Hive_convert_semantic_to_instance_segmentation script

Script to convert a semantic segmentation dataset (with the Patient ID as the folder name) into an instance segmentation dataset. Instance segmentation masks are saved within the same patient folder with the standard format “INST_SEG.nii.gz”. Regions in instance segmentation containing less than 10 voxels are ignored and the number of labels in each instance segmentation mask is saved in a separate json file (‘inst_seg_labels.json’) alongside its ‘Patient ID’.

usage: Hive_convert_semantic_to_instance_segmentation [-h] --data-folder
                                                      DATA_FOLDER
                                                      --sem-seg-suffix
                                                      SEM_SEG_SUFFIX
                                                      --inst-seg-suffix
                                                      INST_SEG_SUFFIX
                                                      --output-json-path
                                                      OUTPUT_JSON_PATH
                                                      [-v | -q]

Named Arguments

--data-folder

AutoPET patient dataset folder.

--sem-seg-suffix

Semantic Segmentation suffix.

--inst-seg-suffix

Instance Segmentation suffix.

--output-json-path

Output path of json file.

-v, --verbose

Enable verbose output in terminal. Add multiple times to increase verbosity.

-q, --silent

Suppress most log outputs in terminal.

Example call:

Hive_convert_semantic_to_instance_segmentation  --data-folder /PATH/TO/SEMANTIC_SEG_DATA --sem-seg-suffix _SEG.nii.gz --inst-seg-suffix _INST_SEG.nii.gz --output-json-path /PATH/TO/JSON/inst_seg_labels.json