Noninvasive medical imaging system and method for the identification and 3-D display of atherosclerosis and the like

• US 4,945,478 A
• Filed: 11/06/1987
• Issued: 07/31/1990
• Est. Priority Date: 11/06/1987
• Status: Expired due to Fees

• Current Assignee: University of Virginia
• Original Assignee: The Center For Innovative Technology
• Inventors: Carman, Charles S., Ayers, Carlos R., Brookeman, James R., Mugler, John P. III, Merickel, Michael B.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): Kilby, Steve
• Application Number:
  US07/117,508
• Time in Patent Office: 998 Days
• Field of Search: 364/413.19, 364/413.22, 364/413.13, 364/521, 364/522, 340/729, 382/9, 382/6, 382/15, 382/22, 324/309
• US Class Current: 382/131
• CPC Class Codes: A61B 5/055 involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
• CPC Class Codes: G01R 33/56 Image enhancement or correction, e.g. subtraction or averaging techniques , e.g. improvement of signal-to-noise ratio and resolution
• CPC Class Codes: G01R 33/5608 Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels image data processing or generation, in general G06T
• CPC Class Codes: G06T 7/00 Image analysis

Abstract

There is disclosed an image processing, pattern recognition and computer graphics system and method for the noninvasive identification and evaluation of atheroscelerosis using multidimensional Magnetic Resonance Imaging (MRI). Functional information, such as plaque tissue type, is combined with structure information, represented by the 3-D vessel and plaque structure, into a single composite 3-D display. The system and method is performed with the application of unsupervised pattern recognition techniques and rapid 3-D display methods appropriate to the simultaneous display of multiple data classes. The results are a high information content display which aids in the diagnosis and analysis of the atherosclerotic disease process, and permits detailed and quantitative studies to assess the effectiveness of therapies, such as drug, exercise and dietary regimens.

Claims

1. A noninvasive imaging system for the display of a mammalian soft tissue live body part such as blood vessels or similar soft tissue for evaluation of soft tissue, using information obtained from multiple sequenced image slices produced by a body scan such as MRI and PET comprising:

an electronic digital storage means for storing said information;

a computer processing means for processing said information from said storage means;

a graphics display station means for displaying an image resulting from said processed information;

said computer processing means including;

means for determining the boundary of said soft tissue body part from said image slices;

means for extracting said soft tissue body part;

means for clustering and determining the different soft tissue types of interest in said body part; and

means for creating a data base for said extracted soft tissue body part including the different soft tissue types with their location and volume for displaying in contrasting delineation on said graphics display station whereby said display shows an extracted view of said body part with said different soft tissues shown delineated in their volumetric quantities and shape at the correct location in said body part.

2. The noninvasive imaging system of claim 1, wherein said image slices are is made by an MRI and each of said image slices is a set of images one of which is produced by T₁ emphasis.

3. The noninvasive imaging system of claim 2, wherein each of said images in said image slice sets are brought into registration with one another prior to determining the boundary of said body part.

4. The noninvasive imaging system of claim 3, wherein:

said MRI scan utilizes a surface intensity coil for increasing the signal to noise ratio and improving the spatial resolution; and

said noninvasive imaging system includes means for reducing the intensity gradient created by said surface intensity coil.

5. The noninvasive imaging system of claim 4, wherein means included for normalizing the intensity between each of said images in each of said image slice sets.

6. The noninvasive imaging system of claim 5, wherein means are provided for calculating a single average image from all of said images produced by T₁ emphasis.

7. The noninvasive imaging system of claim 6, wherein means are provided to convolve said average image by using a convolution mask.

8. The noninvasive imaging system of claim 7, wherein means are provided for thresholding the brightest values of said convolved image.

9. The noninvasive imaging system of claim 8, wherein means are provided for producing the centroid of said thresholded convolved image.

10. The noninvasive imaging system of claim 9, wherein means are provided for extracting a subimage from each of said T₁ emphasized image slices by using said centroid as a central point with said subimages being large enough to include said body part.

11. The noninvasive imaging system of claim 10, wherein means are provided for determining the approximate boundary of said body part by first creating a histogram of each of said subimage slices followed by producing a threshold analysis to determine said approximate boundary.

12. The noninvasive imaging system of claim 11, wherein means are provided for removing artifacts from said approximate boundary and reconstructing the boundary into a new boundary.

13. The noninvasive imaging system of claim 12, wherein means are provided for extracting the body part of interest from each slice by utilizing said new boundary as a mask.

