When I try running edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToads from command line, it stops after few steps:

Bedova_T1_brain_clone: Processing image ...
Bedova_T1_brain_clone: Initialization
Initialization
Atlas alignment

And then it shows this:

Java heap space
Module Failed in an Ungraceful Manner. Please contact the developer with the following message.
edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToadsRuntime Exception:null
edu.jhmi.rad.medic.methods.LesionToadSegmentation.initCentroids(LesionToadSegmentation.java:602)
edu.jhmi.rad.medic.algorithms.AlgorithmLesionToads.calcSegmentation(AlgorithmLesionToads.java:476)
edu.jhmi.rad.medic.algorithms.AlgorithmLesionToads.runAlgorithm(AlgorithmLesionToads.java:336)
edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToads$AlgorithmLesionToadsWrapper.runAlgorithm(MedicAlgorithmLesionToads.java:578)
edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToads$LesionToadsWrapper.execute(MedicAlgorithmLesionToads.java:462)
edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToads.execute(MedicAlgorithmLesionToads.java:605)
edu.jhu.ece.iacl.jist.pipeline.ProcessingAlgorithm.runAlgorithm(ProcessingAlgorithm.java:337)
edu.jhu.ece.iacl.jist.pipeline.ProcessingAlgorithm.runAlgorithm(ProcessingAlgorithm.java:321)
edu.jhu.ece.iacl.jist.cli.run.main(run.java:100)
edu.jhu.ece.iacl.jist.pipeline.CalculationMonitorCalculation monitor join failed.
edu.jhu.ece.iacl.jist.pipeline.CalculationMonitorCalculation monitor shutting down by interrupt.

I found similar issue here http://www.nitrc.org/tracker/?func=detai... so I guess it is also a problem with missing tag. At first I thought it is the -outExecution missing, but I got no idea what value it should have. Then I downloaded the newest version of JIST (JIST-CRUISE-2015Jul14-10-36PM.jar and MIPAV 7.2), but the problem stays the same (but now I guess the tag -outExecution is not recognized).

Great news! We are happy to help. Thank you for using the support forums.

Best wishes,
Bennett

Thanks a milion, it was indeed the heap size. I guess it is trivial for you, but it did not occur to me. I was not able to run that command you recommended (what should come after -classpath flag), but I made a workaround with "export _JAVA_OPTIONS=-Xmx5G".

From the team:


Not much experience with running LesionTOADS from the command line. I
would check the obvious: heap size, atlas files exist and are
readable.

Seeing that it makes it to "Atlas alignment", could be a null return
from a FLIRT registration. Not sure if TOADS uses this or not any
more, but we have definitely had null matrices returned from FLIRT
Registrations.

�

I think the heap size is the first thing to check.
The memory to use should be added into the command line as follow
/mipav/jre/bin/java -Xms5G -XX:+UseSerialGC -Xmx5G -classpath ...