Erode block error

[Tommaso Lucchesi]

Hello,

In order to analyze the surroundings of the nuclei of cells in my histological sections, i am trying to obtain a "rim" around the nucleus using the following procedure. Part of the procedure i am currently trying to develop implies to:

1/ segment the nuclei by applying "HK-means" block on the nuclear signal channel of confocal microscopy images of histological sections. Note that those are hyperstacks, so the output is 3D ROIs. 

2/ use the "convexify" block on the list of ROIs obtained by HK-means segmentation. 

3/ use "Erode ROI" block on the list of convexified ROIs.

As the Erode ROI block starts however, i get the following error:

"An error occured while plugin 'Protocols SDK' was running :

java.lang.ArrayIndexOutOfBoundsException: -17

at plugins.kernel.roi.roi2d.ROI2DArea.setPoint(ROI2DArea.java:1415)

at plugins.kernel.roi.roi3d.ROI3DArea.setPoint(ROI3DArea.java:527)

at plugins.kernel.roi.roi3d.ROI3DArea.removePoint(ROI3DArea.java:517)

at plugins.adufour.blocks.tools.roi.ErodeROI.erodeROI(ErodeROI.java:204)

at plugins.adufour.blocks.tools.roi.ErodeROI.erodeROI(ErodeROI.java:62)

at plugins.adufour.blocks.tools.roi.ErodeROI.run(ErodeROI.java:30)

at plugins.adufour.blocks.lang.BlockDescriptor.run(BlockDescriptor.java:540)

at plugins.adufour.blocks.lang.WorkFlow.run(WorkFlow.java:833)

at plugins.adufour.blocks.lang.BlockDescriptor.run(BlockDescriptor.java:540)

at java.lang.Thread.run(Thread.java:695)"

I already sent a report about this error, but i thought that creating a thread on the forum to give a context to the problem might help solving it. 

Best,

Tommaso

[Tommaso Lucchesi]

Hi everyone,

Any ideas to solve this? I may provide the protocol on Icy and an example input sequence if it can help. 

Thanks,

Tommaso

[JF]

I had the same problem (and same error message) with another protocol using the ROI Erode block.
In fact my protocol works all right running on a 2D image, but produces an error with a 3D image.

Jean-François

Le mardi 27 septembre 2016 17:31:32 UTC+2, Tommaso Lucchesi a écrit :

[Stephane]

Hi Tommaso,

Indeed it would be really useful for us to be able to reproduce the problem, so having an input minimum sample example would help a lot :) Thanks !

- Stephane