System and method for toboggan-based object detection in cutting planes

• US 7,912,294 B2
• Filed: 05/25/2006
• Issued: 03/22/2011
• Est. Priority Date: 05/27/2005
• Status: Active Grant

• Current Assignee: Siemens Healthineers AG (Siemens AG)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Liang, Jianming, Cathier, Pascal, Bogoni, Luca
• Primary Examiner(s): Rashid; David P
• Application Number:
  US11/440,780
• Publication Number:
  US 20070036406A1
• Time in Patent Office: 1,762 Days
• Field of Search: 382/195
• US Class Current: 382/195
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/155 involving morphological operators

Abstract

A system and method for toboggan-based object detection in cutting planes are provided. A method for detecting an object in an image includes: determining a region of interest (ROI) in the image; determining a toboggan potential for each image element in the ROI; extracting a plurality of cutting planes from the ROI; and performing a tobogganing in the cutting planes to form a toboggan cluster to determine a location of the object, wherein image elements inside the toboggan cluster are stored in a cluster-member list, image elements on an outer-border of the toboggan cluster are stored in an outer-border list and image elements on an inner-border of the toboggan cluster are stored in an inner-border list.

Claims

1. A program storage device readable by a computer, tangibly embodying a program of instructions executable by the computer to perform method steps for detecting an object in a three-dimensional (3D) medical image, the method comprising the steps of:

determining a region of interest (ROI) in the 3D medical image;

determining a toboggan potential for each image element in the ROI;

extracting a plurality of two-dimensional (2D) cutting planes from the ROI; and

performing a tobogganing in the cutting planes to form a toboggan cluster to determine a location of the object,

wherein image elements inside the toboggan cluster are stored in a cluster-member list, image elements on an outer-border of the toboggan cluster are stored in an outer-border list and image elements on an inner-border of the toboggan cluster are stored in an inner-border list

wherein the step of tobogganing further comprises;

(a) determining a concentration location, wherein if the concentration location is in the ROI, the tobogganing comprises;

(b) adding the concentration location to the cluster-member list and adding neighbor image elements of the concentration location to an active-elements list;

(c) selecting an image element with a minimum or maximum toboggan potential as a current image element;

(d) determining a sliding direction of the current image element;

(e) if the current image element slides to an image element in the cluster-member list, adding the current image element to the cluster-member list and adding neighbor image elements of the current image element to the active-elements list if the neighbor image elements of the current image element are not already in the active-elements list;

(f) if the current image element does not slide to an image element in the cluster-member list, adding the current image element to the outer-border list and adding neighbor image elements of the current image element in the cluster-member list to the inner-border list;

(g) repeating steps (c-f) until the active-elements list is empty,

wherein the toboggan cluster includes image elements in the cluster-member list and inner-border list.

2. The computer readable program storage device of claim 1, wherein the ROI is an organ or part of an organ including a spherical or ellipsoidal object.

3. The computer readable program storage device of claim 1, wherein the cutting planes are 2D slices of the ROI.

4. The computer readable program storage device of claim 1, further comprising:

extracting a property of the object in the cutting planes using the image elements stored in the outer-border and inner-border lists.

5. The computer readable program storage device of claim 4, further comprising:

identifying the object as a polyp candidate based on the extracted property.

6. A program storage device readable by a computer, tangibly embodying a program of instructions executable by the computer to perform method steps for identifying an object in a three-dimensional (3D) computed tomography (CT) image of a colon, the method comprising the steps of:

determining a region of interest (ROI) in the 3D image, wherein the ROI is determined by thresholding the image to obtain an air region, dilating the air region and subtracting the air region from the dilated air region;

determining a toboggan potential for each image element in the ROI;

extracting a plurality of 2D cutting planes from the ROI; and

performing a tobogganing in the cutting planes to form a toboggan cluster to identify the object, wherein if a concentration location is in the ROI, the tobogganing comprises;

finding a concentration point in the cutting planes; and

expanding from the concentration point by including neighbor image elements of the concentration point in a cluster based on the potentials of the image elements,

wherein image elements inside the cluster are stored in a cluster-member list, image elements on an outer-border of the cluster are stored in an outer-border list and image elements on an inner-border of the cluster are stored in an inner-border list.

7. The computer readable program storage device of claim 6, wherein the ROI is a colon wall.

8. The computer readable program storage device of claim 6, wherein the toboggan potential for each image element in the ROI is acquired when the CT image data is acquired.

9. The computer readable program storage device of claim 6, wherein 13 cutting planes are extracted from the ROI.

10. The computer readable program storage device of claim 6, wherein finding a concentration point comprises:

(a) selecting a point in the image;

(b) finding a neighbor with a maximum or minimum potential with respect to the selected point;

(c) sliding to the neighbor point; and

repeating steps (a-c) until a point is reached that cannot slide to a neighbor point.

11. A program storage device readable by a computer, tangibly embodying a program of instructions executable by the computer to perform method steps for identifying an object in a three-dimensional (3D) computed tomography (CT) image of a colon, the method comprising the steps of:

determining a region of interest (ROI) in the 3D image;

determining a toboggan potential for each image element in the ROI;

extracting a plurality of 2D cutting planes from the ROI; and

performing a tobogganing in the cutting planes to form a toboggan cluster to identify the object, wherein if a concentration location is in the ROI, the tobogganing comprises;

finding a concentration point in the cutting planes; and

expanding from the concentration point by including neighbor image elements of the concentration point in a cluster based on the potentials of the image elements,

wherein image elements inside the cluster are stored in a cluster-member list, image elements on an outer-border of the cluster are stored in an outer-border list and image elements on an inner-border of the cluster are stored in an inner-border list; and

extracting a property of the object in the cutting planes, wherein the property is a circularity measure calculated using the image elements stored in the cluster-member, outer-border and inner-border lists.

12. The computer readable program storage device of claim 11, further comprising:

identifying whether the object is a polyp based on the circularity measure.