keep a list here. Could you please post here the version numbers of the ../jars/fiji-plugins-morphometry-*.jar that you were previously using?

[tischi]

Version 0.5.5

Major changes

• The threshold at which the DNA is found initially has been increased. The hope is that this diminishes connections to nuclei of other cells that are close by.
• The segmentation of the spindle has been spatially limited to a region around the DNA of the cell. The hope is that this diminishes merging with tubular signal of cells close by

Relation to jars

Corresponds to jars starting from version number fiji-plugin-morphometry-0.4.16.jar

[TobiasKletter]

Versions previous to 0.5.5

These versions do not have a version number. But we could keep a list here. Could you please post here the version numbers of the ../jars/fiji-plugins-morphometry-*.jar that you were previously using?

@tischi Currently using: fiji-plugin-morphometry-0.4.15.jar

[tischi]

Are results from fiji-plugin-morphometry-0.4.15.jar already in the graphs (Zwischenergebnisse.pdf) that Nils was posting? If so, how is it called there?

[n-dempe]

Are results from fiji-plugin-morphometry-0.4.15.jar already in the graphs (Zwischenergebnisse.pdf) that Nils was posting? If so, how is it called there?

yes, for Spindle Width fiji-plugin-morphometry-0.4.15.jar is called "new plugin" in the graphs for Spindle Length, DNA longest axis there were no differences between between 0.4.15 and the previous version.

[TobiasKletter]

A little feedback from the analysis of a dataset with rather suboptimal input images:

For reasons, the DNA signal in this dataset was mostly really weak and I anticipated some problems because of that.

The current version (0.5.5) and the version before performed differently:

Plugin Version Pre-0.5.5: total n = 826 n correct (as judged by pole placement) = 139 n incorrect = 606 n "not processed" (java.lang.IndexOutOfBoundsException or java.lang.NullPointerException) = 81

Plugin Version 0.5.5: total n = 826 n correct (as judged by pole placement) = 360 n incorrect = 276 n "not processed" (java.lang.IndexOutOfBoundsException or java.lang.NullPointerException) = 190

Here are some example files: Throws exception only in new version https://www.dropbox.com/s/thy8mtfkix5vuqn/Throws_Exception_Version_055_only.tif?dl=0

Throws exception in both versions https://www.dropbox.com/s/x7omhzw7dwgl3uw/Throws_Exception_Version_pre055_and_055.tif?dl=0

Output images from each version of the same input image https://www.dropbox.com/s/bql0pymjgtlbsg4/Version_pre055-out.tif?dl=0 https://www.dropbox.com/s/02si1lqj9jsca9s/Version_055-out.tif?dl=0 (input image https://www.dropbox.com/s/98dkn8lx926ewik/HighZoom--W0000--P0001-T0030.tif?dl=0)

(In the 0.5.5 output, the channel containing the tubulin mask doesn't seem to be aligned with the other channels?)

Please let me know if you'd be interested in seeing other examples from this set or the full exception messages!?

[tischi]

Thanks!

1. I will try to fix the exceptions!
2. I will see why the tubular mask is not aligned!

[tischi]

The dynamic range is really super low:

I could try to make it work, but, as an alternative: Would it be an option if the plugin would throw some meaningful error there? What would you think?

[tischi]

...another challenge here is that the coverslip seems to be relatively bright in the DNA channel Do you know why that is? Do you maybe have a reflection there?

[tischi]

same issue with the other image: very low dynamic range in DNA and, I think, a reflection signal from the coverslip (which microscope was that?):

[TobiasKletter]

Yes, I think having an error message would be totally fine - I wouldn't want to use these images for quantification anyways because of the low signal-to-noise, so I don't think the plugin needs to be able to cope with these.

Good that you pointed out the reflection issue. I thought we solved it but it seems to appear again when the dynamic range is so low. This is all data from the LSM800.

[tischi]

ok! makes sense. I will implement a proper error message for such cases.

What shall we do in terms of output? Should it simply produce no output files?

[TobiasKletter]

Maybe have the measurements.txt file but with "NaN" for every column? Then we can easily check for these cases in the concatenated table...

[TobiasKletter]

...except the PathInputImage, that's really useful of course.

[tischi]

That makes sense! In addition, there could be a column in the table, called "Comments", and maybe there could be some text, such as "Too low dynamic range in DNA". And in images, where everything went fine we say: "No comment". What do you think?

[TobiasKletter]

Sounds good!

