Add update for dseg/rtstruct [WIP]

models/fmcib_radiomics/config/default.yml

@@ -1,22 +1,34 @@
 general:
   data_base_dir: /app/data
   version: 1.0
-  description: FMCIB pipeline starting from DICOM files and centroids in json files or slicer exports named by their SeriesInstanceUID
+  description: Run fmcib radiomics pipeline on dicom data
 
 execute:
 - DicomImporter
-- FileImporter
+- DsegExtractor
 - NiftiConverter
+- CentroidExtractor
 - FMCIBRunner
 - DataOrganizer
 
 modules:
+  DicomImporter:
+    source_dir: input_data
+    import_dir: sorted_data
+    sort_data: true
+    merge: true
+    meta: 
+      mod: '%Modality'
+      desc: '%SeriesDescription'
 
-  FileImporter:
-    instance_id: sid
-    meta: type=fmcibcoordinates
-    type: json
+  # roi can be specified manually but will otherwise extracted form the dicomseg via the meta.json 
+  # DsegExtractor:
+  #   roi: 
+  #   - LIVER
+  #   - LIVER+NEOPLASM_MALIGNANT_PRIMARY
 
   DataOrganizer:
     targets:
-      - json:type=fmcibfeatures-->[i:sid]/features.json
+    - csv-->[i:sid]/features.csv
+  # - nifit:mod=seg:origin=dicomseg-->[i:sid]/masks/[basename]
+  # - nifti-->[i:sid]/nifti/[d:mod]/[basename]

models/fmcib_radiomics/config/from_centroids.yml

@@ -17,4 +17,4 @@ modules:
 
   DataOrganizer:
     targets:
-      - json:type=fmcibfeatures-->[i:patientID]/features.json
+    - csv-->[i:patientID]/features.csv

models/fmcib_radiomics/config/from_json.yml

@@ -0,0 +1,22 @@
+general:
+  data_base_dir: /app/data
+  version: 1.0
+  description: FMCIB pipeline starting from DICOM files and centroids in json files or slicer exports named by their SeriesInstanceUID
+
+execute:
+- DicomImporter
+- FileImporter
+- NiftiConverter
+- FMCIBRunner
+- DataOrganizer
+
+modules:
+
+  FileImporter:
+    instance_id: sid
+    meta: type=fmcibcoordinates
+    type: json
+
+  DataOrganizer:
+    targets:
+      - csv-->[i:sid]/features.csv

models/fmcib_radiomics/config/from_nifti_mask.yml

@@ -0,0 +1,21 @@
+general:
+  data_base_dir: /app/data
+  version: 1.0
+  description: "FMCIB pipeline starting from a nii.gz file image and a binary mask of the GTV."
+
+execute:
+- FileStructureImporter
+- CentroidExtractor
+- FMCIBRunner
+- DataOrganizer
+
+modules:
+  FileStructureImporter:
+    structures:
+      - $patientID/CT.nii.gz@instance@nifti:mod=ct
+      - $patientID/masks/GTV.nii.gz@nifti:mod=seg
+    import_id: patientID
+
+  DataOrganizer:
+    targets:
+      - csv-->[i:patientID]/features.csv

models/fmcib_radiomics/config/from_rtstruct.yml

@@ -0,0 +1,33 @@
+general:
+  data_base_dir: /app/data
+  version: 1.0
+  description: run pyradiomics pipeline on dicom data
+
+execute:
+- DicomImporter
+- RTStructExtractor
+- NiftiConverter
+- CentroidExtractor
+- FMCIBRunner
+- DataOrganizer
+
+modules:
+  DicomImporter:
+    source_dir: input_data
+    import_dir: sorted_data
+    sort_data: true
+    merge: true
+    meta: 
+      mod: '%Modality'
+      desc: '%SeriesDescription'
+
+  # roi can be specified manually but will otherwise extracted form the rtstruct segemnetation names
+  # RTStructExtractor:
+  #   roi: 
+  #   - LIVER
+  #   - LIVER+NEOPLASM_MALIGNANT_PRIMARY
+
+  DataOrganizer:
+    targets:
+      - csv-->[i:sid]/features.csv
+        # - nifti-->[i:sid]/nifti/[d:mod]/[basename]

models/fmcib_radiomics/config/from_slicer.yml

@@ -17,4 +17,4 @@ modules:
 
   DataOrganizer:
     targets:
-      - json:type=fmcibfeatures-->[i:patientID]/features.json
+      - csv-->[i:patientID]/features.csv

