opentps.core.processing.segmentation package
Submodules
opentps.core.processing.segmentation.segmentation3D module
- applyThreshold(image, thresholdMin, thresholdMax=inf)
Apply a threshold to an image and return a ROIMask object (True if value > thresholdMin and value < thresholdMax)
- Parameters:
image (Image3D) – The image to be thresholded
thresholdMin (float) – The minimum threshold value
thresholdMax (float) – The maximum threshold value (default: np.inf)
- Returns:
mask – The mask of the thresholded image
- Return type:
- getBoxAboveThreshold(data: Image3D, threshold=0.0)
Get the box around an ROI in scanner coordinates and apply a threshold (using the ROI origin and spacing)
- Parameters:
data (Image3D) – The image with the ROI to be thresholded
threshold (float) – The minimum threshold value in the ROI(default: 0.)
- Returns:
boundingBox – The box around which the data is cropped and the threshold applied under the form [[x1, X2], [y1, y2], [z1, z2]]
- Return type:
list of tuples or list
- getBoxAroundROI(ROI) Sequence[Sequence[float]]
Returns the 3D scanner coordinates (in mm) of the smallest box surrounding the given ROI. By convention, the returned box coordinates are the coordinates of centers of the voxels at the extremities of the box. In other words, a box that passes through the center of a voxel necessarily contains this voxel. For example, in 1 dimension, with a spacing of 1mm and origin at 0mm, a box around voxels [1,2] will results in box coordinates [1,2] mm even though the size of the box is 2mm (because it includes 2 voxels of 1mm spacing).
- Parameters:
ROI (an ROIContour or ROIMask) – The contour that is contained in the desired box
- Returns:
boxInUniversalCoords – The box coordinates, under the form [[x1, X2], [y1, y2], [z1, z2]]
- Return type:
list of tuples or list
opentps.core.processing.segmentation.segmentationCT module
- class SegmentationCT(ct)
Bases:
object
Class for CT segmentation
- Parameters:
ct (Image3D) – The CT image to be segmented
- Variables:
ct (Image3D) – The CT image to be segmented
- segmentBody()
Segment the body from the CT image
- Returns:
body – The body mask (bool mask)
- Return type:
- segmentBones(body=None)
Segment the bones from the CT image
- Parameters:
body (Image3D) – The body mask (bool mask)
- Returns:
bones – The bones mask (bool mask)
- Return type:
- compute3DStructuralElement(radiusXYZ, spacing=[1, 1, 1])
Compute a 3D structural element for morphological operations (e.g. dilation, erosion).
- Parameters:
radiusXYZ (list) – The radius of the structural element in each direction
spacing (list) – The spacing of the image in each direction (default: [1,1,1])
- Returns:
filt – The structural element (bool mask)
- Return type:
bool numpy array