MonaiLabel vertebrae segmentation sample-app doesn't work for sample data

After clicking Run Auto Segmentation, the segments of the segmentationNode are still empty.

Another CT I tried only segmented half a vertebrae.

If I try segmentation_vertebrae model instead the monai-label server crashes

Here is the console output for localization_spine

~$ monailabel start_server --app /home/user1/Escritorio/Monai/apps/radiology --studies /home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr --conf models "localization_spine,localization_vertebra,segmentation_vertebra"
Using PYTHONPATH=/home/user1:

2023-01-14 17:57:49,144 - USING:: version = False
2023-01-14 17:57:49,144 - USING:: app = /home/user1/Escritorio/Monai/apps/radiology
2023-01-14 17:57:49,144 - USING:: studies = /home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr
2023-01-14 17:57:49,144 - USING:: verbose = INFO
2023-01-14 17:57:49,144 - USING:: conf = [['models', 'localization_spine,localization_vertebra,segmentation_vertebra']]
2023-01-14 17:57:49,144 - USING:: host = 0.0.0.0
2023-01-14 17:57:49,144 - USING:: port = 8000
2023-01-14 17:57:49,144 - USING:: uvicorn_app = monailabel.app:app
2023-01-14 17:57:49,144 - USING:: ssl_keyfile = None
2023-01-14 17:57:49,144 - USING:: ssl_certfile = None
2023-01-14 17:57:49,144 - USING:: ssl_keyfile_password = None
2023-01-14 17:57:49,144 - USING:: ssl_ca_certs = None
2023-01-14 17:57:49,144 - USING:: workers = None
2023-01-14 17:57:49,144 - USING:: limit_concurrency = None
2023-01-14 17:57:49,144 - USING:: access_log = False
2023-01-14 17:57:49,144 - USING:: log_config = None
2023-01-14 17:57:49,144 - USING:: dryrun = False
2023-01-14 17:57:49,144 - USING:: action = start_server
2023-01-14 17:57:49,144 - ENV SETTINGS:: MONAI_LABEL_API_STR = 
2023-01-14 17:57:49,144 - ENV SETTINGS:: MONAI_LABEL_PROJECT_NAME = MONAILabel
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_APP_DIR = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_STUDIES = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_AUTH_ENABLE = False
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_AUTH_DB = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_APP_CONF = '{}'
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TASKS_TRAIN = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TASKS_STRATEGY = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TASKS_SCORING = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TASKS_BATCH_INFER = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_URL = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_USERNAME = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_PASSWORD = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_API_KEY = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_CACHE_PATH = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_PROJECT = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_ASSET_PATH = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_DSA_ANNOTATION_GROUPS = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_USERNAME = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_PASSWORD = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CACHE_PATH = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_QIDO_PREFIX = None
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_WADO_PREFIX = None
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_STOW_PREFIX = None
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_FETCH_BY_FRAME = False
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CONVERT_TO_NIFTI = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_SEARCH_FILTER = '{"Modality": "CT"}'
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CACHE_EXPIRY = 180
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_PROXY_TIMEOUT = 30.0
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_READ_TIMEOUT = 5.0
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_AUTO_RELOAD = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_READ_ONLY = False
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_FILE_EXT = '["*.nii.gz", "*.nii", "*.nrrd", "*.jpg", "*.png", "*.tif", "*.svs", "*.xml"]'
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_SERVER_PORT = 8000
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_CORS_ORIGINS = '[]'
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_SESSIONS = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_SESSION_PATH = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_SESSION_EXPIRY = 3600
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_INFER_CONCURRENCY = -1
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_INFER_TIMEOUT = 600
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TRACKING_ENABLED = True
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_TRACKING_URI = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_ZOO_SOURCE = github
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_ZOO_REPO = Project-MONAI/model-zoo/hosting_storage_v1
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_ZOO_AUTH_TOKEN = 
2023-01-14 17:57:49,145 - ENV SETTINGS:: MONAI_LABEL_AUTO_UPDATE_SCORING = True
2023-01-14 17:57:49,145 - 
Allow Origins: ['*']
[2023-01-14 17:57:49,432] [850291] [MainThread] [INFO] (uvicorn.error:75) - Started server process [850291]
[2023-01-14 17:57:49,432] [850291] [MainThread] [INFO] (uvicorn.error:45) - Waiting for application startup.
[2023-01-14 17:57:49,432] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.app:38) - Initializing App from: /home/user1/Escritorio/Monai/apps/radiology; studies: /home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr; conf: {'models': 'localization_spine,localization_vertebra,segmentation_vertebra'}
