Method and apparatus for classification of pixels in medical imaging

• US 20050259857A1
• Filed: 02/09/2005
• Published: 11/24/2005
• Est. Priority Date: 05/21/2004
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Muller, Serge, Iordache, Razvan, Jeunehomme, Fanny
• Granted Patent: US 7,298,884 B2
• Time in Patent Office: Days
• US Class Current: 382/132
• CPC Class Codes: G06T 2207/30068 Mammography; Breast
• CPC Class Codes: G06T 7/0012 Biomedical image inspection

Abstract

An apparatus and method for medical imaging, particularly for mammography, wherein a body organ, such as a breast, is exposed to X-rays and the X-rays are collected after attenuation through the object. The recorded attenuations are processed and displaying a result of this processing in the form of a representation of an image of the object. The processing of the recorded attenuations form includes automatic classification of zones of the breast into pathological or non-pathological classes. The automatic classification takes into account at least one classification input into the apparatus in advance in association with data that can be collected by the apparatus, and using this prior classification as a reference in order to produce a classification of the same type if there is similarity between the collected data and the data associated with this reference classification.

Claims

1. An apparatus for medical imaging comprising:

means for exposing an object to radiation;

means for collecting the radiation after attenuation through the object;

means for processing recorded attenuations; and

means for displaying a result of the processing in the form of a representation on an image of the object

wherein the means for processing the recorded attenuations forms means for automatic classification of zones of the object into pathological or non-pathological classes, the means for automatic classification taking account of at least one classification input into the apparatus in advance in association with data that can be collected by the apparatus, and using this prior classification as a reference in order to produce a classification of the same type if there is similarity between the collected data and the data associated with this reference classification.

2. The apparatus according to claim 1 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each corresponding to the value of a grey level at a given instant of a contrast kinetic at the pixel considered.

3. The apparatus according to claim 1 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each representing a parameter among the maximum slope of a contrast variation, the value of a maximum contrast reached, a hold duration of the contrast at its maximum value, at the pixel considered.

4. The apparatus according to claim 2 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each representing a parameter among the maximum slope of a contrast variation, the value of a maximum contrast reached, a hold duration of the contrast at its maximum value, at the pixel considered.

5. The apparatus according to claim 1 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

6. The apparatus according to claim 2 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

7. The apparatus according to claim 3 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

8. The apparatus according to claim 4 wherein the means for automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

9. The apparatus according to claim 1 wherein the means for automatic classification takes account of at least one manual classification of a current image zone of the object as reference classification;

and

automatically establishes the same classification in case of similarity with other data collected on the same object.

10. The apparatus according to claim 1 wherein the means for automatic classification includes means for learning;

and

carrying out recording a collection of reference information comprising several reference classifications associated with data that can be collected by the apparatus.

11. The apparatus according to claim 1 comprising:

means for enabling a user to confirm or contradict an automatic classification made by the apparatus; and

means for taking account of this confirmation or contradiction in order to incorporate the results of this confirmation or contradiction as a reference classification.

12. The apparatus according to claim 1 wherein the means for classification identifies incoherent spatial variations in the classification, and modifying the classification of some locations in case of such incoherent spatial variations.

13. The apparatus according to claim 1 comprising means for accounting of one item of data among the number of changes in the sign of the variation in grey levels, the age of the object, weight or medical data, when creating a classification.

14. A method for medical imaging comprising:

exposing an object to radiation;

collecting the radiation after attenuation through the object;

processing recorded attenuations comprising automatic classifying of zones of the object into pathological or non-pathological classes, the automatic classification taking account of at least one classification input in advance in association with data that can be collected, and using this prior classification as a reference in order to produce a classification of the same type if there is similarity between the collected data and the data associated with this reference classification; and

displaying a result of the processing in the form of a representation on an image of the object the process.

15. The method according to claim 14 wherein the automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each corresponding to the value of a grey level at a given instant of a contrast kinetic at the pixel considered.

16. The method according to claim 15 wherein the automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each corresponding to the value of a grey level at a given instant of a contrast kinetic at the pixel considered.

17. The method according to claim 16 wherein the automatic classification makes a classification of pixels of an image of the object at the same location in a several dimensional space, the dimensions of this space each representing a parameter among the maximum slope of a contrast variation, the value of a maximum contrast reached, a hold duration of the contrast at its maximum value, at the pixel considered.

18. The method according to claim 14 wherein the automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

19. The method according to claim 15 wherein the automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

20. The method according to claim 16 wherein the automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

21. The method according to claim 17 wherein the automatic classification makes a classification of pixels of an image of the object at the same location, in a space with at least two dimensions, these two dimensions each representing the measured signal for two different radiation energies at the pixel considered.

22. The method according to claim 14 wherein the automatic classification (10, 20) takes account of at least one manual classification of a current image zone of the object as reference classification;

and

automatically establishes the same classification in case of similarity with other data collected on the same object.

23. The method according to claim 14 wherein the automatic classification includes learning;

and

carrying out recording a collection of reference information comprising several reference classifications associated with data that can be collected by the apparatus.

24. The method according to claim 14 comprising:

enabling a user to confirm or contradict an automatic classification; and

taking into account of this confirmation or contradiction in order to incorporate the results of this confirmation or contradiction as a reference classification.

25. The method according to claim 14 wherein the classification identifies incoherent spatial variations in the classification, and modifying the classification of some locations in case of such incoherent spatial variations.

26. The method according to claim 14 comprising:

accounting for one item of data among the number of changes in the sign of the variation in grey levels, the age of the object, weight or medical data, when creating a classification.

27. A computer program comprising program code means for implementing the method according to claim 14.

28. A computer program product comprising a computer useable medium having computer readable program code means embodied in the medium, the computer readable program code means implementing the method according to claim 14.

29. An article of manufacture for use with a computer system,.the article of manufacture comprising a computer readable medium having computer readable program code means embodied in the medium, the program code means implementing of the method according to claim 14.

30. A program storage device readable by a machine tangibly embodying a program of instructions executable by the machine to perform the method according to claim 14.