Image enhancement method and system for fiducial-less tracking of treatment targets

• US 7,840,093 B2
• Filed: 12/15/2008
• Issued: 11/23/2010
• Est. Priority Date: 06/30/2004
• Status: Active Grant

• Current Assignee: Accuray Incorporated
• Original Assignee: Accuray Incorporated
• Inventors: Kuduvalli, Gopinath, Fu, Dongshan
• Primary Examiner(s): Le; Vu
• Assistant Examiner(s): TSAI, TSUNG YIN
• Application Number:
  US12/334,789
• Publication Number:
  US 20090091567A1
• Time in Patent Office: 708 Days
• Field of Search: 382/254
• US Class Current: 382/294
• CPC Class Codes: A61B 6/505 for diagnosis of bone
• CPC Class Codes: A61B 6/5235 combining images from the same or different ionising radiation imaging techniques, e.g. PET and CT
• CPC Class Codes: A61N 2005/1061 using an x-ray imaging system having a separate imaging source
• CPC Class Codes: A61N 2005/1062 using virtual X-ray images, e.g. digitally reconstructed radiographs [DRR]
• CPC Class Codes: A61N 5/1049 for verifying the position of the patient with respect to the radiation beam
• CPC Class Codes: G06T 2207/10116 X-ray image
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 5/30 Erosion or dilatation, e.g. thinning
• CPC Class Codes: G06T 5/94 based on local image properties, e.g. for local contrast enhancement
• CPC Class Codes: G06T 7/33 using feature-based methods
• CPC Class Codes: G06T 7/35 using statistical methods
• CPC Class Codes: G06T 7/38 Registration of image sequences

Abstract

A method and system are presented for enhancing one or more images of an object, so as to increase the visibility within the images of one or more structures within the object. The object may be an anatomical region of a patient, and may include one or more reference structures, for example skeletal structures or vertebral structures, and one or more treatment targets, for example tumors or lesions. An operator, for example a top-hat filter operator, selects at least a first neighborhood and a second neighborhood within the images. The operator selects within each neighborhood one or more pixels having an optimal pixel value, and eliminates the remaining pixels in these neighborhoods. When the operator is applied to the selected neighborhoods, only the pixels having the greatest pixel values remain in the selected neighborhoods, and the remaining pixels are eliminated in the selected neighborhoods. As a result, desired features can be located and enhanced in the images.

Claims

1. A method of enhancing an image of an object so as to increase the visibility in the image of at least one skeletal structure within the object, wherein the image is characterized by a plurality of pixels, each pixel having an associated pixel value that represents the image intensity of a corresponding unit volume of the object, the method comprising:

generating a two-dimensional (2D) reconstructed image from a three-dimensional (3D) scan of the object, wherein the 2D reconstructed image includes the skeletal structure;

generating a 2D x-ray image of the object, wherein the 20×

-ray image includes the skeletal structure;

for each of the 2D reconstructed image and the 2D x-ray image

a) selecting at least a first neighborhood and a second neighborhood within the image,

b) constructing an operator configured to select, within the first and second neighborhoods, one or more pixels having an optimal pixel value, and to eliminate the remaining pixels in the neighborhoods;

c) applying the operator to the selected neighborhoods so that only the pixels having the optimal pixel values remain in the selected neighborhoods, and the remaining pixels in the selected neighborhoods are eliminated to produce an enhanced image; and

registering the enhanced 2D x-ray image with the enhanced 2D reconstructed image by deriving a transformation between a coordinate system of the enhanced 2D xray image and a different coordinate system of the enhanced 2D reconstructed image.

2. A method in accordance with claim 1, wherein the first and the second neighborhoods are characterized by different sizes.

3. A method in accordance with claim 2, further comprising the act of constructing the operator so that the operator is configured to perform the following acts:

1) select, from each neighborhood, the pixel having the largest pixel value;

2) compare the pixel selected from the smaller of the first and the second neighborhood with the pixel selected from the larger of the first and the second neighborhood; and

3) if the pixel selected from the smaller neighborhood is greater than the pixel selected from the larger neighborhood by a threshold amount, keep the pixel;

otherwise, eliminate the pixel.

4. A method in accordance with claim 1, wherein the operator comprises a non-linear operator.

5. A method in accordance with claim 4, wherein the operator comprises a non-linear top-hat filter operator.

6. A method in accordance with claim 1, wherein the act of selecting the neighborhoods comprises:

selecting the at least two different neighborhoods within the image so that at least a portion of the structure within the object is included in at least one of the neighborhoods.

7. The method of claim 1, wherein the 2D reconstructed image from a 3D scan is a digitally reconstructed radiograph from a computed tomography scan.

8. An apparatus for enhancing an image of an object so as to increase the visibility in the image of at least one skeletal structure within the object, the image being characterized by a plurality of pixels, each pixel having an associated pixel value that represents the image intensity of a corresponding unit volume of the object, the apparatus comprising:

a) a neighborhood selector configured to select at least a first neighborhood and a second neighborhood within the image for each of a two-dimensional (20) reconstructed image from a three-dimensional (3D) scan of the object and a 2D x-ray image of the object, wherein each image of the object includes the skeletal structure;

b) an operator generator to construct an operator configured to select, within the first and second neighborhoods of each image, one or more pixels having an optimal pixel

value, and to eliminate the remaining pixels in the selected neighborhoods; and

c) a controller to apply the operator to the selected neighborhoods of each image so that only the pixels having the optimal pixel values remain in the selected neighborhoods, and the remaining pixels in the selected neighborhoods are eliminated to produce enhanced images, and to register the enhanced 2D x-ray image with the enhanced 2D reconstructed image by deriving a transformation between a coordinate system of the enhanced 2D x-ray image and a different coordinate system of the enhanced 2D reconstructed image.

9. An apparatus in accordance with claim 8, wherein the operator comprises a non linear operator.

10. The apparatus of claim 8, wherein the 2D reconstructed image from ascan is a digitally reconstructed radiograph from a computed tomography scan.