System for digital bowel subtraction and polyp detection and related techniques

• US 6,947,784 B2
• Filed: 04/06/2001
• Issued: 09/20/2005
• Est. Priority Date: 04/07/2000
• Status: Active Grant

• Current Assignee: The General Hospital Corporation (Mass General Brigham, Inc.)
• Original Assignee: The General Hospital Corporation (Mass General Brigham, Inc.)
• Inventors: Zalis, Michael E.
• Primary Examiner(s): ROBINSON, DANIEL LEON
• Application Number:
  US09/828,268
• Publication Number:
  US 20020097320A1
• Time in Patent Office: 1,628 Days
• Field of Search: 600/429, 600/425, 600/407, 600/426, 600/428, 600/424, 600/410, 600/378, 600/21, 600/25, 600/26, 600/27, 600/427, 324/307, 324/308, 324/309, 378/21, 378/4, 378/22, 378/23, 378/24, 378/25, 378/26, 378/27
• US Class Current: 600/425
• CPC Class Codes: G06T 2207/10081 Computed x-ray tomography [CT]
• CPC Class Codes: G06T 2207/30032 Colon polyp
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/155 involving morphological operators
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

A system for performing a virtual colonoscopy includes a system for generating digital images, a storage device for storing the digital images, a digital bowel subtraction processor coupled to receive images of a colon from the storage device and for removing the contents of the colon from the image and an automated polyp detection processor coupled to receive images of a colon from the storage device and for detecting polyps in the colon image.

Claims

1. A system for performing a virtual colonoscopy comprising:

(a) a digital image generating system;

(b) a storage device, coupled to said digital image generating system, said storage device for storing digital images; and

(c) a digital bowel subtraction processor coupled to receive images of a colon from said storage device, said digital bowel subtraction processor for processing the received digital images of the colon to digitally remove the contents of the colon from the image.

2. The system of claim 1 further comprising:

an automated polyp detection processor coupled to receive images of a colon from said storage device, said automated polyp detection processor for processing the received digital images of the colon to detect polyps in the colon image.

3. The system of claim 2 wherein said automated polyp detection processor processes images from which bowel contents have been removed by said digital bowel subtraction processor.

4. The system of claim 1 wherein said digital bowel subtraction processor corresponds to a raster digital bowel subtraction processor.

5. The system of claim 4 wherein said raster digital bowel subtraction processor comprises:

a threshold circuit for assessing whether absolute threshold values have been crossed and the ratio at which they have been crossed;

an analysis kernal to scan across images and apply threshold values in a predetermined logic sequence;

means for selecting a starting pixel;

means for selecting a direction to apply a threshold;

a threshold application means; and

a pixel reset means.

6. The system of claim 1 wherein said digital bowel subtraction processor corresponds to a gradient processor digital bowel subtraction processor.

7. The system of claim 6 wherein said gradient processor digital bowel subtraction processor comprises:

a threshold circuit for defining a soft tissue threshold value, an air threshold value and a bowel threshold value;

a mask processor for marking elements above the bone threshold value;

a gradient processor for applying a air and bowel threshold values to appropriate regions wherein said gradient processor forms a first mask to capture a first shoulder region in a region corresponding to a transition between air and bowel;

a dilation processor for identifying a second shoulder region in the region corresponding to a transition between air and bowel;

a subtractor for subtracting the combined masks from the gradient processor to leave a desired image.

8. The system of claim 2 wherein said automated polyp detection processor comprises:

a segmentor to extract pixels in a region which includes a boundary between soft tissue and air in the colon;

a rolling ball processor; and

a vector processor for computing a ratio along a travel path and comparing the travel path to a threshold value.

9. The system of claim 2 wherein said automated polyp detection processor comprises:

a convolution processor for performing a convolution between a test polyp and a region under test, to indicates the similarities between the region under test and the test polyp.

10. Thy system of claim 9 wherein said convolution processor generates a correlation matrix and wherein the convolution polyp detector further comprises a filter to identify portions of the region undertest which have relatively high correlation values.