models/fmcib_radiomics/meta.json

@@ -4,28 +4,25 @@
   "title": "Foundation Model for Cancer Imaging Biomarkers",
   "summary": {
     "description": "A foundation model for cancer imaging biomarker discovery trained through self-supervised learning using a dataset of 11,467 radiographic lesions. The model features can be used as a data-driven substitute for classical radiomic features",
-    "inputs": [
-      {
-        "label": "Input CT Image",
-        "description": "CT imaging data containing lesions of interest, such as nodules or tumors",
-        "format": "DICOM",
-        "modality": "CT",
-        "slicethickness": "5mm",
-        "bodypartexamined": "WHOLEBODY",
-        "non-contrast": true,
-        "contrast": true
-      },
-      {
-        "label": "Center of mass",
-        "description": "Center of mass of the lesion in the CT image",
-        "format": "JSON",
-        "modality": "JSON",
-        "slicethickness": "5mm",
-        "bodypartexamined": "WHOLEBODY",
-        "non-contrast": true,
-        "contrast": true
-      }
-    ],
+    "inputs": [ {
+      "label": "Input Image",
+      "description": "The input image.",
+      "format": "DICOM",
+      "modality": "CT",
+      "bodypartexamined": "WHOLEBODY",
+      "slicethickness": "2.5mm",
+      "non-contrast": true,
+      "contrast": true
+    }, {
+      "label": "Input Segmentation",
+      "description": "The input segmentation to be analysed.",
+      "format": "DICOMSEG",
+      "modality": "SEG",
+      "bodypartexamined": "WHOLEBODY",
+      "slicethickness": "2.5mm",
+      "non-contrast": true,
+      "contrast": true
+    } ],
     "outputs": [
       {
         "type": "Prediction",

models/fmcib_radiomics/mhub.toml

@@ -1,2 +1,2 @@
 [model.deployment]
-test = "https://zenodo.org/record/13785615/files/fmcib_radiomics.test.zip"
+test = "https://zenodo.org/records/14205464/files/test.zip"

models/fmcib_radiomics/utils/CentroidExtractor.py

@@ -1,43 +1,58 @@
 """
 ---------------------------------------------------------
-Author: Leonard Nürnberg
-Email:  [email protected]
+Author: Leonard Nürnberg, Suraj Pai
+Email:  [email protected], [email protected]
 Date:   06.03.2024
 ---------------------------------------------------------
 """
 
 import json, jsonschema
-from mhubio.core import Instance, InstanceData, IO, Module
+from mhubio.core import Instance, InstanceData, InstanceDataCollection, IO, Module
 import SimpleITK as sitk
+import numpy as np
 
 class CentroidExtractor(Module):
-  
+
     @IO.Instance()
-    @IO.Input('in_mask', 'nrrd:mod=seg', the='Tumor segmentation mask for the input NRRD file.')
-    @IO.Output('centroids_json', 'centroids.json', "json:type=fmcibcoordinates", the='JSON file containing 3D coordinates of the centroid of the input mask.')
-    def task(self, instance: Instance, in_mask: InstanceData, centroids_json: InstanceData) -> None:
-
-        # read the input mask 
-        mask = sitk.ReadImage(in_mask.abspath)
-
-        # get the center of massk from the mask via ITK
-        label_shape_filter = sitk.LabelShapeStatisticsImageFilter()
-        label_shape_filter.Execute(mask)
-        try:
-            centroid = label_shape_filter.GetCentroid(255)
-        except:
-            centroid = label_shape_filter.GetCentroid(1)
-
-        # extract x, y, and z coordinates from the centroid
-        x, y, z = centroid
-
-        # set up the coordinate dictionary
-        coordinate_dict = {
-            "coordX": x,
-            "coordY": y,
-            "coordZ": z,
-        }
-
-        # write the coordinate dictionary to a json file
-        with open(centroids_json.abspath, "w") as f:
-            json.dump(coordinate_dict, f)
+    @IO.Inputs('in_masks', 'nrrd|nifti:mod=seg', the='Tumor segmentation masks for the input NRRD files')
+    @IO.Outputs('centroid_jsons', '[filename].json', "json:type=fmcibcoordinates", data='in_masks', the='JSON file containing 3D coordinates of the centroid of the input mask.')
+    def task(self, instance: Instance, in_masks: InstanceDataCollection, centroid_jsons: InstanceDataCollection) -> None:
+        for i, in_mask in enumerate(in_masks):
+            seg_rois = in_mask.type.meta['roi'].split(',')
+            mask = sitk.ReadImage(in_mask.abspath)
+            mask_array = sitk.GetArrayFromImage(mask)
+            unique_values = np.unique(mask_array)
+            print(f"Unique values: {unique_values}")
+            label_shape_filter = sitk.LabelShapeStatisticsImageFilter()
+            seg_roi_coordinates = []
+
+            for channel_id, seg_roi in enumerate(seg_rois):
+                # Check if the label exists in the mask
+                label = channel_id + 1
+                if label not in unique_values:
+                    print(f"Warning: Label {label} (ROI: {seg_roi}) not found in the mask. Skipping.")
+                    continue
+
+                # Calculate centroid if the label exists
+                label_shape_filter.Execute(mask)
+                try:
+                    centroid = label_shape_filter.GetCentroid(label)
+                    # Extract x, y, and z coordinates from the centroid
+                    x, y, z = centroid
+
+                    # Set up the coordinate dictionary
+                    coordinate_dict = {
+                        "Mhub ROI": seg_roi,
+                        "coordX": x,
+                        "coordY": y,
+                        "coordZ": z,
+                    }
+
+                    seg_roi_coordinates.append(coordinate_dict)
+                except Exception as e:
+                    print(f"Error processing label {label} (ROI: {seg_roi}): {e}")
+
+            centroid_json = centroid_jsons.get(i)
+            # Write the coordinate dictionary to a json file
+            with open(centroid_json.abspath, "w") as f:
+                json.dump(seg_roi_coordinates, f)