Medical image processing apparatus and medical image processing method

Patent metadata

Field	Value
Publication number	US-10152651-B2
Application number	US-201514926341-A
Country	US
Kind code	B2
Filing date	Oct 29, 2015
Priority date	Oct 31, 2014
Publication date	Dec 11, 2018
Grant date	Dec 11, 2018

Abstract

A medical image processing apparatus according to an embodiment includes acquisition circuitry, extraction circuitry, calculation circuitry, and determination circuitry. The acquisition circuitry configured to acquire a first photographic image and a second photographic image that contain tree structures of a subject. The extraction circuitry configured to extract branch points of each of the tree structures in the first photographic image and the second photographic image. The calculation circuitry configured to calculate the similarities between the branch points in the first photographic image and the branch points in the second photographic image based on the feature quantities of the branch points. The determination circuitry configured to determine the corresponding path between the branch points in the first photographic image and the branch points in the second photographic image based on the similarities between the branch points calculated by the calculation circuitry.

First claim

What is claimed is:

1. A medical image processing apparatus, comprising:
acquisition circuitry configured to acquire a first photographic image and a second photographic image that contain tree structures of a subject;
extraction circuitry configured to extract branch points of each of the tree structures in the first photographic image and the second photographic image;
calculation circuitry configured to calculate similarities between the branch points in the first photographic image and the branch points in the second photographic image based on feature quantities of the branch points; and
determination circuitry configured to determine a corresponding path between the branch points in the second photographic image and the branch points in the first photographic image based on the similarities between the branch points calculated by the calculation circuitry, wherein the acquisition circuitry is configured to include:
multi-level acquisition circuitry configured to divide a tree structure into a plurality of levels, taking a level where seed points locate as start points, acquire corresponding portions of the tree structure in the respective levels in turn, and combine the acquired portions into a complete tree structure, and
method determination circuitry configured to evaluate acquisition results on a former level and, based on a result of the evaluation, determine acquisition method to be performed by the multi-level acquisition circuitry for a current level, and

the method determination circuitry is configured to take the acquisition method for the level where the seed points locate as region growing and repeat a process of adaptively acquiring segmentation results of the current level by adaptively adjusting parameters of the current level based on feedback of the segmentation results on the former level until growth of a whole tree structure has been completed.

2. The medical image processing apparatus according to claim 1 , further comprising division circuitry configured to divide the tree structure into a plurality of areas,
wherein the extraction circuitry is configured to extract the branch points in the areas based on the areas divided by the division circuitry.

3. The medical image processing apparatus according to claim 2 , wherein
the division circuitry is configured to determine an area of concern, and
the extraction circuitry is configured to extract the branch points for only the area of concern.

4. The medical image processing apparatus according to claim 1 , wherein the determination circuitry is configured to take the branch points extracted by the extraction circuit as branch point candidates and rank various branch point candidates based on the similarities between the branch points.

5. The medical image processing apparatus according to claim 1 , wherein the calculation circuitry is configured to calculate the similarities between the branch points based on positions of the branch points in the tree structure after weighting the various branch points.

6. The medical image processing apparatus according to claim 1 , wherein the determination circuitry is configured to determine the corresponding branch points between the first photographic image and the second photographic image based on the similarities between the branch points calculated by the calculation circuitry, and take a path consisting of various determined branch points as the corresponding path between the tree structures.

7. The medical image processing apparatus according to claim 1 , wherein
when the determination circuitry is configured to determine a plurality of paths, the calculation circuitry is configured to calculate similarities between a reference path in the first photographic image and each of the paths in the second photographic image, based on a set of feature quantities of various branch points constituting the paths, and
the determination circuitry is configured to determine the corresponding path between the tree structures, based on similarities between the paths calculated by the calculation circuitry.

8. A medical image processing apparatus comprising:
multi-level acquisition circuitry configured to divide a tree structure into a plurality of levels, taking a level where a seed point locates as a start point, acquire corresponding portions of the tree structure in the respective levels in turn, and combine the acquired portions into a complete tree structure; and
method determination circuitry configured to evaluate acquisition results on a former level and, based on results of the evaluation, determine acquisition method to be performed by the multi-level acquisition circuitry for a current level, wherein
the method determination circuitry is configured to take the acquisition method for the level where the seed point locates as region growing, estimate a range of volume of a whole tree structure based on segmentation results of a first level and repeat a process of adaptively acquiring segmentation results of the current level by adaptively adjusting parameters of the current level based on feedback of segmentation results on the former level and the estimated range of volume of the whole tree structure until growth of the whole tree structure has been completed.

9. A medical image processing method, comprising:
by multi-level acquisition circuitry, dividing a tree structure into a plurality of levels, taking level where a seed point locates as a start point, acquiring corresponding portions of the tree structure in the respective levels in turn, and combining the acquired portions into a complete tree structure; and
by method determination circuitry, evaluating acquisition results on a former level and, based on results of the evaluation, determining acquisition method to be performed by the multi-level acquisition circuitry for a current level, wherein
the determining the acquisition comprises taking the acquisition method for the level where the seed point locates as region growing, estimating a range of volume of a whole tree structure based on segmentation results of a first level and repeating a process of adaptively acquiring the segmentation results of the current level by adaptively adjusting parameters of the current level based on feedback of the segmentation results on the former level and the estimated range of volume of the whole tree structure until growth of the whole tree structure has been completed.

10. A medical image processing method, comprising:
acquiring, by acquisition circuitry, a first photographic image and a second photographic image that contain tree structures of a subject;
extracting, by extraction circuitry, branch points of each of the tree structures in the first photographic image and the second photographic image;
calculating, by calculation circuitry, similarities between the branch points in the first photographic image and the branch points in the second photographic image based on feature quantities of the branch points;
determining, by determination circuitry, a corresponding path between the branch points in the second photographic image and the branch points in the first photographic image based on the similarities between the branch points calculated by the calculating, wherein
the acquiring, by the acquisition circuitry, includes:
dividing a tree structure into a plurality of levels, taking a level where seed points locate as start points, acquiring corresponding portions of the tree structure in the respective levels in turn, and combining the acquired portions into a complete tree structure, and
evaluating acquisition results on a former level and, based on a result of the evaluation, determining acquisition method to be performed by the a

Assignees

• Toshiba Medical Sys Corp

Inventors

• Wang Shaobin
• Chen Qi
• Zhang Yang
• Wang Yanhua
• Zhang Jialin
• Xu Minfeng
• Zhao Jianchun
• Yang Xiaotian

Classifications

• G06V10/757

  Matching configurations of points or features · CPC title

• G06T7/11 Primary

  Region-based segmentation · CPC title

• G06T2207/10081

  Computed x-ray tomography [CT] · CPC title

• G06T2207/30061

  Lung · CPC title

• G06T7/162

  involving graph-based methods · CPC title