System and method for assessing cancer risk

1. A method for assessing risk of developing cancer, comprising:

receiving an image of a person;

analyzing the image to determine texture values representing characteristics of the image;

calculating a unique identifier for the person using the determined texture values, the texture values for calculating the unique identifier including skewness, coarseness, and contrast;

obtaining risk factors associated with the person;

determining a probabilistic assessment of the person developing cancer based on the determined texture values and the obtained risk factors; and

storing the probabilistic assessment.

2. The method of claim 1 , wherein the received image is a breast image.

3. The method of claim 2 , wherein the breast image is a digital mammography (DM) image.

4. The method of claim 2 , wherein the breast image is a digital breast tomosynthesis (DBT) image.

5. The method of claim 1 , wherein the analyzing step includes:

selecting a region of interest within the image; and

analyzing the image within the selected region of interest to obtain the values representing characteristics of the image.

6. The method of claim 5 , wherein the selecting step comprises:

manually selecting the region of interest within the image.

7. The method of claim 5 , wherein the selecting step comprises:

automatically selecting the region of interest within the image by comparing the image to other images to establish anatomic correspondences, the anatomic correspondences established by segmentation, statistical correlations and texture metrics.

8. The method of claim 7 , wherein the image is a mediolateral oblique (MLO) view of a breast image and the automatically selecting includes:

locating a pectoral muscle;

locating a nipple;

drawing a line perpendicular from the pectoral muscle to the nipple; and

selecting the region of interest at a point along the perpendicular line.

9. The method of claim 7 , wherein the image is a craniocaudal (CC) view of a breast image and the automatically selecting includes:

locating a side of the image;

locating a nipple;

drawing a line perpendicular from the side of the image to the nipple; and

selecting the region of interest at a point along the perpendicular line.

10. The method of claim 5 , wherein the determined texture values representing the characteristics of the image further include energy of the region of interest (ROI), ratios between the pixel values in the ROI and pixel values in a segmented portion of the ROI.

11. The method of claim 5 , wherein a computer is programmed to determine the probabilistic assessment by developing a logistic regression model based on at least one image feature of the region of interest and the risk factors associated with the person and determining the probabilistic estimation of the person developing cancer based on the logistic regression model.

12. The method of claim 11 , wherein the at least one image feature is a texture value.

13. The method of claim 5 , wherein the selecting step comprises:

comparing the image with other images to establish anatomic correspondences; and

mapping a region identifier onto the image to select the region of interest based on the anatomic correspondences.

14. The method of claim 5 , wherein a computer is programmed to determine the probabilistic assessment by developing a linear regression model based on at least one image feature of the selected region of interest and breast density and determining the probabilistic estimation of the person developing cancer based on the linear regression model.

15. The method of claim 14 , wherein the at least one image feature is a texture value.

16. The method of claim 1 , further comprising:

developing a linear regression model based on texture features of the region of interest and signal to noise ratio (SNR) of the image; and

determining image quality based on the linear regression model.

17. The method of claim 1 , wherein a computer is programmed to determine the probabilistic assessment of the person developing cancer.

18. A method for developing a logistic regression model for a person developing breast cancer, comprising:

receiving an image of breast tissue for a person;

selecting a region of interest within the image;

analyzing the region of interest to determine texture values of the image;

calculating a unique identifier for the person using the determined texture values, wherein the texture values for calculating the unique identifier include skewness, coarseness, and contrast;

developing a logistic regression model, by a computer programmed to develop the logistic regression model based on the texture values in the region of interest and risk factors associated with the person for use in determining a probabilistic assessment; and

storing the logistic regression model.

19. The method of claim 18 , wherein the logistic regression model is developed utilizing the risk factors, the risk factors computed from person information.

20. The method of claim 19 , wherein the risk factors include Gail factors.

21. The method of claim 20 , wherein the Gail factors comprise one or more of: current age of the person, age when the person started menstruating, previous breast biopsies of the person, age of person at first birth, and persons family history of breast cancer in first-degree relatives.

22. The method of claim 18 , wherein the texture values further include at least one of ratios between the pixel values in the region of interest (ROI) and pixel values in a segmented portion of the ROI or energy.

23. A system for assessing risk of developing cancer, comprising:

means for receiving an image of a person;

means for analyzing the image to determine texture values representing characteristics of the image;

means for calculating a unique identifier for the person using the determined texture values, the texture values for calculating the unique identifier including skewness, coarseness, and contrast;

means for obtaining risk factors associated with the person;

means for determining a probabilistic assessment of the person developing cancer based on the determined texture values and the obtained risk factors; and

means for storing the probabilistic assessment.