Medical image processing device, treatment system and medical image processing method

Patent metadata

Field	Value
Publication number	US-9786054-B2
Application number	US-201414478577-A
Country	US
Kind code	B2
Filing date	Sep 5, 2014
Priority date	Sep 30, 2013
Publication date	Oct 10, 2017
Grant date	Oct 10, 2017

Abstract

According to an embodiment, a medical image processing device includes a processor, and a memory. The memory that stores processor-executable instructions that, when executed by the processor, cause the processor to execute acquiring a first perspective image of a subject viewed in a first direction; setting a first region and a second region on the first perspective image, the first region including a first group of pixels around a target pixel, the second region including a second group of pixels, the second group including a pixel not included in the first group; calculating a likelihood of the target pixel, wherein the likelihood increases as a difference between pixel values included in the first group decreases and a difference between pixel values of the first group and the second group increases; and detecting a position of an object in the subject based on the likelihood.

First claim

What is claimed is:

1. A medical image processing device comprising:
a processor; and
a memory that stores processor-executable instructions that, when executed by the processor, cause the processor to execute:
acquiring a first perspective image of a subject viewed in a first direction;
setting a first region and a second region on the first perspective image, the first region including a first group of pixels around a target pixel, the second region including a second group of pixels, the second group including a pixel not included in the first group;
calculating a likelihood of a pixel value for the target pixel based on pixel values for pixels included in the first group and the second group, wherein the likelihood increases as a difference between pixel values for pixels included in the first group decreases and a difference between pixel values for pixels included in the first group and the second group increases;
detecting a position of an object in the subject based on the likelihood; and
outputting information related to the position which is usable by a controller for controlling irradiation of a treatment beam to the object.

2. The device according to claim 1 , wherein the first region has a shape depending on a shape of the object.

3. The device according to claim 2 , wherein the acquiring acquires object information regarding a shape and a size of the object, and a projection matrix that is used to project three-dimensional coordinates onto the perspective image;
the calculating calculates a shape and a size of an image of the object appearing on the perspective image using the object information and the projection matrix; and
the setting sets the first region with the shape and the size of the image.

4. The device according to claim 1 , wherein the likelihood is a correlation ratio between the pixel values for pixels of the first group and the second group.

5. The device according to claim 1 , wherein the detecting detects that the position of the object is detected by detecting a target pixel having the likelihood whose value satisfies a predetermined condition.

6. The device according to claim 1 , wherein the calculating acquires region information indicating a third region where an image of the object is expected to exist in the perspective image; and calculates, for the target pixel, the likelihood in the third region indicated by the region information.

7. The device according to claim 1 , wherein the calculating estimates a third region where an image of the object is expected to exist in the perspective image based on a predetermined motion model; and calculates, for the target pixel, the likelihood in the third region.

8. The device according to claim 7 , wherein the detecting sets a prior probability to the third region, a value of the prior probability becoming larger the closer to a position to which the image moves based on the predetermined motion model; and detects a position of the object based on the likelihood and the prior probability of the target pixel.

9. The device according to claim 1 , wherein
the processor detects a plurality of objects in the subject,
the calculating estimates a third region where each of images of the plurality of the objects is expected to exist in the perspective image, based on a predetermined motion model; and calculates the likelihood of the target pixel in each of the plurality of third regions, and
the detecting sets a prior probability to the third region where the image is expected to exist for each of the images, a value of the prior probability becoming larger the closer to a position to which the image moves based on the predetermined motion model, a kurtosis of the prior probability distribution being higher when a distance among the plurality of images is not more than a threshold; and detects, for each of the third regions, a position of the object based on the likelihood and the prior probability of the target pixel.

10. The device according to claim 1 , further comprising:
a storage to which each pixel constituting the perspective image is written in a predetermined order; and wherein
the processor-executable instructions further cause the processor to execute managing a timing to write, to the storage, all pixels necessary for calculation of the likelihood for a target pixel in which calculation of the likelihood is to be first performed, and
the calculating starts calculation of the likelihood for the target pixel after passage of the timing.

11. The device according to claim 1 , wherein the perspective image is an image of a color component corresponding to a transmittance of the object.

12. The device according to claim 11 , wherein when the object is an affected area of the subject, the perspective image is an image of a blue or green component.

13. The device according to claim 11 , wherein when the object is a marker indwelled around an affected area of the subject, the perspective image is an image of a red component.

14. The device according to claim 1 , wherein
the calculating calculates the likelihood for the target pixel in each of a plurality of perspective images perspectively seen an inside of the subject in different directions, and
the detecting acquires a projection matrix that is used to project three-dimensional coordinates to the perspective image for each perspective image; detects a target pixel having a likelihood whose value size satisfies a predetermined condition for each perspective image; and detects a three-dimensional position of the object using the projection matrix for each perspective image and the target pixel detected from each perspective image.

15. The device according to claim 14 , wherein
the processor detects a plurality of objects in the subject, and
the detecting acquires region information indicating a third region where each of images of the plurality of objects is expected to exist for each perspective image; detects a plurality of target pixels each having the likelihood whose value satisfies the predetermined condition for each perspective image; detects in which third region each of the plurality of target pixels detected from the perspective image exists, using the region information of the perspective image for each of the perspective images; defines correspondence of the plurality of target pixels detected in each of the plurality of perspective images among the plurality of perspective images; and detects three-dimensional positions of the plurality of objects using the projection matrix for each of the plurality of perspective images and the plurality of target pixels detected from each of the plurality of perspective images and corresponded among the plurality of perspective images.

16. The device according to claim 14 , wherein
the processor detects a plurality of objects in the subject, and
the detecting detects target pixels the likelihood for each of which satisfies the predetermined condition for each perspective image; defines correspondence of the plurality of target pixels detected in each perspective image among the plurality of perspective images by epipolar geometry using the projection matrix for each perspective image; and detects three-dimensional positions of the plurality of objects using the projection matrix for each perspective image and the target pixels detected from each perspective image and corresponded among the perspective images.

17. A treatment system, comprising:
a first irradiator that irradiates a subject with radiation;
a first detector that detects the radiation having passed through the subje

Assignees

• Toshiba Kk

Inventors

• Taguchi Yasunori
• Mita Takeshi
• Hirai Ryusuke
• Sugiura Kyoka
• Takeguchi Tomoyuki
• Sakata Yukinobu

Classifications

• G06T2207/10081

  Computed x-ray tomography [CT] · CPC title

• G06T7/74

  involving reference images or patches · CPC title

• G06T2207/30096

  Tumor; Lesion · CPC title

• A61N5/103

  Treatment planning systems · CPC title

• G06T7/337 Primary

  involving reference images or patches · CPC title