11. The system of claim 2 wherein said automated polyp detection processor comprises a distance processor for aligning a distance template over the image and for computing a plurality of distance values, each of the plurality of distance values corresponding to a distance between a selected point on the distance template and a point on a bowel perimeter with each of the plurality of distance values computed using a different point on the bowel perimeter.

12. The system of claim 11 wherein said automated polyp detection processor further comprises a comparison circuit for comparing each of the plurality of distance values.

13. The system of claim 12 wherein said comparison circuit includes means for comparing each of the plurality of distance values by subtracting the values to provide a relative distance value.

14. The system of claim 11 wherein said automated polyp detection processor further comprises a segmentor to extract pixels from the image to provide a region which includes a boundary between soft tissue and air in the colon and wherein said distance processor for aligns the distance template over the region provided by said segmentor.

15. The system of claim 1 further comprising:

a detection system coupled to said digital bowel subtraction processor for detecting, in the digital images of the colon having contents digitally removed therefrom, regions having one or more characteristics similar to the characteristics of a polyp;

an indicator system for indicating on the images those regions detected by said detection system.

16. The system of claim 1 further comprising a mucosa insertion processor for providing a mucosa layer.

17. The system of claim 16 wherein said mucosa insertion processor comprises:

a gradient processor for applying a gradient operator to an image from which bowel contents have been digitally extracted to identify pixels corresponding to a boundary between bowel wall and non-bowel wall regions in the image; and

a smoothing processor for smoothing the boundary over a predetermined width along the boundary.

18. The system of claim 17 wherein said smoothing processor comprises a filter which filters the pixels which form the boundary.

19. The system of claim 18 wherein said filter is provided having a constrained Gaussian filter characteristic.

20. The system of claim 19 wherein the predetermined width corresponds to one of:

a width of ten pixels;

a width in the range of two to five pixels; and

a width of three pixels.

21. The system of claim 1 further comprising a contrast agent adapted for ingestion by a patient prior said digital image generating system generating one or more bowel images of the patient.

22. The system of claim 16 wherein said digital bowel subtraction processor comprises:

a threshold circuit for applying a threshold function to the image of the colon; and

a gradient analyzer for performing a gradient analysis of the image of the colon to define a first region corresponding to a wall region and a second region corresponding to a bowel contents region.

23. The system of claim 6 wherein said gradient processor digital bowel subtraction processor comprises:

a transition region identifier for identifying a portion of an image corresponding to an image transition, region having a first portion which corresponds to bowel contents, a second portion which corresponds to bowel wall and a third portion disposed between the first and second portions which corresponds to a transition between the bowel contents and the bowel wall;

a shoulder region identifier for identifying first and second shoulder regions in the image transition region;

means for dilating a boundary of the first shoulder region by applying a first threshold function to the bowel image;

a gradient processor for applying a gradient to the bowel image to define a first region corresponding to a wall region and a second region corresponding to a bowel contents region; and

means for digitally removing the second region from the bowel image.

24. The system of claim 23 wherein said a shoulder region identifier comprises:

means for dilating pixels around a portion of the image transition region corresponding to the second shoulder region;

a comparator for comparing each of the pixel values to a gradient threshold value;

means, responsive to a pixel value being less than the gradient threshold value, for setting that pixel value equal to a first predetermined value corresponding to a first one of a contents region and a wall region; and

means, responsive to a pixel value being greater than the gradient threshold value, for setting that pixel value equal to a second predetermined value corresponding to a second one of the contents region and the wall region.

25. The system of claim 16 wherein said digital bowel subtraction processor comprises:

a pixel analysis map processor for generating a first pixel analysis map (PAM) from the image;

means for stepping the first PAM across the image in a raster pattern;

means for locating an air region in the image by examining a predetermined set of elements defined by the PAM;

means for searching an area around the air element to locate a bowel region adjacent the air region; and

means for subtracting the bowel contents from the image.

