Method and apparatus for detecting a desired behavior in digital image data

• US 5,633,948 A
• Filed: 01/30/1995
• Issued: 05/27/1997
• Est. Priority Date: 11/30/1992
• Status: Expired due to Fees

• Current Assignee: Kegelmeyer W. Philip Jr
• Original Assignee: Kegelmeyer W. Philip Jr
• Inventors: Kegelmeyer, W. Philip Jr.
• Primary Examiner(s): COUSO, YON JUNG
• Application Number:
  US08/380,474
• Time in Patent Office: 848 Days
• Field of Search: 382/128, 382/132, 382/133, 382/168, 382/199, 382/131, 382/171, 382/170, 364/413.13
• US Class Current: 382/132
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound

Abstract

A method for detecting stellate lesions in digitized mammographic image data includes the steps of prestoring a plurality of reference images, calculating a plurality of features for each of the pixels of the reference images, and creating a binary decision tree from features of randomly sampled pixels from each of the reference images. Once the binary decision tree has been created, a plurality of features, preferably including an ALOE feature (analysis of local oriented edges), are calculated for each of the pixels of the digitized mammographic data. Each of these plurality of features of each pixel are input into the binary decision tree and a probability is determined, for each of the pixels, corresponding to the likelihood of the presence of a stellate lesion, to create a probability image. Finally, the probability image is spacially filtered to enforce local consensus among neighboring pixels and the spacially filtered image is output.

Claims

1. An apparatus for detecting a desired behavior in digital image data comprising:

input means for inputting digital image data;

accessing means for accessing, for each of a plurality of pixels of the digital image data, digital image data of each pixel and pixels surrounding each of the plurality of pixels in predetermined window dimensions;

orientation means for computing edge orientation values of the accessed digital image data, for each of the plurality of pixels;

histogramming means for histogramming, for each of the plurality of pixels, the computed edge orientation values; and

detecting means for detecting a presence of the desired behavior in each of the plurality of pixels for which the orientation histogram indicates a characteristic degree of orientation heterogeneity.

2. The apparatus of claim 1, wherein the input means includes a keyboard.

3. The apparatus of claim 1, wherein the input means includes a scanner for scanning an object to create the digital image data.

4. The apparatus of claim 1, wherein the detecting means further is for detecting an absence of the desired behavior in each of the plurality of pixels for which the orientation histogram fails to indicate a characteristic degree of orientation heterogeneity.

5. The apparatus of claim 4, further comprising:

indicating means for indicating, for each of the plurality of pixels, the detected presence or absence of the desired behavior.

6. The apparatus of claim 5, wherein the indicating means includes a display.

7. The apparatus of claim 5, wherein the indicating means includes a printer.

8. The apparatus of claim 1, wherein the desired behavior is an abnormality.

9. The apparatus of claim 3, wherein the desired behavior is an abnormality.

10. The apparatus of claim 8, wherein the abnormality is a stellate lesion and the digital image data is digitized mammographic image data.

11. The apparatus of claim 1, wherein the accessing means accesses predetermined window dimensions of 30 mm by 30 mm.

12. The apparatus of claim 1, wherein the orientation means includes an edge operator to compute edge orientation values.

13. The apparatus of claim 12, wherein the edge operator is a Sobel edge operator which computes an arc-tangent of a ratio of vertical and horizontal derivatives of each pixel.

27. The apparatus of claim 1, further comprising:

computing means for computing a standard deviation of each histogram, wherein the identifying means identifies a presence of the desired behavior based on the standard deviation of the orientation histogram indicating a characteristic degree of orientation heterogeneity.

28. The apparatus of claim 1, wherein the orientation means further converts the computed edge orientation values into discrete values prior to input to the histogramming means.

29. The apparatus of claim 4, further comprising:

computing means for computing a standard deviation of each histogram, wherein the identifying means identifies a presence of the desired behavior based on the standard deviation of the orientation histogram indicating a characteristic degree of orientation heterogeneity and an absence of the desired behavior based on the standard deviation of the orientation histogram failing to indicate a characteristic degree of orientation heterogeneity.

14. A method for detecting a desired behavior in digital image data, comprising the steps of:

(a) inputting digital image data;

(b) accessing, for each of a plurality of pixels of the digital image data, digital image data of each pixel and pixels surrounding each of the plurality of pixels in predetermined window dimensions;

(c) computing edge orientation values of the accessed digital image data, for each of the plurality of pixels;

(d) histogramming, for each of the plurality of pixels, the computed edge orientation values; and

(e) identifying a presence of the desired behavior in each of the plurality of pixels for which the orientation histogram indicates a characteristic degree of orientation heterogeneity.

15. The method of claim 14, wherein the digital input data of step (a) is input through the keyboard.

16. The method of claim 14, further comprising scanning an object to generate the input digital image data.

17. The method of claim 14, wherein the identifying step (e) further includes detecting an absence of the desired behavior in each of the plurality of pixels for which the orientation histogram fails to indicate a characteristic degree of orientation heterogeneity.

18. The method of claim 17, further comprising the step:

(f) indicating, for each of the plurality of pixels, the detected presence or absence of the desired behavior.

19. The method of claim 18, wherein the indicating step (f) includes displaying.

20. The method of claim 18, wherein the indicating step (f) includes printing.

21. The method of claim 14, wherein the desired behavior is an abnormality.

22. The method of claim 16, wherein the desired behavior is an abnormality.

23. The method of claim 21, wherein the abnormality is a stellate lesion and the digital image data is digitized mammographic image data.

24. The method of claim 14, wherein the accessing of step (b) includes accessing predetermined window dimensions of 30 mm by 30 mm.

25. The method of claim 14, wherein step (c) includes computing edge orientation values with an edge operator.

26. The method of claim 25, wherein the edge operator is a sobel edge operator which computes an arc-tangent of a ratio of vertical and horizontal derivatives of each pixel in step (c).

30. The apparatus of claim 14, further comprising the step of:

(f) computing a standard deviation of each histogram, wherein the identifying of step (e) identifies a presence of the desired behavior based on the standard deviation of the orientation histogram indicating a characteristic degree of orientation heterogeneity.

31. The method of claim 14, wherein the computed edge orientation values are converted into discrete values prior to histogramming in step (d).

32. The method of claim 17, further comprising the step of:

(f) computing a standard deviation of each histogram of step (d), wherein the identifying of step (e) identifies a presence of the desired behavior based on the standard deviation of the orientation histogram indicating a characteristic degree of orientation heterogeneity and an absence of the desired behavior based on the standard deviation of the orientation histogram failing to indicate a characteristic degree of orientation heterogeneity.