Method and system for identifying regions in an image
A method and system for visualizing regions in an image is provided. The method comprises computing a regional response around a region in the image, deriving a region score based on from the regional response for the region and labeling the region in the image by comparing the region score to a plurality of probabilistic models.
1. A method for assigning labels to regions in an image, the method comprising:
deriving, by a processor, a plurality of probabilistic models for a plurality of geometrical structures;
computing, by a processor, a regional response around a region in the image;
computing, by a processor, a region score for each geometrical structure using the plurality of probabilistic models; and
labeling, by the processor, the region in the image based on the region score.
2. The method of claim 1 , wherein the geometrical structures comprise anatomical structures and the labeling comprising labeling the anatomical structures.
3. The method of claim 2 , wherein the anatomical structure is one of a nodule, a vessel, a polyp, a tumor, a lesion, a fold, an aneurysm and a pulmonary embolism.
4. The method of claim 1 , wherein the deriving the probabilistic models comprises:
modeling, by a processor, the anatomical structures using a plurality of geometric models; and
representing, by the processor, the anatomical structures using model parameters.
5. The method of claim 4 , wherein computing the regional response comprises computing the regional response as a function of the geometrical models and the model parameters.
6. The method of claim 5 , further comprising deriving, by the processor, a distribution of the regional response as a function of the model parameter.
7. The method of claim 6 , wherein the deriving the distribution comprises applying a knowledgebase of anatomical and functional information.
8. The method of claim 6 , further comprising deriving, by the processor, a distribution of the regional responses for noisy regions in the image.
9. The method of claim 1 , wherein the regional response comprises of a response at a pixel of interest or a voxel of interest.
10. The method of claim 1 , wherein the regional response comprises of a response at a neighborhood around a pixel of interest or a voxel of interest.
11. The method of claim 1 , wherein the regional response comprises of principal curvatures for the region.
12. The method of claim 1 , wherein the regional response comprises of a function of the intensity or texture data for the region.
13. The method of claim 1 , further comprising identifying, by the processor, a set of regions from the image.
14. The method of claim 13 , wherein the identifying comprises thresholding each pixel or voxel in the image to identify the regions of interest.
15. The method of claim 1 , wherein the image comprises a medical image.
16. The method of claim 1 , wherein the image comprises a two-dimensional image, a three-dimensional image, a four-dimensional image, or a five-dimensional image.
17. The method of claim 1 , further comprising displaying the labeled regions and assigning a respective color for each label for visualizing the corresponding regions of interest.
18. A medical imaging system for labeling anatomical structures in an image, the system comprising,
an image processor configured to:
compute a regional response around a voxel of interest in the image;
compute a voxel score for each anatomical structures based on a volume measure of the anatomical structure for a plurality of probabilistic models;
label the voxel of interest in the image based on the voxel score; and
a display unit configured to display the image including the labeled anatomical regions.
19. The imaging system of claim 18 , wherein the probabilistic models comprise histograms.
20. The imaging system of claim 19 , wherein the histograms include parameters obtained from fixed shapes and a distribution of shapes.
21. The imaging system of claim 18 , wherein the regional response comprises a geometric response for the image voxel.
22. The imaging system of claim 18 , wherein the regional response comprises an intensity response for the image voxel.
23. The imaging system of claim 18 , wherein the anatomical regions include vessels, nodules, polyps, folds, aneurysm or junctions of vessels trees.
24. The imaging system of claim 18 , wherein the imaging system comprises at least one of a computed tomography (CT) system, positron emission tomography (PET) system, a single photon emission computed tomography (SPECT) system, magnetic resonance imaging system, microscopy or a digital radiography system.
25. A computed tomography (CT) system for labeling anatomical structures in a CT image, the system comprising,
an image processor configured to:
compute a regional response around a voxel of interest in the CT image;
compute a voxel score for each anatomical structures based on a plurality of probabilistic models;
label the voxel of interest in the image based on the voxel score; and
a display unit configured to display the CT image including the labeled anatomical structures.
26. The CT system of claim 25 , wherein image processor is configured to develop the plurality of probabilistic models using a distribution of geometrical parameters.
27. The CT system of claim 25 , wherein image processor is configured to label the voxel of interest using probabilistic models for curvature data in a neighborhood of the voxel of interest.
28. A non-transitory computer-readable medium storing computer instructions for instructing a computer system to code uncompressed data, the computer instructions including:
deriving, by a processor, a probabilistic model for a plurality of geometrical structures;
computing, by a processor, a regional response around a region in the image;
computing, by a processor, a region score for each geometrical structure using the plurality of probabilistic models; and
labeling, by a processor, the region in the image based on the region score.
29. The system of claim 28 , wherein the geometrical models comprise anatomical structures.
30. The system of claim 29 , wherein the deriving, processor, the probabilistic models comprises:
modeling, by a processor, the anatomical structures using a plurality of geometric models; and
representing the anatomical structures using model parameters.
31. The system of claim 28 , wherein computing the regional response comprises computing the regional response as a function of the geometrical models and the model parameters.
32. The system of claim 30 , further comprising deriving, by the processor, a distribution of the regional response as a function of the model parameter, wherein the deriving the distribution comprises applying a knowledgebase of anatomical and functional information.
33. The system of claim 31 , further comprising deriving, by the processor, a distribution of the regional responses for noisy regions in the image.