[2023-01-14 17:57:49,455] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for MONAILabelApp Found: <class 'main.MyApp'>
[2023-01-14 17:57:49,458] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepgrow_3d.Deepgrow3D'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepedit.DeepEdit'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.localization_vertebra.LocalizationVertebra'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation_spleen.SegmentationSpleen'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepgrow_2d.Deepgrow2D'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation_vertebra.SegmentationVertebra'>
[2023-01-14 17:57:49,459] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation.Segmentation'>
[2023-01-14 17:57:49,460] [850291] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.localization_spine.LocalizationSpine'>
[2023-01-14 17:57:49,460] [850291] [MainThread] [INFO] (main:93) - +++ Adding Model: localization_spine => lib.configs.localization_spine.LocalizationSpine
[2023-01-14 17:57:49,564] [850291] [MainThread] [INFO] (main:93) - +++ Adding Model: localization_vertebra => lib.configs.localization_vertebra.LocalizationVertebra
[2023-01-14 17:57:49,648] [850291] [MainThread] [INFO] (main:93) - +++ Adding Model: segmentation_vertebra => lib.configs.segmentation_vertebra.SegmentationVertebra
[2023-01-14 17:57:49,733] [850291] [MainThread] [INFO] (main:96) - +++ Using Models: ['localization_spine', 'localization_vertebra', 'segmentation_vertebra']
[2023-01-14 17:57:49,733] [850291] [MainThread] [INFO] (monailabel.interfaces.app:135) - Init Datastore for: /home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr
[2023-01-14 17:57:49,733] [850291] [MainThread] [INFO] (monailabel.datastore.local:129) - Auto Reload: True; Extensions: ['*.nii.gz', '*.nii', '*.nrrd', '*.jpg', '*.png', '*.tif', '*.svs', '*.xml']
[2023-01-14 17:57:49,734] [850291] [MainThread] [INFO] (monailabel.datastore.local:593) - Adding New Image: CTChest => CTChest.nii.gz
[2023-01-14 17:57:49,734] [850291] [MainThread] [INFO] (monailabel.datastore.local:576) - Invalidate count: 1
[2023-01-14 17:57:49,735] [850291] [MainThread] [INFO] (monailabel.datastore.local:150) - Start observing external modifications on datastore (AUTO RELOAD)
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:126) - +++ Adding Inferer:: localization_spine => <lib.infers.localization_spine.LocalizationSpine object at 0x7fb1700a5f30>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:126) - +++ Adding Inferer:: localization_vertebra => <lib.infers.localization_vertebra.LocalizationVertebra object at 0x7fb16cb0d480>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:126) - +++ Adding Inferer:: segmentation_vertebra => <lib.infers.segmentation_vertebra.SegmentationVertebra object at 0x7fb16cb0d4b0>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:191) - {'localization_spine': <lib.infers.localization_spine.LocalizationSpine object at 0x7fb1700a5f30>, 'localization_vertebra': <lib.infers.localization_vertebra.LocalizationVertebra object at 0x7fb16cb0d480>, 'segmentation_vertebra': <lib.infers.segmentation_vertebra.SegmentationVertebra object at 0x7fb16cb0d4b0>, 'Histogram+GraphCut': <monailabel.scribbles.infer.HistogramBasedGraphCut object at 0x7fb16cb0e560>, 'GMM+GraphCut': <monailabel.scribbles.infer.GMMBasedGraphCut object at 0x7fb16cb0e590>, 'vertebra_pipeline': <lib.infers.vertebra_pipeline.InferVertebraPipeline object at 0x7fb16cb0e5f0>}
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:206) - +++ Adding Trainer:: localization_spine => <lib.trainers.localization_spine.LocalizationSpine object at 0x7fb16cb0d8d0>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:206) - +++ Adding Trainer:: localization_vertebra => <lib.trainers.localization_vertebra.LocalizationVertebra object at 0x7fb16cb0d780>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (main:206) - +++ Adding Trainer:: segmentation_vertebra => <lib.trainers.segmentation_vertebra.SegmentationVertebra object at 0x7fb16cb0d7e0>
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (monailabel.utils.sessions:51) - Session Path: /home/user1/.cache/monailabel/sessions
[2023-01-14 17:57:49,736] [850291] [MainThread] [INFO] (monailabel.utils.sessions:52) - Session Expiry (max): 3600
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (monailabel.interfaces.app:475) - App Init - completed
[2023-01-14 17:57:49,737] [timeloop] [INFO] Starting Timeloop..
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (timeloop:60) - Starting Timeloop..
[2023-01-14 17:57:49,737] [timeloop] [INFO] Registered job <function MONAILabelApp.on_init_complete.<locals>.run_scheduler at 0x7fb16d05cc10>
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (timeloop:42) - Registered job <function MONAILabelApp.on_init_complete.<locals>.run_scheduler at 0x7fb16d05cc10>
[2023-01-14 17:57:49,737] [timeloop] [INFO] Timeloop now started. Jobs will run based on the interval set
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (timeloop:63) - Timeloop now started. Jobs will run based on the interval set
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (uvicorn.error:59) - Application startup complete.
[2023-01-14 17:57:49,737] [850291] [MainThread] [INFO] (uvicorn.error:206) - Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
