Computer-aided method for image feature analysis and diagnosis in mammography

• US 5,740,268 A
• Filed: 09/29/1995
• Issued: 04/14/1998
• Est. Priority Date: 04/29/1994
• Status: Expired due to Term

• Current Assignee: ARCH Development Corporation (University of Chicago)
• Original Assignee: ARCH Development Corporation (University of Chicago)
• Inventors: Yoshida, Hiroyuki, Ema, Takehiro, Nishikawa, Robert M., Doi, Kunio
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Tran, Phuoc
• Application Number:
  US08/536,253
• Time in Patent Office: 928 Days
• Field of Search: 382/130, 382/132, 382/199, 382/202, 382/203, 382/260, 382/263, 382/264, 382/262, 382/266, 382/270, 382/275, 364/413.13, 364/413.14, 364/413.19, 364/413.23
• US Class Current: 382/132
• CPC Class Codes: A61B 6/5258 involving detection or reduction of artifacts or noise
• CPC Class Codes: G06T 2207/10116 X-ray image
• CPC Class Codes: G06T 2207/20064 Wavelet transform [DWT]
• CPC Class Codes: G06T 2207/30068 Mammography; Breast
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06V 10/52 Scale-space analysis, e.g. wavelet analysis multi-scale boundary representations G06V10/42
• CPC Class Codes: Y10S 128/925 Neural network

Abstract

A method for automated detection of abnormal anatomic regions, wherein a mammogram is digitized to produce a digital image and the digital image is processed using local edge gradient analysis and linear pattern analysis in addition to feature extraction routines to identify abnormal anatomic regions. Noise reduction filtering and pit-filling/spike-removal filtering techniques are also provided. Multiple difference imaging techniques are also used in which difference images employing different filter characteristics are obtained and processing results logically OR'"'"'ed to identify abnormal anatomic regions. In another embodiment the processing results with and without noise reduction filtering are logically AND'"'"'ed to improve detection sensitivity. Also, in another embodiment the wavelet transform is utilized in the identification and detection of abnormal regions. The wavelet transform is preferably used in conjunction with the difference imaging technique with the results of the two techniques being logically OR'"'"'ed.

Claims

1. A method for automated detection in mammography of an abnormal anatomic region, comprising:

obtaining a digital image of an object including said anatomic region;

processing said digital image to identify in said digital image locations which correspond to potential abnormal regions;

determining an edge gradient for each of the locations identified in said processing step;

comparing each edge gradient with at least one threshold, comprising comparing each edge gradient with a predetermined number; and

eliminating locations identified in said processing step from consideration as an abnormal region based on a result of said comparing step, comprising eliminating those regions having an edge gradient exceeding said predetermined number.

2. The method of claim 1, wherein:

said comparing step comprises comparing each edge gradient with a varying threshold which varies inversely as a function of the average pixel value for said location; and

said eliminating step comprises eliminating those locations having an edge gradient less than said varying threshold.

3. The method of claim 2, wherein:

said processing step comprises identifying locations of microcalcifications and locations of microcalcification clusters;

said determining step comprises determining edge gradients for the locations of said microcalcifications and for the locations of said microcalcification clusters; and

said comparing step comprises comparing the edge gradients determined for said microcalcification locations and for said microcalcification cluster locations with respective thresholds; and

said eliminating step comprises eliminating locations based on the comparing of microcalcification location edge gradients with respective thresholds and based on the comparing of microcalcification cluster edge gradients with respective thresholds.

4. The method of claim 3, comprising:

said processing step identifying locations of microcalcification signals and microcalcification clusters;

said determining step comprising,

determining, for each microcalcification signal, an edge gradient value for each of plural selected pixels adjacent said microcalcification signal;

determining, for each microcalcification signal, an average value of said edge gradient values of adjacent pixels to be the edge gradient of the respective microcalcification signal;

averaging, in each microcalcification cluster, the edge gradients and the average pixel values of a selected number of microcalcification signals having the largest edge gradients in the cluster and using the averaged edge gradient as the edge gradient of said cluster and the averaged pixel value as the average pixel value of said cluster in said comparing and eliminating steps.

