mouse-imaging-centre / minclaplace Goto Github PK

Hi!

I want to use your great codes to deform one parameterized boundary surface (generated by SPHARM-PDM, its format is 'vtk') to its medial surface (already generated, and its format is also 'vtk') along the gradient of a laplacian field, and I am doing so to guarantee shape-inherent point correspondence on the medial surface across subjects, but I don't know how to deal with it.

Can you give me some advice about that? Thank you for your kind assistance!

Zeng

Documenting:

The root cause of the bug is that the if conditions in lines 61:78 in the cythonLaplaceStep function are sometimes not satisfied, and since counter starts at 0.0, and never increments, line tmpvalue = tmpvalue / counter results in a divide by zero error which crashes the program. For now I’ve wrapped the division in a check for if the counter is >0, but I need to check if perhaps there should be an else clause that would be more appropriate. The solution for the non-crashing left surface seem to be identical, so this looks okay.

Dear Dr Lerch, if you have an example usage file, could you please post it?
Thank you,
Sohrab

Hi, I am running MINC on a ubuntu (16.04) VM. Minc is installed (checked via mincinfo - returns Usage etc). Pyminc is also installed (locate pyminc gives location in '/lib/python2.7/dist-packages/pyminc'). Minclaplace is also installed correctly as command 'minclaplace' gives subsequent instruction for input. However, when I try to run 'minclaplace' with my template I get ' ***Error in 'usr/bin/python': double free or corruption (!prev)'. I think this may be a memory allocation problem, but I am unsure. I have also downloaded the VM from the cobralab github with MINC/PYMINC already installed. I have installed minclaplace on this also and get an error of 'malloc_consolidate(): invalid chunk size, aborted (core dumped)' when running the command with my template. Could you please advise on the amount of memory and storage required for these processes. If possible, is there any test images online, e.g. the template from the minclaplace/laplacegrid and representative csv file that I could download and use to compare with my images? (The attached file is the output from the ubuntu vm (not the pre-made MINC lubuntu setup from cobralab)). My template is consistent with the one outlined in the LaplaceGrid page (with central portion of the brain (non-cortex) being 0, the middle (or cortical column) having a voxel value of 5, the resistive portion = 20 and outside the brain being set to 10.

Thanks in advance! Kind regards,
Jolyon

minc_laplace_output (2).txt