[tischi]

Version 0.5.6

Major changes

• Added a "Comment" column to the output table, which informs, e.g., whether the analysis finished successfully.
• Added a "Minimum dynamic range" to the settings. If the threshold for either the DNA or spindle is below that value the analysis will be interrupted. Values in the table will be NaN and there will be an appropriate comment.
• Changed the spindle threshold algorithm: Now using intensity values close to the later DNA periphery as thresholds. This seems to yield lower threshold values and thus more compact spindles during the segmentation.

Relation to jars

Corresponds to jars starting from version number fiji-plugin-morphometry-0.4.18.jar

@TobiasKletter @n-dempe Could you please test it? (Update Fiji to get it)

[TobiasKletter]

[tischi]

Version 0.5.7

Major changes

• Spindle mask is now correctly saved into output image
• Input image can have more than 2 channels
• Output image format is .zip to save disk space
• Added version to output table

Relation to jars

Corresponds to jars starting from version number fiji-plugin-morphometry-0.4.19.jar

@TobiasKletter @n-dempe Could you please test it? (Update Fiji to get it)

[n-dempe]

Just to clarify :

If 0.4.19.jar. is version 0.5.7 then 0.4.15.jar is which version?

[TobiasKletter]

@n-dempe I think 0.4.15.jar should correspond to version 0.5.4

[TobiasKletter]

@tischi https://www.dropbox.com/sh/wmltcno4312rd69/AACJfsQiwnj9VE4ZL5cRwCiga?dl=0&lst= The two last ones in this folder are throwing exceptions and we can't figure out why... Signal-to-noise in DNA channel should be fine, no? Or are the bright non-mitotic DNA signals disrupting something?

[tischi]

Version 0.5.8

Major changes

• Improved Spindle threshold
• Improved Spindle length measurement

Relation to jars

Corresponds to jars starting from version number fiji-plugin-morphometry-0.5.0.jar

[TobiasKletter]

@tischi

Hi! Since the latest update, I can no longer use the batch processing function, and the following message pops up: Same happened on a different computer after updating. Re-installing Fiji didn't help so I figured it's on the plugin end?

[tischi]

I tried this with other Commands and it seems to me not specific to our Spindle Morphometry.

1. Could you please try if that also happens with a fresh Fiji for some other plugin (before you add my update site)?
2. If yes, could you please post this issue on www.image.sc?

[TobiasKletter]

Sadly no reply yet on image.sc ... For the time being, I tried to call it within an Imagej macro. Would that work, in theory? My code did not (it opens the GUI and wants me to click "ok"), but I might be just missing something:

parent_Dir = "/Path/To/Parent/Directory/";
in_Dir = parent_Dir + "InputFolder/";
out_Dir = parent_Dir + "OutputFolder/";
Voxel_size = 0.25;
Dynamic_range = 7;
DNA_Channel = 2;
Spindle_Channel = 1;
Intermediate_Results = false;  

fileList = getFileList(in_Dir);

setBatchMode(true);
for (i = 0; i < fileList.length; i++) 
    {
    fileName = in_Dir + fileList[i];
    print("Processing: " + fileName);
    import_Args = "Input Image File=" + fileName + "Input Image Files Parent Directory=" + parent_Dir + "Output Directory=" + out_Dir + "Voxel Size for Analysis=" + Voxel_size + "Minimum dynamic range [segmentation threshold gray value]=" + Dynamic_range + "DNA Channel [one-based index]=" + DNA_Channel + "Spindle Channel [one-based index]=" + Spindle_Channel + "Show Intermediate Results=" + Intermediate_Results; 
    run("Spindle Morphometry", import_Args);
    }
setBatchMode(false);
print("Finished analysis.")

Also, it appears that there is some hick-up anyways. When executing the plugin (in non-batch mode...), it throws the following error:

## Spindle Morphometry Measurement
Processing file HighZoom--W0000--P0001-T0002.tif
Create isotropic images...
DNA initial threshold: 49.0
Remove small regions...
Determining DNA axes...
[ERROR] Module threw error
java.lang.NoClassDefFoundError: mcib3d/image3d/ImageInt
    at de.embl.cba.morphometry.spindle.SpindleMorphometry.determineDnaAxes(SpindleMorphometry.java:755)
    at de.embl.cba.morphometry.spindle.SpindleMorphometry.measure(SpindleMorphometry.java:134)
    at de.embl.cba.morphometry.spindle.SpindleMorphometry.run(SpindleMorphometry.java:102)
    at de.embl.cba.morphometry.commands.SpindleMorphometryCommand.processFile(SpindleMorphometryCommand.java:120)
    at de.embl.cba.morphometry.commands.SpindleMorphometryCommand.run(SpindleMorphometryCommand.java:77)
    at org.scijava.command.CommandModule.run(CommandModule.java:199)
    at org.scijava.module.ModuleRunner.run(ModuleRunner.java:168)
    at org.scijava.module.ModuleRunner.call(ModuleRunner.java:127)
    at org.scijava.module.ModuleRunner.call(ModuleRunner.java:66)
    at org.scijava.thread.DefaultThreadService.lambda$wrap$2(DefaultThreadService.java:228)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.ClassNotFoundException: mcib3d.image3d.ImageInt
    at java.net.URLClassLoader.findClass(URLClassLoader.java:382)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
    at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:349)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:357)
    ... 14 more

