Optimization of parameters for segmenting an image

Patent metadata

Field	Value
Publication number	US-10134142-B2
Application number	US-201515314527-A
Country	US
Kind code	B2
Filing date	Jun 11, 2015
Priority date	Jun 12, 2014
Publication date	Nov 20, 2018
Grant date	Nov 20, 2018

Abstract

The present invention relates to a device for segmenting an image ( 12 ) of a subject ( 14 ) comprising a data interface ( 16 ) for receiving an image ( 12 ) of said subject ( 14 ) and at least one contour ( 18 ) or at least one part of a contour ( 18 ), said contour ( 18 ) indicating a structure ( 19 ) within said image ( 12 ), a selection unit ( 20 ) for selecting a region ( 22 ) in said image ( 12 ) divided into a first and a second disjoint part ( 24, 26 ) by said contour ( 18 ) or said part of said contour ( 18 ), said selected region ( 22 ) comprising a drawn region and/or a computed region, a classifier ( 28 ) for classifying said selected region ( 22 ) based on at least one parameter for image segmentation, an analysis unit ( 29 ) for defining an objective function based on the classification result, an optimizer ( 30 ) for optimizing said parameter set by varying an output of said objective function and an image segmentation unit ( 32 ) for segmenting said image ( 12 ) using said optimized parameter set.

First claim

The invention claimed is:

1. Device for segmenting an image of a subject, comprising:
a data interface for receiving an image of said subject and at least one contour or at least one part of a contour, said contour indicating a structure within said image;
a selection unit for selecting a region in said image divided into a first and a second disjoint part by said contour or said part of said contour, said selected region comprising a region drawn by a user and/or a region generated by a computer;
a classifier for classifying said selected region into a first image class and a second image class based on at least one parameter for image segmentation;
an analysis unit for defining an objective function based on at least one of said first is a class and said second image class;
an optimizer for optimizing said at least one parameter by varying an output of said objective function; and
an image segmentation unit for segmenting said image using said optimized at least one parameter;
wherein said selection unit is configured to select said region to divide said contour into a first contour part within said region and a second contour part outside said region, and
wherein the image segmentation unit is configured to transfer said at least one parameter optimized for said region to at least one part selected in said second contour part.

2. The device according to claim 1 , wherein said first image class is a major image class of said first disjoint part.

3. The device according to claim 1 , wherein said second image class is a major image class of said second disjoint part.

4. The device according to claim 1 , wherein said optimizer being configured to determine a first area from said first disjoint part, said first area corresponding to said first image class.

5. The device according to claim 4 , wherein said objective function comprises a summation of said first and said second area.

6. The device according to claim 1 , wherein said optimizer being configured to determine a second area from said second disjoint part, said second area corresponding to said second image class.

7. The device according to claim 1 , wherein said optimizer is configured to maximize said output of said objective function.

8. The device according to claim 7 , wherein said optimizer is configured to maximize said output of said objective function according to a ground truth classification of said first and/or said second disjoint part.

9. The device according to claim 1 , wherein said image segmentation unit is configured to adjust said contour or said part of said contour using said optimized parameter set.

10. The device according to claim 1 , wherein said contour or said part of said contour comprises a drawn contour or a drawn part of contour and/or a computed contour.

11. System for segmenting an image of a subject, comprising:
an imaging apparatus for generating at least one image of said subject; and
a device claimed in claim 1 for segmenting said generated at least one image.

12. A method for segmenting an image of a subject, comprising the steps of:
receiving an image said subject and at least one contour at least one part of a contour, said contour indicating a structure within said image;
selecting a region in said image divided into a first and a second disjoint part by said contour or said part of said contour, said selected region comprising a region drawn by a user and/or a region generated by a computer;
classifying said selected region into a first image class and a second image class based on at least one parameter for image segmentation;
defining an objective function based on at least one of said first image class and said second image class;
optimizing said at least one parameter by varying an output of said objective function; and
segmenting said image using said optimized at least one parameter;

wherein said method further comprises the steps of selecting said region to divide said contour into a first contour part within said region and a second contour part outside of said region and transferring said at least one parameter optimized for said region to at least one part selected in said second contour part.

13. Computer program comprising program code means for causing a computer to carry out the steps of the method as claimed in claim 12 when said computer program is carried out on the computer.

Assignees

• Koninklijke Philips Nv

Inventors

• Bystrov Daniel
• Schulz Heinrich
• Vik Torbjoern

Classifications

• G06T7/11 Primary

  Region-based segmentation · CPC title

• G06T7/12 Primary

  Edge-based segmentation · CPC title

• G06T12/10

  Image preprocessing, e.g. calibration, positioning of sources or scatter correction · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06T2207/20112

  Image segmentation details · CPC title