[2023-01-14 17:58:05,865] [850291] [MainThread] [INFO] (monailabel.endpoints.activelearning:43) - Active Learning Request: {'strategy': 'random', 'client_id': 'user-xyz'}
[2023-01-14 17:58:05,865] [850291] [MainThread] [INFO] (monailabel.tasks.activelearning.random:47) - Random: Selected Image: CTChest; Weight: 6180
[2023-01-14 17:58:05,865] [850291] [MainThread] [INFO] (monailabel.endpoints.activelearning:59) - Next sample: {'id': 'CTChest', 'weight': 6180, 'path': '/home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr/CTChest.nii.gz', 'ts': 1673729869, 'name': 'CTChest.nii.gz'}
[2023-01-14 17:58:13,440] [850291] [MainThread] [INFO] (monailabel.endpoints.infer:160) - Infer Request: {'model': 'localization_spine', 'image': 'CTChest', 'device': 'cpu', 'result_extension': '.nrrd', 'result_dtype': 'uint8', 'client_id': 'user-xyz'}
[2023-01-14 17:58:13,440] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:279) - Infer Request (final): {'device': 'cpu', 'model': 'localization_spine', 'image': '/home/user1/Escritorio/Monai/datasets/vertebrae/imagesTr/CTChest.nii.gz', 'result_extension': '.nrrd', 'result_dtype': 'uint8', 'client_id': 'user-xyz', 'description': 'A pre-trained model for volumetric (3D) spine localization from CT image'}
[2023-01-14 17:58:13,441] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:76) - PRE - Run Transform(s)
[2023-01-14 17:58:13,441] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:77) - PRE - Input Keys: ['device', 'model', 'image', 'result_extension', 'result_dtype', 'client_id', 'description', 'image_path']
[2023-01-14 17:58:15,869] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (LoadImaged): Time: 2.4272; image: torch.Size([512, 512, 139])(torch.float32)
[2023-01-14 17:58:15,870] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (EnsureTyped): Time: 0.0009; image: torch.Size([512, 512, 139])(torch.float32)
[2023-01-14 17:58:15,870] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (EnsureChannelFirstd): Time: 0.0001; image: torch.Size([1, 512, 512, 139])(torch.float32)
[2023-01-14 17:58:15,884] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (CacheObjectd): Time: 0.0134; image: torch.Size([1, 512, 512, 139])(torch.float32)
[2023-01-14 17:58:16,779] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (Spacingd): Time: 0.8947; image: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 17:58:16,836] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (ScaleIntensityRanged): Time: 0.0569; image: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 17:58:17,448] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (GaussianSmoothd): Time: 0.6115; image: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 17:58:17,494] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (ScaleIntensityd): Time: 0.0461; image: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 17:58:17,494] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:471) - Inferer:: cpu => SlidingWindowInferer => {'roi_size': (96, 96, 96), 'sw_batch_size': 2, 'overlap': 0.4, 'mode': gaussian, 'sigma_scale': 0.125, 'padding_mode': 'replicate', 'cval': 0.0, 'sw_device': None, 'device': None, 'progress': False, 'cpu_thresh': None, 'roi_weight_map': None}
[2023-01-14 17:58:17,494] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:420) - Infer model path: /home/user1/Escritorio/Monai/apps/radiology/model/pretrained_localization_spine.pt
[2023-01-14 18:00:39,455] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:76) - POST - Run Transform(s)
[2023-01-14 18:00:39,455] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:77) - POST - Input Keys: ['device', 'model', 'image', 'result_extension', 'result_dtype', 'client_id', 'description', 'image_path', 'image_meta_dict', 'latencies', 'image_cached', 'pred']
[2023-01-14 18:00:39,455] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (EnsureTyped): Time: 0.0001; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([25, 300, 300, 266])(torch.float32)
[2023-01-14 18:00:39,759] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (Activationsd): Time: 0.3031; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([25, 300, 300, 266])(torch.float32)
[2023-01-14 18:00:42,576] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (AsDiscreted): Time: 2.8175; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 18:00:42,839] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (KeepLargestConnectedComponentd): Time: 0.2623; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 18:00:42,844] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (BinaryMaskd): Time: 0.0051; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([1, 300, 300, 266])(torch.float32)
[2023-01-14 18:00:42,938] [850291] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (Restored): Time: 0.094; image: torch.Size([1, 300, 300, 266])(torch.float32); pred: torch.Size([512, 512, 139])(torch.float32)
[2023-01-14 18:00:43,013] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:571) - Writing Result...
[2023-01-14 18:00:43,014] [850291] [MainThread] [INFO] (monailabel.transform.writer:189) - Result ext: .nrrd; write_to_file: True; dtype: uint8
[2023-01-14 18:00:43,213] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:334) - ++ Latencies => Total: 149.7728; Pre: 4.0537; Inferer: 141.9601; Invert: 0.0000; Post: 3.5590; Write: 0.1997
[2023-01-14 18:00:43,213] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:358) - Result File: /tmp/tmpt87i1c7f.nrrd
[2023-01-14 18:00:43,213] [850291] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:359) - Result Json Keys: ['label_names', 'latencies']

