lodurality / 35_methods_miccai_2017 Goto Github PK

arXiv preprint

data (coming) -- sample data for running scripts. Full data will be available at IGC website.

results -- folder with full original results and outputs of scripts. raw_results -- data computed in preparation of the paper, script_results -- outputs of sample scripts

*scripts (coming) * -- scripts we used to calculate pairwise and gender classification and ICC.

pics -- some of the graphs.

pdf -- paper pdf.

Scripts currently requires Python 3.4 or later to run. Please install Python and pip via the package manager of your operating system if it is not included already. Also you need R language:

$ apt-get install r-base

And you need psych package. To install it run following in R shell:

> install.packages('psych', dependencies=TRUE, repos='http://cran.us.r-project.org')

To install needed Python dependencies run next in terminal:

$ pip3 install -r https://github.com/lodurality/35_methods_MICCAI_2017/blob/master/requirements.txt

There are three main scripts in scripts directory:

• calculate_icc.py
• classificate_gender.py
• classificate_pairwise.py

An argument for these scripts is a string of format:

$ ./classificate_pairwise.py dataset-normalization-tractography-reconstruction_model-atlas_type

For example:

$ ./classificate_pairwise.py BNU_1-max-deter-csd-con_aparc68+subcort

All calculated combinations you can find in combinations_for_gender_classification_and_icc.txt and `combinations_for_pairwise_classification.txt'. More information about parameters you can find in the article.

To test calculated results you can use test_results.py script. Just run it and choose directory you want to check:

$ ./test_results.py

0 /gender_results/
1 /icc_results/BNU_1/
2 /icc_results/HNU_1/
3 /icc_results/IPCAS_1/
4 /pairwise_results/

Please choose what directories to check:
$ 0
/gender_results/


	1/327 are similar
	0 ain't similar

Here we calculated results for one combination and script said us the result the same as we used for the article.