Automatic image segmentation methods and analysis
The invention provides methods and apparatus for image processing that perform image segmentation on data sets in two- and/or three-dimensions so as to resolve structures that have the same or similar grey values (and that would otherwise render with the same or similar intensity values) and that, thereby, facilitate visualization and processing of those data sets.
1. An apparatus for revealing a volumetric region comprising:
(a) a storage device;
(b) a network to which a processed image can be transmitted;
(c) a Digital Data Processor (DDP) comprising a central processing unit, a graphics processing unit, a dynamic memory, a software program, an input device and an output device, where the output device is adapted to enable the processed image to be transmitted to the storage device, where the software program is adapted to receive and/or generate a plurality of 2D image slices in a CT runoff study corresponding to a portion of a patient's body, where each 2D image slice of the plurality of 2D image slices comprises a plurality of pixels, where the software program is further adapted to:
(i) generate a first histogram of the plurality of 2D image slices;
(ii) compute a cross-correlation between the first histogram and an atlas of average histograms;
(iii) determine based on the cross-correlation a matching body part;
(iv) assign the CT runoff study to the matching body part based on step(c)(iii);
(v) determine a threshold;
(vi) perform a segmentation based on the threshold to determine one or more parameters selected from the group consisting of a first connected component, a total area occupied by the first connected component, a second histogram of the first connected component, and/or a third histogram of the total area occupied by the first connected component;
(vii) determine a volumetric region of the first connected component based on step(c)(vi);
(viii) generate a plurality of extracted 2D image slices by extracting the first connected component from the plurality of 2D image slices; and
(viii) label a second connected component based on the plurality of extracted 2D image slices.
2. The apparatus of claim 1 , where the plurality of 2D image slices correspond to an upper leg.
3. The apparatus of claim 2 , where one or more pixels of the plurality of pixels in a 2D image slice of the plurality of 2D image slices are identified as corresponding to a bone structure based on one or both the segmentation and an intensity of the one or more pixels above the threshold.
4. The apparatus of claim 3 , where a diameter is determined for the bone structure occupied by the one or more pixels in the 2D image slice identified as corresponding to the bone structure.
5. The apparatus of claim 4 , where one or more pixels connected with the one or more pixels in the 2D image slice identified as corresponding to the bone structure are identified as one or more additional bone structure pixels.
6. The apparatus of claim 4 , where one or more pixels adjacent with the one or more pixels in the 2D image slice identified as corresponding to the bone structure are identified as one or more additional bone structure pixels.
7. The apparatus of claim 4 , where one or more pixels with an area greater than the diameter of a largest vessel in the volumetric region and bounded by the diameter are identified as one or more additional bone structure pixels.
8. The apparatus of claim 4 , where one or more pixels with a geometric characteristic are identified as one or more additional bone structure pixels, where the geometric characteristic is selected from the group consisting of shape of the bone structure in the 2D image slice and size of the bone structure in the 2D image slice.
9. The apparatus of claim 8 , where all pixels not assigned as the bone structure are labelled as vessel.
10. The apparatus of claim 9 , where all pixels assigned as the bone structure are extracted from the 2D image slice.
11. The apparatus of claim 10 , where removing the bone structure from the 2D image slice reveals a vessel.
12. The apparatus of claim 3 , where a pixel of the plurality of pixels is not assigned to the bone structure based on an intensity of the pixel below the threshold.
13. The apparatus of claim 1 , where the threshold is between:
a lower limit of 130 Hounsfield Units (HU); and
an upper limit of 400 HU.
14. The apparatus of claim 13 , where steps (v) through (vii) are repeated with an increased threshold.
15. The apparatus of claim 14 , where the increased threshold is the threshold increased by 1 HU.
16. The apparatus of claim 1 , where the threshold is between:
a lower limit of −1000 Hounsfield Units (HU); and
an upper limit of 400 HU.
17. The apparatus of claim 16 , where steps (v) through (vii) are repeated with an increased threshold.
18. The apparatus of claim 17 , where the increased threshold is the threshold increased by 1 HU.