Thanks for the help

Hi

The server does work for sample radiology app “segmentation_spleen” on the same virtualenviroment

~/Escritorio/Monai$ monailabel start_server --app apps/radiology --studies datasets/Task09_Spleen/imagesTr --conf models segmentation_spleen
Using PYTHONPATH=/home/user1:

2023-01-20 12:13:57,337 - USING:: version = False
2023-01-20 12:13:57,337 - USING:: app = /home/user1/Escritorio/Monai/apps/radiology
2023-01-20 12:13:57,337 - USING:: studies = /home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr
2023-01-20 12:13:57,337 - USING:: verbose = INFO
2023-01-20 12:13:57,337 - USING:: conf = [['models', 'segmentation_spleen']]
2023-01-20 12:13:57,337 - USING:: host = 0.0.0.0
2023-01-20 12:13:57,337 - USING:: port = 8000
2023-01-20 12:13:57,337 - USING:: uvicorn_app = monailabel.app:app
2023-01-20 12:13:57,337 - USING:: ssl_keyfile = None
2023-01-20 12:13:57,337 - USING:: ssl_certfile = None
2023-01-20 12:13:57,337 - USING:: ssl_keyfile_password = None
2023-01-20 12:13:57,337 - USING:: ssl_ca_certs = None
2023-01-20 12:13:57,337 - USING:: workers = None
2023-01-20 12:13:57,337 - USING:: limit_concurrency = None
2023-01-20 12:13:57,337 - USING:: access_log = False
2023-01-20 12:13:57,337 - USING:: log_config = None
2023-01-20 12:13:57,337 - USING:: dryrun = False
2023-01-20 12:13:57,337 - USING:: action = start_server
2023-01-20 12:13:57,337 - ENV SETTINGS:: MONAI_LABEL_API_STR = 
2023-01-20 12:13:57,337 - ENV SETTINGS:: MONAI_LABEL_PROJECT_NAME = MONAILabel
2023-01-20 12:13:57,337 - ENV SETTINGS:: MONAI_LABEL_APP_DIR = 
2023-01-20 12:13:57,337 - ENV SETTINGS:: MONAI_LABEL_STUDIES = 
2023-01-20 12:13:57,337 - ENV SETTINGS:: MONAI_LABEL_AUTH_ENABLE = False
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_AUTH_DB = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_APP_CONF = '{}'
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TASKS_TRAIN = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TASKS_STRATEGY = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TASKS_SCORING = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TASKS_BATCH_INFER = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_URL = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_USERNAME = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_PASSWORD = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_API_KEY = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_CACHE_PATH = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_PROJECT = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_ASSET_PATH = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_DSA_ANNOTATION_GROUPS = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_USERNAME = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_PASSWORD = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CACHE_PATH = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_QIDO_PREFIX = None
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_WADO_PREFIX = None
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_STOW_PREFIX = None
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_FETCH_BY_FRAME = False
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CONVERT_TO_NIFTI = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_SEARCH_FILTER = '{"Modality": "CT"}'
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_CACHE_EXPIRY = 180
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_PROXY_TIMEOUT = 30.0
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DICOMWEB_READ_TIMEOUT = 5.0
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_AUTO_RELOAD = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_READ_ONLY = False
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_DATASTORE_FILE_EXT = '["*.nii.gz", "*.nii", "*.nrrd", "*.jpg", "*.png", "*.tif", "*.svs", "*.xml"]'
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_SERVER_PORT = 8000
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_CORS_ORIGINS = '[]'
