OBTAINING BREAST DENSITY MEASUREMENTS AND CLASSIFICATIONS

20160256126

15/035743

November 10, 2014

Breast density measurements are used to perform Breast Imaging Reporting and Data System (BI-RADS) classification during breast cancer screenings. The accuracy of breast density measurements can be improved by quantitatively processing digital mammographic images. For example, breast segmentation may be performed on a mammographic image to isolate the breast tissue from the background and pectoralis tissue, while a breast thickness adjustment may be performed to compensate for decreased tissue thickness near the skin line of the breast. In some instances, BI-RADS density categorization may consider the degree to which dense tissue is dispersed throughout the breast. A breast density dispersion parameter can also be obtained using quantitative techniques, thereby providing objective BI-RADS classifications that are less susceptible to human error.

1. A method for breast density computation comprising: performing, by a processor, breast segmentation on a mammographic image to obtain breast and pectoral muscle segmentation maps; performing breast thickness correction on the mammographic image to obtain a flattened breast image; and computing a breast density measurement in accordance with the breast and pectoral muscle segmentation maps and the flattened breast image.

2. The method of claim 1, wherein the breast and pectoral muscle segmentation maps classify pixels in the mammographic image as representing breast tissue, pectoral tissue, or background.

3. The method of claim 2, wherein the breast and pectoral muscle segmentation maps indicate a first boundary line in the mammographic image, the first boundary line separating pixels representing pectoralis tissue from pixels representing breast tissue.

4. The method of claim 3, wherein the breast and pectoral muscle segmentation maps indicate a second boundary line in the mammographic image, the second boundary line separating pixels representing a breast tissue from pixels representing background.

5. The method of claim 4, wherein performing breast thickness correction on the mammographic image to obtain the flattened breast image comprises: computing an estimate of fatty tissue as a function of distance from the second boundary line; and subtracting the estimate of fatty tissue from the mammographic image to obtain the flattened image.

6. The method of claim 1, wherein computing the breast density measurement in accordance with the breast and pectoral muscle segmentation maps and the flattened breast image comprises: performing breast density estimation to obtain an initial density map; performing breast density segmentation on the initial density map to obtain a final density map; and computing the breast density measurement in accordance with the final density map.

7. The method of claim 6, further comprising performing breast density estimation to obtain the initial density map by: filtering the mammographic image using multiple scales to obtain filtered outputs, wherein each filtered output corresponds to a different scale; analyzing the filtered outputs to obtain signals within corresponding regions of interest (ROI), wherein each corresponding ROI is a function of the scale used to obtain the filtered output; and combining the signals in accordance with a weighted value to generate the initial density map.

8. The method of claim 7, wherein filtering the mammographic image using multiple scales to obtain filtered outputs comprises: filtering the mammographic image in a spatial domain.

9. The method of claim 7, wherein filtering the mammographic image using multiple scales to obtain filtered outputs comprises: filtering the mammographic image in a frequency domain.

10. The method of claim 7, wherein filtering the mammographic image using multiple scales to obtain filtered outputs comprises: filtering the mammographic image in a joint spatial-frequency domain.

11. The method of claim 6, wherein performing breast density segmentation on the initial density map to obtain the final density map comprises: refining the initial density map to obtain a refined density map, wherein refining the initial density map includes extracting fatty tissue from the initial density map and filling in holes in parenchyma regions of the initial density map; segmenting a flattened breast image to obtain a flattened-image density map; and combining the flattened-image density map with the refined density map to obtain the final density map.

12. The method of claim 11, wherein combining the flattened-image density map with the refined density map to obtain the final density map comprises: taking a logical AND of the flattened-image density map and the refined density map to generate the final density map.

13. The method of claim 6, wherein computing the breast density measurement in accordance with the final density map comprises: calculating a dense area in accordance with the final density map; calculating a breast area in accordance with a breast segmentation map; and calculating the breast density measurement in accordance with the dense area and the breast area.

14. The method of claim 13, wherein calculating the breast density measurement in accordance with the dense area and the breast area comprises: dividing the dense area by the breast area to obtain a ratio; and multiplying the ratio by one hundred percent to obtain the breast density measurement.

15. An apparatus comprising: a processor; and a computer readable storage medium storing programming for execution by the processor, the programming including instructions to: perform breast segmentation on a mammographic image to obtain breast and pectoral muscle segmentation maps; perform breast thickness correction on the mammographic image to obtain a flattened breast image; and compute a breast density measurement in accordance with the breast and pectoral muscle segmentation maps and the flattened breast image.

16. The apparatus of claim 15, wherein the breast and pectoral muscle segmentation maps classify pixels in the mammographic image as representing breast tissue, pectoral tissue, or background.

17. A method for breast density classification, the method comprising: computing, by a processor, a breast density ratio for a mammographic image, the breast density ratio corresponding to a percentage of dense breast tissue in a breast depicted by the mammographic image; computing a dispersion measurement value for the mammographic image, the dispersion measurement value corresponding to a distribution of the dense tissue in the breast depicted by the mammographic image; and assigning a breast density classification to the mammographic image in accordance with the breast density ratio and the dispersion measurement value.

18. The method of claim 17, wherein computing the dispersion measurement value for the mammographic image comprises: performing breast density estimation to obtain a breast density map; and computing the dispersion measurement in accordance with the breast density map.

19. The method of claim 17, wherein computing the dispersion measurement value for the mammographic image comprises: performing breast density estimation to obtain an initial breast density map; performing breast density segmentation on the initial breast density map to obtain a final breast density map; and computing the dispersion measurement in accordance with the final breast density map.

20. An apparatus comprising: a computer readable storage medium storing programming for execution by the processor, the programming including instructions to: compute a breast density ratio for a mammographic image, the breast density ratio corresponding to a percentage of dense breast tissue in a breast depicted by the mammographic image; compute a dispersion measurement value for the mammographic image, the dispersion measurement value corresponding to a distribution of the dense tissue in the breast depicted by the mammographic image; and assign a breast density classification to the mammographic image in accordance with the breast density ratio and the dispersion measurement value.

61; 06

edspap.20160256126