14. The noninvasive imaging system of claim 1, wherein means are provided for classification which utilizes unsupervised analysis of said body part by finding the natural groupings or pixel clusters in said image slices.

15. The noninvasive imaging system of claim 14, wherein means are provided for carrying out said unsupervised analysis by use of CAIC criterion and assignment of each pixel with a cluster which has the closest Mahalanobis distance.

16. The non-invasive imaging system of claim 15, wherein means are provided for removing all pixels from further consideration which are assigned to clusters insufficiently large.

17. The noninvasive imaging system of claim 16, wherein means are provided for merging and splitting pixel clusters until the best set of clusters are found.

18. The noninvasive imaging system of claim 17, wherein means are provided for labelling each cluster with its tissue class name;

labelling each pixel by assigning to them the label now associated with their respective cluster and using said labeled pixels to form a sequence of tissue maps.

19. The noninvasive imaging system of claim 1, wherein means are provided for labeling the pixel in each of said image slices with its tissue class and using said labeled pixels to form a sequence of tissue maps.

20. The noninvasive imaging system of claim 19, wherein means are provided for determining the volumes of each tissue type.

21. The noninvasive imaging system of claim 20, wherein said tissue type volumes are determined by using a quadtree data structure.

22. The noninvasive imaging system of claim 20, wherein means are provided for determining the cross-sectional areas of each tissue in each slice.

23. A method for noninvasive display of a mammalian soft tissue live body part using information obtained from multiple image slices produced by a body scan such as MRI and PET comprising the following steps:

produce multiple sequenced image slices by noninvasively scanning a mammalian body;

segment the entire scene represented by the image slices and identify the boundary of said soft tissue body part;

extract said body part by extracting a region enclosed within said boundary;

cluster the different soft tissue types of interest in said soft tissue body part; and

create a data base for the different soft tissue types including their location and volume in contrasting delineation for displaying on a graphics display station whereby said display shows an extracted view of said body part with said different soft tissues shown delineated in their volumetric quantities and shape at the correct location in said body part.

24. The noninvasive 3-D display method of claim 23, wherein said body scan is made by MRI and the production of each of said images slices includes production of a set of images one of which is produced by T₁ emphasis.

25. The noninvasive 3-D display method of claim 24, which includes the steps of bringing into registration each of said images in said image slice sets with one another prior to determining the boundary of said body part.

26. The noninvasive 3-D display method of claim 25, which includes the the steps of making said MRI scan utilizing a surface intensity coil for increasing the signal to noise ratio and improving the spatial resolution;

reducing the intensity gradient created by said surface intensity coil.

27. The noninvasive 3-D display method of claim 26, which includes the step of normalizing the intensity between each of said images in each of said image slice sets.

28. The noninvasive 3-D display method of claim 27, which includes the step of calculating a single average image from all of said images produced by T₁ emphasis.

29. The noninvasive 3-D display method of claim 28, which includes the step of convolving said average image by using a convolution mask.

30. The noninvasive 3-D display method of claim 29, which includes the step of thresholding the brightest values of said convolved image.

31. The noninvasive 3-D display method of claim 30, which includes the step of producing the centroid of said thresholded convolved image.

32. The noninvasive 3-D display method of claim 31, which includes the step of extracting a subimage from each of said T₁ emphasized image slices by using said centroid as a central point with said subimage being large enough to include said body part.

33. The noninvasive 3-D display method of claim 32, which includes the steps of determining the approximate boundary of said body part by first creating a histogram of each of said subimage slices followed by making a threshold analysis to determing said approximate boundary.

34. The noninvasive 3-D display method of claim 33, which include the steps of removing artifacts from said approximate boundary and reconstructing the boundary into a new boundary.

35. The noninvasive 3-D display method of claim 34, which includes the step of extracting the body part of interest from each slice by utilizing said new boundary as a mask.

36. The noninvasive 3-D display method of claim 23, which includes the step of extracting from said slices said body part of interest.

37. The noninvasive 3-D display method of claim 36, which includes the step of classification utilizing the unsupervised analysis of said body part by finding the natural groupings or pixel clusters in said image slices.

38. The noninvasive 3-D display method of claim 37, which includes the step of carrying out said unsupervised analysis by use of CAIC criterion and assignment of each pixel with a cluster which has the closest Mahalanobis distance.

39. The noninvasive 3-D display method of claim 38, which includes the step of removing all pixels from further consideration which are assigned to clusters insufficiently large.

40. The noninvasive 3-D display method of claim 39, which includes the steps of merging and splitting pixel clusters until the best set of clusters are found.