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_SESSIONS = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_SESSION_PATH = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_SESSION_EXPIRY = 3600
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_INFER_CONCURRENCY = -1
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_INFER_TIMEOUT = 600
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TRACKING_ENABLED = True
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_TRACKING_URI = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_ZOO_SOURCE = github
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_ZOO_REPO = Project-MONAI/model-zoo/hosting_storage_v1
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_ZOO_AUTH_TOKEN = 
2023-01-20 12:13:57,338 - ENV SETTINGS:: MONAI_LABEL_AUTO_UPDATE_SCORING = True
2023-01-20 12:13:57,338 - 
Allow Origins: ['*']
[2023-01-20 12:13:57,611] [1927570] [MainThread] [INFO] (uvicorn.error:75) - Started server process [1927570]
[2023-01-20 12:13:57,611] [1927570] [MainThread] [INFO] (uvicorn.error:45) - Waiting for application startup.
[2023-01-20 12:13:57,611] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.app:38) - Initializing App from: /home/user1/Escritorio/Monai/apps/radiology; studies: /home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr; conf: {'models': 'segmentation_spleen'}
[2023-01-20 12:13:57,636] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for MONAILabelApp Found: <class 'main.MyApp'>
[2023-01-20 12:13:57,638] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepgrow_3d.Deepgrow3D'>
[2023-01-20 12:13:57,639] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepedit.DeepEdit'>
[2023-01-20 12:13:57,639] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.localization_vertebra.LocalizationVertebra'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation_spleen.SegmentationSpleen'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.deepgrow_2d.Deepgrow2D'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation_vertebra.SegmentationVertebra'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.segmentation.Segmentation'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.class_utils:37) - Subclass for TaskConfig Found: <class 'lib.configs.localization_spine.LocalizationSpine'>
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (main:93) - +++ Adding Model: segmentation_spleen => lib.configs.segmentation_spleen.SegmentationSpleen
[2023-01-20 12:13:57,640] [1927570] [MainThread] [INFO] (monailabel.utils.others.generic:185) - Downloading resource: /home/user1/Escritorio/Monai/apps/radiology/model/pretrained_segmentation_spleen.pt from https://github.com/Project-MONAI/MONAILabel/releases/download/pretrained/radiology_segmentation_unet_spleen.pt
pretrained_segmentation_spleen.pt: 18.4MB [00:06, 3.09MB/s]                                                                                                                                                              
2023-01-20 12:14:03,890 - INFO - Downloaded: /home/user1/Escritorio/Monai/apps/radiology/model/pretrained_segmentation_spleen.pt
2023-01-20 12:14:03,890 - INFO - Expected md5 is None, skip md5 check for file /home/user1/Escritorio/Monai/apps/radiology/model/pretrained_segmentation_spleen.pt.
[2023-01-20 12:14:04,938] [1927570] [MainThread] [INFO] (lib.configs.segmentation_spleen:75) - EPISTEMIC Enabled: False; Samples: 5
[2023-01-20 12:14:04,938] [1927570] [MainThread] [INFO] (main:96) - +++ Using Models: ['segmentation_spleen']
[2023-01-20 12:14:04,938] [1927570] [MainThread] [INFO] (monailabel.interfaces.app:135) - Init Datastore for: /home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr
[2023-01-20 12:14:04,939] [1927570] [MainThread] [INFO] (monailabel.datastore.local:129) - Auto Reload: True; Extensions: ['*.nii.gz', '*.nii', '*.nrrd', '*.jpg', '*.png', '*.tif', '*.svs', '*.xml']