26. The system of claim 1 further comprising:

means for representing an image of the colon by an array of pixels; and

means for selecting pixels in the image which represent bowel contents wherein said means for selecting pixels comprises;

means for forming a pixel analysis map comprising an array of pixels, with predetermined ones of the pixels in the map having a spatial relationship and a predetermined threshold value;

means for applying the pixel analysis map to the image;

a comparison processor for comparing the values of the predetermined ones of the pixels in the pixel analysis map to pixels in the image and indicating whether the corresponding pixel in the image represents one of bowel contents or bowel wall; and

means for setting values of pixels in the image corresponding to bowel contents to one of a pixel value corresponding to air and a pixel value corresponding to a bowel.

27. The system of claim 26 wherein said pixel analysis map processor comprises:

means for defining a plurality of central pixels at the center of the pixel analysis map;

means for defining a plurality of boundary pixels disposed about the plurality of central pixels; and

means for defining a plurality of outer boundary pixels disposed about the plurality of boundary.

28. The system of claim 27 wherein said comparison processor comprises:

means for selecting a first pixel of the plurality of central pixels;

means for computing a difference value between a pixel in the image which is in a position adjacent to the position of the first pixel and an outer boundary pixel proximate the first pixel; and

means, responsive to the difference value being above a gradient threshold, for setting the values of all pixels between the first pixel and the outer boundary pixel to a predetermined value.

29. The system of claim 28 wherein the predetermined value corresponds to one of an air value and a bowel value.

30. The system of claim 2 wherein said automated polyp detection processor comprises:

means for obtaining a test structure template from a first CT image;

means for selecting a portion of a second CT image;

means for performing a convolution between the test structure template and the selected portion of the second CT image to produce a correlation matrix; and

means for identifying regions in the second image having values with a high correlation to characteristics of a polyp.

31. The system of claim 30 wherein said means for obtaining a test structure template comprises;

means for obtaining a first CT image which includes a structure of the type to be detected; and

means for excising the structure from the CT image to provide the test structure.

32. The system of claim 30 wherein said means for identifying regions in the second image comprises:

a filter for filtering values which are below a predetermined correlation threshold value; and

means for designating a region of an image as having a likelihood of containing a structure with a shape which is similar to the shape of the test structure.

33. The method of claim 31 wherein the test structure template corresponds to a polyp template and the structure of the type to be detected in the first CT image corresponds to a polyp.

34. The system of claim 2 wherein said automated polyp detection processor comprises:

means for moving a test element along a boundary of a bowel;

means for tracking a path defined by the movement of the test element;

means for detecting changes in direction of the test element in the path;

means for classifying bowel features based upon the path of the test element and the changes in direction of the test element.

35. The system of claim 34 wherein said means for detecting changes in direction of the test element in the path comprises:

means for computing changes in slope of the path defined by the test element;

means for marking a point at each location at which the direction of the slope of the path

means for changes;

computing distances between each of the marked points; and

means for computing a ratio value using each of the distances.

36. The system of claim 34 wherein said means for classifying bowel features comprises means for identifying bowel features using the ratio value computed by said means for computing a ratio value.

37. The system of claim 35 wherein the test element is provided having a circular shape;

and a spherical shape.

38. A system of claim 2 wherein said automated polyp detection processor comprises:

means for applying a template to a region of an image which contains a bowel perimeter;

means for computing a plurality of distances between the template and the bowel perimeter, each of the plurality of distances corresponding to a distance from at least one point on the template to one of a plurality of points on the bowel perimeter; and

means for determining whether the distances are equal.

39. The system of claim 38 further comprising means for identifying the image region as containing a structure having a shape which is similar to the shape of the template in response to the distances being equal.

40. The system of claim 38 wherein applying a template to a bowel perimeter comprises:

means for segmenting a bowel image to produce to an image from which pixels which do not form the bowel perimeter have been filtered.

41. The system of claim 38 wherein said means for computing a plurality of distances between the template and the bowel perimeter comprises means for computing the distances between points on a perimeter of the template and points on the perimeter of the bowel.

42. The system of claim 38 wherein said means for determining whether the distances are equal comprises means for computing a standard deviation between the distances.