5. The method according to claim 4, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified in said processing steps;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

6. The method according to claim 5, wherein said step of determining a degree of linearity, comprises:

placing templates around pixels corresponding to microcalcification signals;

determining local gradients for corresponding portions of said templates at plural orientations around each microcalcification signal;

comparing the local gradients for each template orientation with a predetermined local gradient threshold and determining for each template orientation the percentage of determined local gradients which exceed said predetermined local gradient threshold; and

determining said degree of linearity as being the largest percentage of the determined percentages of local gradients.

7. The method according to claim 2, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified in said processing steps;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

8. The method according to claim 7, wherein said step of determining a degree of linearity, comprises:

placing templates around pixels corresponding to microcalcification signals;

determining local gradients for corresponding portions of said templates at plural orientations around each microcalcification signal;

comparing the local gradients for each template orientation with a predetermined local gradient threshold and determining for each template orientation the percentage of determined local gradients which exceed said predetermined local gradient threshold; and

determining said degree of linearity as being the largest percentage of the determined percentages of local gradients.

9. The method of claim 1, wherein locations identified in said processing step correspond to microcalcification signals and said determining step comprises:

determining, for each microcalcification signal, an edge gradient value for each of plural selected pixels adjacent said microcalcification signal; and

determining, for each microcalcification signal, an average value of said edge gradient values of adjacent pixels to be the edge gradient of the respective microcalcification signal.

10. The method according to claim 9, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified in said processing steps;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

11. The method according to claim 10, wherein said step of determining a degree of linearity, comprises:

placing templates around pixels corresponding to microcalcification signals;

determining local gradients for corresponding portions of said templates at plural orientations around each microcalcification signal;

comparing the local gradients for each template orientation with a predetermined local gradient threshold and determining for each template orientation the percentage of determined local gradients which exceed said predetermined local gradient threshold; and

determining said degree of linearity as being the largest percentage of the determined percentages of local gradients.

12. The method according to claim 1, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified in said processing steps;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

13. The method according to claim 12, wherein said step of determining a degree of linearity, comprises:

placing templates around pixels corresponding to microcalcification signals;

determining local gradients for corresponding portions of said templates at plural orientations around each microcalcification signal;

comparing the local gradients for each template orientation with a predetermined local gradient threshold and determining for each template orientation the percentage of determined local gradients which exceed said predetermined local gradient threshold; and

determining said degree of linearity as being the largest percentage of the determined percentages of local gradients.

14. A method for automated detection in mammography of an abnormal anatomic region, comprising:

obtaining a digital image of an object including said anatomic region;

processing said digital image to identify in said digital image locations which correspond to potential abnormal regions;

determining a degree of linearity for each of the locations identified in said processing step;

comparing the determined degree of linearity for each location with a predetermined linearity threshold; and

elimination locations having a degree of linearity exceeding said linearity threshold from consideration as an abnormal region.

15. The method according to claim 14, wherein said step of determining a degree of linearity, comprises:

placing templates around pixels corresponding to microcalcification signals;

determining local gradients for corresponding portions of said templates at plural orientations around each microcalcification signal;

comparing the local gradients for each template orientation with a predetermined local gradient threshold and determining for each template orientation the percentage of determined local gradients which exceed said predetermined local gradient threshold; and

determining said degree of linearity as being the largest percentage of the determined percentages of local gradients.

16. A method for automated detection of an abnormal anatomic region, comprising:

obtaining a digital image of an object including said anatomic region;

first filtering said digital image using a first set of filters having a first transfer function to produce a first difference image based on a difference between a signal-enhanced and a signal suppressed image of said digital image;

second filtering said digital image using a second set of filters having a second transfer function different from said first transfer function to produce a second difference image based on a difference between a signal-enhanced and a signal suppressed image of said digital image;

processing said first difference image to identify in said first difference image first locations which correspond to potential abnormal regions;

processing said second difference image to identify in said second difference image second locations which correspond to potential abnormal regions;

logically OR'"'"'ing the first and second locations to identify as candidate abnormal regions all those first and second locations which are separated by a predetermined distance or more; and

,

processing the candidate abnormal regions to identify abnormal regions from among said candidate abnormal regions.

17. The method of claim 16, wherein:

said first filtering step comprises spatially filtering using a first set of spatial filters having a first optical transfer factor; and

said second filtering step comprises spatially filtering using a second set of spatial filters having a second optical transfer factor.