[2023-01-20 12:14:04,943] [1927570] [MainThread] [INFO] (monailabel.datastore.local:576) - Invalidate count: 0
[2023-01-20 12:14:04,943] [1927570] [MainThread] [INFO] (monailabel.datastore.local:150) - Start observing external modifications on datastore (AUTO RELOAD)
[2023-01-20 12:14:04,943] [1927570] [MainThread] [INFO] (main:126) - +++ Adding Inferer:: segmentation_spleen => <lib.infers.segmentation_spleen.SegmentationSpleen object at 0x7f3ed429dfc0>
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (main:191) - {'segmentation_spleen': <lib.infers.segmentation_spleen.SegmentationSpleen object at 0x7f3ed429dfc0>, 'Histogram+GraphCut': <monailabel.scribbles.infer.HistogramBasedGraphCut object at 0x7f3ed3d1e6b0>, 'GMM+GraphCut': <monailabel.scribbles.infer.GMMBasedGraphCut object at 0x7f3ed3d1e6e0>}
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (main:206) - +++ Adding Trainer:: segmentation_spleen => <lib.trainers.segmentation_spleen.SegmentationSpleen object at 0x7f3ed3d982e0>
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (monailabel.utils.sessions:51) - Session Path: /home/user1/.cache/monailabel/sessions
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (monailabel.utils.sessions:52) - Session Expiry (max): 3600
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (monailabel.interfaces.app:475) - App Init - completed
[2023-01-20 12:14:04,944] [timeloop] [INFO] Starting Timeloop..
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (timeloop:60) - Starting Timeloop..
[2023-01-20 12:14:04,944] [timeloop] [INFO] Registered job <function MONAILabelApp.on_init_complete.<locals>.run_scheduler at 0x7f3ed3d0ee60>
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (timeloop:42) - Registered job <function MONAILabelApp.on_init_complete.<locals>.run_scheduler at 0x7f3ed3d0ee60>
[2023-01-20 12:14:04,944] [timeloop] [INFO] Timeloop now started. Jobs will run based on the interval set
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (timeloop:63) - Timeloop now started. Jobs will run based on the interval set
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (uvicorn.error:59) - Application startup complete.
[2023-01-20 12:14:04,944] [1927570] [MainThread] [INFO] (uvicorn.error:206) - Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
[2023-01-20 12:14:40,865] [1927570] [MainThread] [INFO] (monailabel.endpoints.activelearning:43) - Active Learning Request: {'strategy': 'first', 'client_id': 'user-xyz'}
[2023-01-20 12:14:40,865] [1927570] [MainThread] [INFO] (monailabel.tasks.activelearning.first:38) - First: Selected Image: spleen_10
[2023-01-20 12:14:40,868] [1927570] [MainThread] [INFO] (monailabel.endpoints.activelearning:59) - Next sample: {'id': 'spleen_10', 'path': '/home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr/spleen_10.nii.gz', 'ts': 1673700860, 'name': 'spleen_10.nii.gz', 'strategy': {'first': {'ts': 1674227680, 'client_id': 'user-xyz'}}}
[2023-01-20 12:14:53,752] [1927570] [MainThread] [INFO] (monailabel.endpoints.activelearning:43) - Active Learning Request: {'strategy': 'first', 'client_id': 'user-xyz'}
[2023-01-20 12:14:53,752] [1927570] [MainThread] [INFO] (monailabel.tasks.activelearning.first:38) - First: Selected Image: spleen_10
[2023-01-20 12:14:53,754] [1927570] [MainThread] [INFO] (monailabel.endpoints.activelearning:59) - Next sample: {'id': 'spleen_10', 'path': '/home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr/spleen_10.nii.gz', 'ts': 1673700860, 'name': 'spleen_10.nii.gz', 'strategy': {'first': {'ts': 1674227693, 'client_id': 'user-xyz'}}}
[2023-01-20 12:15:01,791] [1927570] [MainThread] [INFO] (monailabel.endpoints.infer:160) - Infer Request: {'model': 'segmentation_spleen', 'image': 'spleen_10', 'device': 'cpu', 'result_extension': '.nrrd', 'result_dtype': 'uint8', 'client_id': 'user-xyz'}
