Method of lung lobe segmentation and computer system

• US 7,315,639 B2
• Filed: 03/03/2004
• Issued: 01/01/2008
• Est. Priority Date: 03/03/2004
• Status: Active Grant

• Current Assignee: Mevis GmbH
• Original Assignee: Mevis GmbH
• Inventors: Kuhnigk, Jan-Martin
• Primary Examiner(s): Werner; Brian P.
• Assistant Examiner(s): RADKIEWICZ, JARED
• Application Number:
  US10/793,669
• Publication Number:
  US 20050196024A1
• Time in Patent Office: 1,399 Days
• Field of Search: None
• US Class Current: 382/131
• CPC Class Codes: G06V 10/26 Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
• CPC Class Codes: G06V 2201/031 of internal organs

Abstract

A method of lung lobe segmentation having the steps of: providing of three dimensional first image data of a lung; segmentation of a tubular structure to identify first voxels of the first image data belonging to the tubular structure and to identify second voxels of the first image data which do not belong to the tubular structure; determining a distance measure for each one of the second voxels, by determining a minimum distance between each one of the second voxels to one of the first voxels to provide second image data; and performing a lobe segmentation based on the second image data.

Claims

1. A method of lung lobe segmentation comprising the steps of:

providing of three dimensional first image data of a lung,

performing segmentation of the first image data with respect to a tubular structure located within the lung to identify first voxels of the first image data belonging to the tubular structure and to identify second voxels of the first image data which do not belong to the tubular structure,

performing an image transformation on the first image data being expressive of a degree of association between the identified first and the second voxels to provide second image data,

performing a weighted combination of the first image data and the second image data to provide third image data,

performing a lobe segmentation based on the third image data, and

obtaining an output of the lobe segmentation indicative of the lobes of the lung.

2. The method of claim 1, the image transformation comprising determining a distance measure for each one of the second voxels to one of the first voxels.

3. The method of claim 2, the distance measure being determined by determining a minimum distance between each one of the second voxels to one of the first voxels to provide the second image data.

4. The method of claim 2, whereby the world coordinates of the first and second voxels are used for determining the distance.

5. The method of claim 1, the image transformation being performed by means of a cost function between the first and second voxels.

6. The method of claim 5, the cost function being the sum of all density values of the voxels on a path between a pair of first and second voxels and further comprising determining a minimum cost path for each one of the second voxels to one of the first voxels to provide the second image data.

7. The method of claim 1, the step of providing first image data comprising acquisition of thoracic image data and lung segmentation of the thoracic image data to provide the first image data.

8. The method of claim 1, whereby the tubular structure is a vascular structure or a bronchial structure.

9. The method of claim 1, whereby the first and the second voxels are identified by means of a first segmentation mask.

10. The method of claim 1, whereby the output is displayed, printed or stored.

11. The method of claim 1, whereby the lobe segmentation is performed by means of a watershed transformation.

12. The method of claim 11, whereby the first voxels are inverted prior to the watershed transformation.

13. A computer readable medium containing program instructions for performing lung lobe segmentation;

performing segmentation of a three dimensional first image data of a lung with respect to a tubular structure located within the lung to identify first voxels of the first image data belonging to the tubular structure and to identify second voxels of the first image data which do not belong to the tubular structure;

performing an image transformation on the first image data being expressive of a degree of association between the identified first and the second voxels to provide second image data;

performing a weighted combination of the first image data and the second image data to provide third image data, performing a lobe segmentation based on the third image data, and obtaining an output of the lobe segmentation indicative of the lobes of the lung.

14. A computer system comprising:

a memory component for storing of three dimensional first image data of a lung,

a first program component for performing segmentation of the first image data with respect to a tubular structure located within the lung to identify first voxels of the first image data belonging to the tubular structure and to identify second voxels of the first image data which do not belong to the tubular structure,

a second program component for determining a distance measure for each one of the second voxels, by determining a minimum distance between each one of the second voxels to one of the first voxels to provide second image data

a third program component for performing a weighted combination of the second image data and the first image data to provide third image data,

a fourth program component for performing a lobe segmentation based on the third image data, and

a fifth program component for obtaining an output of the lobe segmentation indicative of the lobes of the lung.

15. The computer system of claim 14 further comprising means for acquisition of thoracic image data and a sixth program component for lung segmentation of the thoracic image data to provide the first image data.

16. The computer system of claim 14, the first program component being adapted to provide a first segmentation mask for identification of the first and second voxels.

17. The computer system of claim 14, whereby the second program component is adapted to calculate an Euclidean distance between a pair of first and second voxels.

18. The computer system of claim 14, further comprising seventh program means for providing world coordinates of the first and second voxels for determination of the distance.

19. The computer system of claim 14, the fourth program component being adapted to perform a watershed transformation.

20. The computer system of claim 19, further comprising an eighth program component for inversion of the first voxels prior to the watershed transformation.