18. The method of claim 17, wherein at least one of said first and second filtering steps comprises:

filtering using a spatial filter having an optical transfer factor in the frequency range of 1.0 to 1.5 cycles/mm.

19. The method of claim 18, wherein said step of processing the candidate abnormal regions comprises:

performing predetermined feature extraction routines to identify microcalcifications;

identifying microcalcification clusters based on the identified microcalcifications; and

identifying abnormal regions based on the identified microcalcification clusters.

20. The method of claim 18, wherein said steps of processing said first difference image and processing said second difference image each comprise:

performing global thresholding; and

performing local thresholding.

21. The method of claim 20, wherein said step of processing the candidate abnormal regions comprises:

performing predetermined feature extraction routines to identify microcalcifications;

identifying microcalcification clusters based on the identified microcalcifications; and

identifying abnormal regions based on the identified microcalcification clusters.

22. The method of claim 20, wherein said steps of processing said first difference image and processing said second difference image in each case further comprise:

performing predetermined feature extraction routines to identify microcalcifications; and

identifying microcalcification clusters based on the identified microcalcifications.

23. The method of claim 16, wherein:

said first filtering step and said second filtering step comprise linearly filtering said digital image.

24. The method of claim 16, wherein:

said first filtering step and said second filtering step comprise morphological filtering of said digital image.

25. The method of claim 16, wherein said step of processing the candidate abnormal regions comprises:

performing predetermined feature extraction routines to identify microcalcifications;

identifying microcalcification clusters based on the identified microcalcifications; and

identifying abnormal regions based on the identified microcalcification clusters.

26. The method of claim 16, wherein said steps of processing said first difference image and processing said second difference image in each case comprise:

performing global thresholding; and

performing local thresholding.

27. The method of claim 26, wherein said step of processing the candidate abnormal regions comprises:

performing predetermined feature extraction routines to identify microcalcifications;

identifying microcalcification clusters based on the identified microcalcifications; and

identifying abnormal regions based on the identified microcalcification clusters.

28. The method of claim 26, wherein said steps of processing said first difference image and processing said second difference image in each case further comprise:

performing predetermined feature extraction routines to identify microcalcifications; and

identifying microcalcification clusters based on the identified microcalcifications.

29. The method of claim 16, wherein said step of processing the candidate abnormal regions comprises:

determining an edge gradient for each of the candidate abnormal regions;

comparing each edge gradient determined in said determining step with at least one threshold; and

eliminating candidate abnormal regions identified in said processing step from consideration as an abnormal region based on a result of said comparing step.

30. The method of claim 29, wherein:

said comparing step comprises comparing each edge gradient with a predetermined number; and

said eliminating step comprises eliminating those candidate abnormal regions having an edge gradient exceeding said predetermined number.

31. The method of claim 30, wherein:

said comparing step comprises comparing each edge gradient with a varying threshold which varies inversely as a function of the average pixel value for said location; and

said eliminating step comprises eliminating those candidate abnormal regions having an edge gradient less than said varying threshold.

32. The method of claim 31, wherein said candidate abnormal region processing step comprises:

identifying locations of microcalcifications and locations of microcalcification clusters;

determining edge gradients for the locations of said microcalcifications and for the locations of said microcalcification clusters; and

comparing the edge gradients determined for said microcalcification locations and for said microcalcification cluster locations with respective thresholds; and

eliminating locations based on the comparing of microcalcification location edge gradients with respective thresholds and based on the comparing of microcalcification cluster edge gradients with respective thresholds.

33. The method according to claim 32, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

34. The method according to claim 30, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

35. The method according to claim 31, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

36. The method of claim 29, wherein:

said comparing step comprises comparing each edge gradient with a varying threshold which varies inversely as a function of the average pixel value for said location; and

said eliminating step comprises eliminating those candidate abnormal regions having an edge gradient less than said varying threshold.

37. The method according to claim 36, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

38. The method according to claim 29, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.

39. The method according to claim 16, wherein:

said determining step comprises determining a degree of linearity for each of said locations identified;

said comparing step comprises comparing the degree of linearity determined for each location with a predetermined linearity threshold; and

said eliminating step comprises eliminating from consideration as abnormal regions locations identified in said processing step and having a linearity factor exceeding said predetermined linearity threshold.