[2023-01-20 12:15:01,791] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:279) - Infer Request (final): {'device': 'cpu', 'model': 'segmentation_spleen', 'image': '/home/user1/Escritorio/Monai/datasets/Task09_Spleen/imagesTr/spleen_10.nii.gz', 'result_extension': '.nrrd', 'result_dtype': 'uint8', 'client_id': 'user-xyz', 'description': 'A pre-trained model for volumetric (3D) segmentation of the spleen from CT image'}
[2023-01-20 12:15:01,792] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:76) - PRE - Run Transform(s)
[2023-01-20 12:15:01,792] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:77) - PRE - Input Keys: ['device', 'model', 'image', 'result_extension', 'result_dtype', 'client_id', 'description', 'image_path']
[2023-01-20 12:15:02,083] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (LoadImaged): Time: 0.2906; image: torch.Size([512, 512, 55])(torch.float32)
[2023-01-20 12:15:02,084] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (EnsureTyped): Time: 0.0002; image: torch.Size([512, 512, 55])(torch.float32)
[2023-01-20 12:15:02,084] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (EnsureChannelFirstd): Time: 0.0001; image: torch.Size([1, 512, 512, 55])(torch.float32)
[2023-01-20 12:15:04,614] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (Spacingd): Time: 2.53; image: torch.Size([1, 500, 500, 271])(torch.float32)
[2023-01-20 12:15:04,797] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - PRE - Transform (ScaleIntensityRanged): Time: 0.1831; image: torch.Size([1, 500, 500, 271])(torch.float32)
[2023-01-20 12:15:04,798] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:471) - Inferer:: cpu => SlidingWindowInferer => {'roi_size': (160, 160, 160), 'sw_batch_size': 1, 'overlap': 0.25, 'mode': constant, 'sigma_scale': 0.125, 'padding_mode': constant, 'cval': 0.0, 'sw_device': None, 'device': None, 'progress': False, 'cpu_thresh': None, 'roi_weight_map': None}
[2023-01-20 12:15:04,798] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:420) - Infer model path: /home/user1/Escritorio/Monai/apps/radiology/model/pretrained_segmentation_spleen.pt
[2023-01-20 12:15:17,355] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:76) - POST - Run Transform(s)
[2023-01-20 12:15:17,355] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:77) - POST - Input Keys: ['device', 'model', 'image', 'result_extension', 'result_dtype', 'client_id', 'description', 'image_path', 'image_meta_dict', 'latencies', 'pred']
[2023-01-20 12:15:17,356] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (EnsureTyped): Time: 0.0001; image: torch.Size([1, 500, 500, 271])(torch.float32); pred: torch.Size([2, 500, 500, 271])(torch.float32)
[2023-01-20 12:15:17,420] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (Activationsd): Time: 0.0644; image: torch.Size([1, 500, 500, 271])(torch.float32); pred: torch.Size([2, 500, 500, 271])(torch.float32)
[2023-01-20 12:15:22,801] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (AsDiscreted): Time: 5.3802; image: torch.Size([1, 500, 500, 271])(torch.float32); pred: torch.Size([1, 500, 500, 271])(torch.float32)
[2023-01-20 12:15:22,858] [1927570] [MainThread] [INFO] (monailabel.interfaces.utils.transform:122) - POST - Transform (Restored): Time: 0.0571; image: torch.Size([1, 500, 500, 271])(torch.float32); pred: torch.Size([512, 512, 55])(torch.float32)
[2023-01-20 12:15:22,885] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:571) - Writing Result...
[2023-01-20 12:15:22,886] [1927570] [MainThread] [INFO] (monailabel.transform.writer:189) - Result ext: .nrrd; write_to_file: True; dtype: uint8
[2023-01-20 12:15:24,913] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:334) - ++ Latencies => Total: 23.1218; Pre: 3.0060; Inferer: 12.5575; Invert: 0.0000; Post: 5.5303; Write: 2.0278
[2023-01-20 12:15:24,913] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:358) - Result File: /tmp/tmpq5s1_hsr.nrrd
[2023-01-20 12:15:24,913] [1927570] [MainThread] [INFO] (monailabel.tasks.infer.basic_infer:359) - Result Json Keys: ['label_names', 'latencies']

