Processing medical image information to detect anatomical abnormalities
View Patent ↗A method, system, and related computer program products are provided for processing a medical image of a body part according to a computer-aided detection (CAD) algorithm, the medical image having an image border, the body part comprising imaged tissue appearing inside the image border and outlying tissue not appearing in the medical image, wherein likely anatomical abnormalities in the outlying tissue near the imaged tissue border are detected by the CAD algorithm. In one example, the detected likely abnormalities in the outlying tissue are located within a first distance from the imaged tissue border, wherein the first distance corresponds to a spatial ambit of a neighborhood-based feature computed by the CAD algorithm.
1. A method for detecting anatomical abnormalities in a body part, the body part including (i) an imaged tissue region captured in a medical image by a medical imaging process, and (ii) a non-imaged tissue region not captured in the medical image by the medical imaging process, the non-imaged tissue region of the body part adjacently outlying the imaged tissue region of the body part with respect to the medical imaging process, the method being implemented by a computing device comprising at least one processor and memory, comprising:
receiving, by the computing device, the medical image; and
processing, by the computing device, the medical image to identify likely anatomical abnormalities centered in the non-imaged tissue region of the body part, wherein said processing the medical image comprises:
defining a first image array that includes both (i) the medical image and (ii) an extended region adjacently outlying the medical image, the extended region being spatially representative of the non-imaged tissue region, the extended region being devoid of image information captured in the medical imaging process; and
processing the first image array according to a computer-aided detection (CAD) algorithm, including computing at least one neighborhood-based CAD feature for at least one location in said extended region.
2. The method of claim 1 , further comprising:
displaying the medical image; and
for each identified likely anatomical abnormality in the non-imaged tissue region, displaying at least one annotation that communicates a location thereof relative to the imaged tissue region.
3. The method of claim 2 , further comprising displaying said extended region adjacent to the medical image, wherein said at least one annotation comprises a visible marker positioned on the extended region at a location corresponding to a computed location of the anatomical abnormality in the non-imaged tissue region.
4. The method of claim 3 , further comprising:
processing the medical image to identify likely anatomical abnormalities in the imaged tissue region; and
for each identified likely anatomical abnormality in the imaged tissue region, displaying at least one annotation that communicates a location thereof within the imaged tissue region.
5. The method of claim 1 , wherein said processing the first image array according to the CAD algorithm comprises:
identifying a first candidate location in said extended region; and
computing the at least one neighborhood-based CAD feature for the identified first candidate location.
6. The method of claim 5 , said processing the first image array according to the CAD algorithm further comprising:
identifying a plurality of additional candidate locations within the first image array;
computing said at least one neighborhood-based CAD feature for each of said additional candidate locations; and
applying a classification algorithm to said first and additional candidate locations, said classification algorithm being based at least upon said at least one neighborhood-based CAD feature.
7. The method of claim 5 , wherein said identifying the first candidate location comprises:
processing the first image array to identify candidate locations therein using a forward projection and accumulation method, wherein an accumulation region associated with said forward projection and accumulation method includes said extended region; and
thresholding said accumulation region.
8. The method of claim 5 , wherein said medical image comprises an x-ray mammogram, wherein said extended region corresponds to chest wall tissue not captured in the x-ray mammogram, and wherein said at least one neighborhood-based CAD feature comprises a mass stellateness feature.
9. The computer-implemented method of claim 8 , said defining said first image array comprising defining said extended region as extending a predetermined distance outward from a chest wall edge of the x-ray mammogram.
10. A system comprising:
a source of a medical x-ray image of a body part in digital form, the medical x-ray image having an image border, the body part comprising imaged tissue appearing inside the image border and non-imaged tissue adjacently outlying the imaged tissue, the non-imaged tissue not appearing in the medical x-ray image;
a processor processing said medical x-ray image digitally to detect suspected abnormalities in both the imaged tissue and the non-imaged tissue, said processor creating as a result of said processing an annotated map comprising symbols identifying the suspected abnormalities by type and location, wherein said processing comprises:
defining a first image array that includes both (i) the medical x-ray image and (ii) an extended region adjacently outlying the medical x-ray image, the extended region being spatially representative of the non-imaged tissue adjacently outlying the imaged tissue; and
processing the first image array according to a computer-aided detection (CAD) algorithm, including computing at least one neighborhood-based CAD feature for at least one location in said extended region;
and
a display coupled with said processor to display the medical x-ray image, the extended region, and the annotation map.
11. The system of claim 10 , wherein said processing the first image array according to the CAD algorithm further comprises:
identifying a first candidate location in the extended region; and
computing the at least one neighborhood-based CAD feature for the identified first candidate location.
12. The system of claim 11 , wherein said processing the first image array according to the CAD algorithm further comprises:
identifying a plurality of additional candidate locations within the first image array;
computing said at least one neighborhood-based CAD feature for each of said additional candidate locations; and
applying a classification algorithm to said first and additional candidate locations, said classification algorithm being based at least upon said at least one neighborhood-based CAD feature.
13. The system of claim 11 , wherein said identifying the first candidate location comprises:
processing the first image array to identify candidate locations therein using a forward projection and accumulation method, wherein an accumulation region associated with said forward projection and accumulation method includes said extended region; and
thresholding said accumulation region.
14. The system of claim 11 , wherein said medical x-ray image comprises an x-ray mammogram, wherein said extended region corresponds to chest wall tissue not captured in the x-ray mammogram, and wherein said at least one neighborhood-based CAD feature comprises a mass stellateness feature.
15. The system of claim 14 , said defining said first image array comprising defining said extended region as extending a predetermined distance outward from a chest wall edge of the x-ray mammogram.
16. A computer program product embodied in a computer-readable medium for detecting anatomical abnormalities in a body part, the body part including (i) an imaged tissue region captured in a medical image by a medical imaging process, and (ii) a non-imaged tissue region not captured in the medical image by the medical imaging process, the non-imaged tissue region of the body part adjacently outlying the imaged tissue region of the body part with respect to the medical imaging process, comprising:
computer code for receiving the medical image; and
computer code for processing the medical image to identify likely anatomical abnormalities centered in the non-imaged tissue region of the body part, comprising:
computer code for defining a first image array that includes both (i) the medical image and (ii) an extended region adjacently outlying the medical image, the extended region being spatially representative of the non-imaged tissue region, the extended region being devoid of image information captured in the medical imaging process; and
computer code for processing the first image array according to a computer-aided detection (CAD) algorithm, including computing at least one neighborhood-based CAD feature for at least one location in said extended region.
17. The computer program product of claim 16 , further comprising:
computer code for displaying the medical image; and
computer code for displaying, for each identified likely anatomical abnormality in the non-imaged tissue region, at least one annotation that communicates a location thereof relative to the imaged tissue region.
18. The computer program product of claim 17 , further comprising:
computer code for processing the medical image to identify likely anatomical abnormalities in the imaged tissue region; and
computer code for displaying, for each identified likely anatomical abnormality in the imaged tissue region, at least one annotation that communicates a location thereof within the imaged tissue region.
19. The computer program product of claim 16 , wherein said computer code for processing the first image array according to the CAD algorithm comprises:
computer code for identifying a first candidate location in said extended region; and
computer code for computing the least one neighborhood-based CAD feature for the identified first candidate location.
20. The computer program product of claim 19 , further comprising:
computer code for identifying a plurality of additional candidate locations within the first image array;
computer code for computing said at least one neighborhood-based CAD feature for each of said additional candidate locations; and
computer code for applying a classification algorithm to said first and additional candidate locations, said classification algorithm being based at least upon said at least one neighborhood-based CAD feature.