41. The noninvasive 3-D display method of claim 40, which includes the steps of labelling each cluster with its tissue class name;

labelling each pixel by assigning to them the label now associated with their respective cluster and using said labeled pixels to form a sequence of tissue maps.

42. The noninvasive 3-D display method of claim 37, which includes the steps of labeling the pixel in each of said image slices with its tissue class and using said labeled pixels to form a sequence of tissue maps.

43. The noninvasive 3-D display method of claim 42, which includes the step of determining the volumes of each tissue type.

44. The noninvasive 3-D display method of claim 43, which includes the step of determining by said tissue type volumes using a quadtree data structure.

45. The noninvasive 3-D display method of claim 44, which includes the step of determining the cross-sectional areas of each tissue in each slice.

46. A noninvasive imaging system for the 3-D display of atherosclerosis of a living human blood vessel using information obtained from multiple image slices produced by a MRI body scan comprising:

an electronic digital storage means for storing said information;

a computer processing means for processing said information from said storage means;

a graphics display station means for displaying a 3-D image resulting from said processed information;

said computer processing means including;

means for taking multiple sequenced image slice sets wherein each set includes three images representing proton emphasis, T₁ emphasis and T₂ emphasis and normalizing the images so that they represent a common intensity,

means for segmenting said normalized image for delineating the boundary of and extracting said living human blood vessel;

classification means for using unsupervised clustering with no a priori knowledge to define the label lipid tissue, fibrous tissue and calcified tissue and assign different colors to each tissue; and

means for creating a 3-D data base for the tissue types including their location and volume in different colors for displaying on said graphics display stations whereby said display shows a 3-D extracted view of the vessel with said lipid tissue, fibrous tissue and calcified tissue, to the extent present, shown in different colors in their volumetric quantities and shape at the correct location in said vessel.

47. The noninvasive imaging system of claim 46, wherein means are included for bringing each of said images in said image slice sets into registration with one another prior to determining the boundary of said body part.

48. The noninvasive imaging system of claim 47, wherein:

said MRI scan utilizes a surface intensity coil for increasing the signal to noise ratio and improve the spatial resolution; and

means are included for reducing the intensity gradient created by said surface intensity coil.

49. The noninvasive imaging system of claim 48, wherein means are provided for normalizing the intensity between each of said images in each of said image slice sets.

50. The noninvasive imaging system of claim 49, wherein means are provided for calculating a single average image from all of said images produced by T₁ emphasis.

51. The noninvasive imaging system of claim 50, wherein means are provided to convolve said average image by using a convolution mask.

52. The noninvasive imaging system of claim 51, wherein means are provided for thresholding the brightest values of said convolved image.

53. The noninvasive imaging system of claim 52, wherein means are provided for producing the centroid of said thresholded convolved image.

54. The noninvasive imaging system of claim 53, wherein means are provided for extracting a subimage from each of said T₁ emphasized imaged slices by using said centroid as a central point with said subimages being large enough to include said body part.

55. The noninvasive imaging system of claim 54, wherein means are provided for determining the approximate boundary of said body part by first creating a histogram of each of said subimage slices followed by producing a threshold analysis to determine said approximate boundary.

56. The noninvasive imaging system of claim 55, wherein means are provided for removing artifacts from said approximate boundary and reconstructing the boundary into a new boundary.

57. The noninvasive imaging system of claim 56, wherein means are provided for extracting the body part of interest from each slice by utilizing said new boundary as a mask.

58. The noninvasive imaging system of claim 46, wherein means are provided for carrying out said unsupervised clustering by use of CAIC criterion and assignment of each pixel with a cluster which has the closest Mahalanobis distance.

59. The non-invasive imaging system of claim 58, wherein means are provided for removing all pixels from further consideration which are assigned to clusters insufficiently large.

60. The noninvasive imaging system of claim 59, wherein means are provided for merging and splitting pixel clusters until the best set of clusters are found.

61. The noninvasive imaging system of claim 60, wherein means are provided for labelling each cluster with its tissue class name;

labelling each pixel by assigning to them the label now associated with their respective cluster and using said labeled pixels to form a sequence of tissue maps.

62. The noninvasive imaging system of claim 61, wherein means are provided for determining the volumes of each tissue type.

63. The noninvasive imaging system of claim 62, wherein said tissue type volumes are determined by using a quadtree data structure.

64. The noninvasive imaging system of claim 63, wherein means are provided for determining the cross-sectional areas of each tissue in each slice.