- deform v0.2
- ANTs v2.3.1
- SimpleITK (e.g.
pip install SimpleITK) - Adult atlas (30 subjects, 95 regions) (https://brain-development.org/brain-atlases/adult-brain-atlases/)
First configure the environment by setting necessary environment variables.
export DEFORM_EVAL_DATA_PATH <path>
export DEFORM_EVAL_RESULTS_PATH <path>
export ANTS_BIN_PATH <path>
export DEFORM_BIN_PATH <path>
| Variable | Description |
|---|---|
DEFORM_EVAL_DATA_PATH |
Path for storing intermediate imaging data. |
DEFORM_EVAL_RESULTS_PATH |
Path for the outputted results |
ANTS_BIN_PATH |
Path to the ANTs binaries. |
DEFORM_BIN_PATH |
Path to the deform binaries. |
Generate intermediate imaging data by first extracting atlas retrieved from https://brain-development.org. Then run the setup script, providing the full path to the extracted Hammers_mith-n30r95 path:
python run.py setup <Hammers_mith-n30r95/>
To run all labeling experiments simply run
python run.py
nvprof can further assess the performance of deform. This requires deform to be compiled with profiling enabled. See scripts run_nvprof.py and parse_nvprof.py.
Results are stored in DEFORM_EVAL_RESULTS_PATH (results/ by default). Provided notebooks can be used to generate figures and tables.