python3 dicom_to_hdf5_json.py -i ../dicom_data -h ../hdf5_data/hdf5_data.hdf5 -j ../json_data/json_data.json
"../dicom_data" ------ path to input DICOM directory
"../hdf5_data/hdf5_data.hdf5" ------ path to output hdf5 file and filename
"../json_data/json_data.json" ------ path to output JSON file
python3 hdf5_to_dicom.py -h ../hdf5_data/hdf5_data.hdf5 -d ../template.dcm -o ../dicom_output/
"../hdf5_data/hdf5_data.hdf5" ------ path to input hdf5 file
"../template.dcm" ------ path to the template DICOM directory
"../dicom_output/" ------ path to output DICOM directory
At first, I implement the gaussian algorithm to blur a 2D image, and then extend to blur 3D volumne, pixel data computaion is based on hdf5 file.
python3 gaussian_blur3d_starter.py
python3 inference_pipeline_starter.py
I wrote a simple front end page can send Post and Get request through that textfield. Involve Interference Pipeline to execute background job task.
Run server at directory of Part3
./manage.py runserver
Can use postman or just type path in the post textfield, I use 3dblurr function as example in Inference Pipeline, it might take a long time to run this function at background.
Example: /job/?job_name=3dblur&in_dir=./part3_backend/input_volume/mytestfile.hdf5
Can type path in the get textfield or type in broswer url field.
Example: query/601