Method and system for segmentation of medical images

• US 6,978,039 B2
• Filed: 12/13/2001
• Issued: 12/20/2005
• Est. Priority Date: 12/13/2001
• Status: Expired due to Fees

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Cline, Harvey Ellis, Ludke, Siegwalt
• Primary Examiner(s): Werner, Brian
• Assistant Examiner(s): LAVIN, CHRISTOPHER L
• Application Number:
  US09/683,321
• Publication Number:
  US 20030113003A1
• Time in Patent Office: 1,468 Days
• Field of Search: 382/128, 382/131, 382/132, 382/164, 382/171, 382/173, 382/180, 382/199, 382/205, 128/922
• US Class Current: 382/128
• CPC Class Codes: G06T 2207/10072 Tomographic images
• CPC Class Codes: G06T 2207/20064 Wavelet transform [DWT]
• CPC Class Codes: G06T 2207/20101 Interactive definition of point of interest, landmark or seed
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/168 involving transform domain methods
• CPC Class Codes: G06T 7/187 involving region growing; involving region merging; involving connected component labelling
• CPC Class Codes: Y10S 128/922 including image analysis

Abstract

A method and system for segmenting three-dimensional (3D) medical images containing an object of interest are provided. The method comprises generating a plurality of successive layers of fixed radius spheres about a circumference of a sphere containing at least one seed point placed within the object of interest when a plurality of respective voxels contained within the spheres exceed a selected threshold. The generation of the layers is repeated until no further voxels contained within an outer surface of each respective layer exceed the selected threshold or a stop seed point is encountered. The layers form a segmented representation of the object of interest.

Claims

1. A method for segmenting three-dimensional (3D) medical images containing an object of interest comprising:

generating a plurality of successive layers of fixed radius spheres about a circumference of a sphere containing at least one seed point placed within the object of interest when a plurality of respective voxels contained within the spheres exceed a selected threshold; and

,

repeating generation of the layers until no further voxels contained within an outer surface of each respective layer exceed the selected threshold, the layers forming a segmented representation of the object of interest.

2. The method of claim 1 wherein the radius is selected in accordance with a desired radius of curvature of the segmented representation.

3. The method of claim 1 wherein the layers of spheres within the segmented representation of the object is wholly contained within the object of interest.

4. The method of claim 1 wherein the three-dimensional images are acquired by at least one of magnetic resonance imaging (MRI), computed tomography (CT), position emission tomography (PET), and x-ray systems.

5. A method for segmenting three-dimensional medical images containing an object of interest comprising:

placing at least one seed point in the object of interest;

generating at least one spherical wavelet having a selected radius about the seed point;

comparing a plurality of voxels contained with the wavelet with a selected threshold;

generating a plurality of additional spherical wavelets circumferentially about the wavelet when the plurality of voxels exceed the selected threshold;

the plurality of spherical wavelets having an advancing surface about an outside circumference;

repeating the comparing and generating steps for each of the plurality of voxels contained within the additional spherical wavelets on the advancing surface to generate layers of spherical wavelets to form a segmented representation of the object of interest, the repeating step ending when no further voxels contained within the advancing surface exceed the selected threshold.

6. The method of claim 5 wherein the radius is selected in accordance with a desired radius of curvature of the segmented representation.

7. The method of claim 5 wherein the three-dimensional images are acquired by at least one of magnetic resonance imaging (MRI), computed tomography (CT), position emission tomography (PET), and x-ray systems.

8. A method for segmenting three-dimensional (3D) medical images containing an object of interest comprising:

placing at least one start seed point within the object of interest;

placing at least one stop seed point outside the object of interest;

generating successive layers of spheres about a circumference of a sphere containing at the least one start seed point when a plurality of respective voxels contained within the spheres exceed an selected initial threshold;

repeating generation of the layers until no further voxels contained within an outer surface of each respective layer exceed the selected initial threshold or until the at least one stop seed is encountered to form a segmented representation of the object of interest; and

,

adjusting the selected threshold in response to encountering the stop seed point.

9. The method of claim 8 wherein the spheres have a selected radius based on desired radius of curvature.

10. The method of claim 8 wherein the three-dimensional images are acquired by at least one of magnetic resonance imaging (MRI), computed tomography (CT), position emission tomography (PET), and x-ray systems.

11. The method of claim 8 wherein the adjusting step is performed with a binary search and the selected initial threshold is an average of a minimum and maximum intensity of the object.