Could someone help? or give pointers?

Thanks a lot

1 Like

@diazandr3s can maybe comment and will present vertrebrae segmentation with 3D Slicer sample datasets in the upcoming MONAILabel workshop on Wednesday January 25th 9-11am EST

2 Likes

Thanks for the ping, @rbumm.

@mau_igna_06:

The vertebra segmentation model has been trained on a portion of the VerSe dataset. This means the model may need training to work on other datasets.

Can you try the localization_vertebra or vertebra_pipeline? It should be under the list of Auto Segmentation models.

You may also be interested on the whole-body CT segmentation model that includes vertebra segmentation. This model was trained on the Total Segmentator dataset.

Here you can find the app folder: radiology_full_ct-upgraded-HYBRID - Google Drive

Please let us know how that goes.

1 Like

Forgot to add the command to start the server:

monailabel start_server --app radiology_full_ct-upgraded-HYBRID/ --studies PATH_TO_IMAGES/ --conf models segmentation_full_ct

1 Like

Hi @diazandr3s

Thank you for your answer

None of them worked, one gave an empty result and the other an error of mismatching tensor size on the server

This one worked. Thank you

The original CT had 0.65mm spacing, but the automatic segmentation result looks much rougher than when I do a manual threshold. Please see the picture below:

Can this be corrected?

1 Like

Yes, this is expected - @diazandr3s has told me that this model is only trained at 1.5mm and another at 2mm which is the same as TotalSegmentator. In my experience the MONAI Label ct_full and TotalSegmentator give comparable results, just with different resource requirements (MONAI can be faster but uses more memory). I understand either could be trained to run inference at higher resolution but I don’t believe that has happened yet. We plan to do some spine-specific higher resolution models with @Ron Alkalay.

2 Likes

Thanks. flying out today, will have internet by 9 or 10 est tomorrow

Ron N Alkalay
Associate Professor,
Dept of Orthopedic Surgery, Harvard Medical School
Center for Advanced
Orthopedic Studies
Beth Israel
Deaconess Medical Center
1 Overland Street
Boston, MA, 02215
Tel. 617-667-5185
Fax. 617-667-7175
email:
rn_alkalay@bidmc.harvard.edu

1 Like

Thanks for sharing the results, @mau_igna_06.

As @pieper said, we’ve trained two models on the Total Segmentator dataset: one using 1.5mm and the other using 2mm.

Just to check how it goes, could you try running inference after changing the self.target_spacing values to (0.6, 0.6, 0.6) in the file lib/configs/segmentation_full_ct?

Screenshot from 2023-01-22 21-09-11

Retraining the model on better resolution shouldn’t be difficult at all. You just need to change the self.target_spacing values to (0.6, 0.6, 0.6), or the one that fits on your GPU/CPU, in the file lib/configs/segmentation_full_ct and trigger training.

Hi @diazandr3s

Thank you for your answer

I’ll do

I wonder if we could train a preliminar stage of a bone segmentation model using just binary thresholds (then they could be separated by connected components by another stage or a human) or create an expert threshold limit chooser for bones segmentation using the training data from totalSegmentator.
On our application, making anatomic surgical guide bases, only surface of cortical bone is important (considering other geometric restrictions). Currently we create them with the segment editor.

1 Like

Definitely! that can be done. This is a good question/topic to talk about during the workshop this Wednesday :slight_smile:

1 Like

Hi

My laptop could not allocate 58GB of RAM

Thank you anyway, I’ll keep exploring in the near future with better (cloud?) hardware

1 Like

I suspected this is the challenge with high-resolution images

1 Like