[tischi]

java.lang.NoClassDefFoundError: mcib3d/image3d/ImageInt

you need to also install the update site: 3D Image Suite

[TobiasKletter]

Aaah, I wasn't aware! Thanks, works perfectly now.

[TobiasKletter]

Hi @tischi since the last update, we get the following error:

(Fiji Is Just) ImageJ 2.0.0-rc-69/1.52p; Java 1.8.0_202 [64-bit]; Mac OS X 10.14.5; 201MB of 4670MB (4%)

java.lang.RuntimeException: java.util.concurrent.ExecutionException: java.lang.NullPointerException
    at net.imagej.legacy.LegacyService.runLegacyCompatibleCommand(LegacyService.java:309)
    at net.imagej.legacy.DefaultLegacyHooks.interceptRunPlugIn(DefaultLegacyHooks.java:163)
    at ij.IJ.runPlugIn(IJ.java)
    at ij.Executer.runCommand(Executer.java:137)
    at ij.Executer.run(Executer.java:66)
    at java.lang.Thread.run(Thread.java:748)
Caused by: java.util.concurrent.ExecutionException: java.lang.NullPointerException
    at java.util.concurrent.FutureTask.report(FutureTask.java:122)
    at java.util.concurrent.FutureTask.get(FutureTask.java:192)
    at net.imagej.legacy.LegacyService.runLegacyCompatibleCommand(LegacyService.java:305)
    ... 5 more
Caused by: java.lang.NullPointerException
    at de.embl.cba.morphometry.commands.SpindleMorphometryCommand.processFile(SpindleMorphometryCommand.java:159)
    at de.embl.cba.morphometry.commands.SpindleMorphometryCommand.run(SpindleMorphometryCommand.java:94)
    at org.scijava.command.CommandModule.run(CommandModule.java:199)
    at org.scijava.module.ModuleRunner.run(ModuleRunner.java:168)
    at org.scijava.module.ModuleRunner.call(ModuleRunner.java:127)
    at org.scijava.module.ModuleRunner.call(ModuleRunner.java:66)
    at org.scijava.thread.DefaultThreadService.lambda$wrap$2(DefaultThreadService.java:228)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    ... 1 more

Any idea?

[tischi]

I do not have an immediate idea. Are there outputs in the IJ log Window?

[tischi]

I found the error. I introduced an error on purpose just to stop the code while trying to develop an improved method to find the cell center in images with more than one cell. I fixed it and uploaded to Fiji.

[TobiasKletter]

Works again, thanks!

[tischi]

@TobiasKletter

There is a new version!

Spindle Morphometry Version 0.5.9

Major changes

• Renamed: Spindle_CoV to Spindle_Tubulin_CoV
• Added: Spindle_Tubulin_Sum
  • Measures an offset corrected sum intensity of all pixels within the spindle mask. As offset we are using Spindle_Intensity_Threshold
• Bug fix: The coordinates of derivative plots were not exactly identical to the coordinates in the parent plot. This could lead to crashed when the code was trying to find back, e.g. the coordinate of a maximal derivative in the parent plot, because it could be non-existent. This is fixed now.

Some code snippets

Spindle CoV (Coefficient of Variation)

final ArrayList< Double > intensitiesWithMask =
        getValuesWithinMaskAsList( intensities, mask );

final double mean = Utils.mean( intensitiesWithMask );
final double sdev = Utils.sdev( intensitiesWithMask, mean );

double cov = sdev / ( mean - meanOffset );

with meanOffset here being the Spindle_Intensity_Threshold.

Jars

Jars with a version number >= 0.5.3

[tischi]

@TobiasKletter

Could you please:

1. Download a fresh Fiji
2. Download, unzip, and put those two files into the jars folder: https://oc.embl.de/index.php/s/SvKMxKttl7nhbDO

You should see at the very bottom of the plugins menu:

[TobiasKletter]

Works nicely 👍 I will try to quality control how the new DNA segmentation and the spindle angles perform! Anything else I should test?

[tischi]

Just test everything and make issues, but now here: https://github.com/tischi/spindle3d/issues I will be on holidays for 3 weeks from Monday on. Tomorrow I could still do some fixes.