Method and system for the automated delineation of lung regions and costophrenic angles in chest radiographs

• US 6,282,307 B1
• Filed: 02/23/1998
• Issued: 08/28/2001
• Est. Priority Date: 02/23/1998
• Status: Expired due to Term

• Current Assignee: ARCH Development Corporation (University of Chicago)
• Original Assignee: ARCH Development Corporation (University of Chicago)
• Inventors: Giger, Maryellen L., Armato, Samuel G. III, MacMahon, Heber
• Primary Examiner(s): Bella, Matthew C.
• Application Number:
  US09/028,518
• Time in Patent Office: 1,282 Days
• Field of Search: 382/128, 382/131, 382/132, 382/199, 382/172, 382/266
• US Class Current: 382/132
• CPC Class Codes: A61B 6/50 specially adapted for specific body parts; specially adapted for specific clinical applications
• CPC Class Codes: G06T 2207/10116 X-ray image
• CPC Class Codes: G06T 2207/20012 Locally adaptive
• CPC Class Codes: G06T 2207/30061 Lung
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/66 of image moments or centre of gravity
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns

Abstract

A method, system, and computer product for the automated segmentation of the lung fields and costophrenic angle (CP) regions in posteroanterior (PA) chest radiographs wherein image segmentation based on gray-level threshold analysis is performed by applying an iterative global gray-level thresholding method to a chest image based on the features of a global gray-level histogram. Features of the regions in a binary image constructed at each iteration are identified and analyzed to exclude regions external to the lung fields. The initial lung contours that result from this global process are used to facilitate a local gray-level thresholding method. Individual regions-of-interest (ROIs) are placed along the initial contour. A procedure is implemented to determine the gray-level thresholds to be applied to the pixels within the individual ROIs. The result is a binary image, from which final contours are constructed. Smoothing processes are applied, including a unique adaptation of a rolling ball method. CP angles are identified and delineated by using the lung segmentation contours as a means of placing ROIs that capture the CP angle regions. Contrast-based information is employed on a column-by-column basis to identify initial diaphragm points, and maximum gray-level information is used on a row-by-row basis to identify initial costal points. Analysis of initial diaphragm and costal points allows for appropriate adjustment of CP angle ROI positioning. Polynomial curve-fitting is used to combine the diaphragm and costal points into a continuous, smooth CP angle delineation. This delineation is then spliced into the final lung segmentation contours. In addition, quantitative information derived from the CP angle delineations is used to assess the presence of abnormal CP angles.

Claims

1. A method for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

performing, based on said first initial lung segmentation contours, local threshold analysis to construct second initial lung segmentation contours for said posteroanterior chest image; and

applying a rolling ball filter to said second initial lung segmentation contours to smooth the shape of said second initial lung segmentation contours.

2. The method according to claim 1, further comprising:

performing global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level designating a gray-level threshold range;

wherein said performing global threshold analysis further comprises;

a) generating a binary image having pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range;

b) repeating substep a) a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range; and

c) processing said binary image to eliminate pixels outside the lung fields.

3. The method according to claim 1, wherein said performing local threshold analysis comprises:

positioning regions-of-interest (ROIs) along said first initial lung segmentation contours;

performing local gray-level thresholding within each of said ROIs;

generating a composite binary image based on said performing local gray-level thresholding; and

constructing based on said composite binary image said second initial lung segmentation contours for said posteroanterior chest image.

4. The method according to claim 1, further comprising:

delineating costophrenic angle margins for each of said lung fields; and

constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins.

5. The method according to claim 4, wherein said constructing final lung segmentation contours comprises:

splicing said costophrenic angle margins to said second initial lung segmentation contours to construct said final lung segmentation contours.

6. The method according to claim 4, wherein said delineating costophrenic angle margins comprises:

determining whether said posteroanterior chest image exhibits an abnormal hemithorax.

7. A method for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said lung fields identified in said processed image global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image and;

performing global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level designating a gray-level threshold range;

wherein said performing global threshold analysis further comprises;

a) generating a binary image having pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range,

b) repeating substep a) a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range, and

c) processing said binary image to eliminate pixels outside the lung fields;

wherein substep c) comprises;

constructing intermediate lung segmentation contours representing boundaries of groups of contiguous pixels having said first logic level,

determining the centroid for each of said intermediate lung segmentation contours;

detecting centroids outside the lung fields based on a gray-level profile, and

prohibiting regions of the binary image defined by intermediate lung segmentation contours having centroids outside the lung fields from having said first logic level during subsequent iterations of substep a) generating.

8. The method according to claim 7, wherein said performing global threshold analysis further comprises:

smoothing the intermediate lung segmentation contours constructed during the last iteration of said generating a binary image to provide said first initial lung segmentation contours.

9. A method for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said lung fields identified in said processed image global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

delineating costophrenic angle margins for each of said lung fields; and

constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins;

wherein said delineating costophrenic angle margins comprises;

a) placing costophrenic regions of interest (ROIs) over approximate locations of each costophrenic angle as determined based on the first initial lung segmentation contours;

b) identifying diaphragm border points within each of said ROIs;

c) identifying costal border points within each of said ROIs;

d) checking the placement of said ROIs at least once to determine whether the ROIs were positioned accurately in substep a); and

e) repeating substeps a) through d) if it is determined in substep d) that the ROIs were not accurately positioned in substep a).

10. A computer readable medium storing computer instructions for identification of lung fields within a chest region based on posteroanterior chest radiographic images, by performing the steps of:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

performing, based on said first initial lung segmentation contours, local threshold analysis to construct second initial lung segmentation contours for said posteroanterior chest image; and

applying a rolling ball filter to said second initial lung segmentation contours to smooth the shape of said second initial lung segmentation contours.

11. The medium of claim 10, wherein the computer instructions further comprise:

performing global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level designating a gray-level threshold range;

and wherein the computer instructions for said performing global threshold analysis further comprise;

a) generating a binary image having pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range;

b) repeating substep a) a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range; and

c) processing said binary image to eliminate pixels outside the lung fields.

12. The medium of claim 10, wherein the computer instructions for said performing local threshold analysis comprise:

positioning regions-of-interest (ROIs) along said first initial lung segmentation contours;

performing local gray-level thresholding within each of said ROIs;

generating a composite binary image based on said performing local gray-level thresholding; and

constructing based on said composite binary image said second initial lung segmentation contours for said posteroanterior chest image.

13. The medium of claim 10, wherein the computer instructions further comprise:

delineating costophrenic angle margins for each of said lung fields; and

constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins.

14. The medium of claim 13, wherein the computer instructions for said constructing final lung segmentation contours comprise:

splicing said costophrenic angle margins to said second initial lung segmentation contours to construct said final lung segmentation contours.

15. The medium of claim 13, wherein the computer instructions for said delineating costophrenic angle margins comprises:

determining whether said posteroanterior chest image exhibits an abnormal hemithorax.

16. A computer readable medium storing computer instructions for identification of lung fields within a chest region based on posteroanterior chest radiographic images, by performing the steps of:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image; and

performing global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level designating a gray-level threshold range;

wherein the computer instructions for said performing global threshold analysis further comprise;

a) generating a binary image having pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range,

b) repeating substep a) a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range, and

c) processing said binary image to eliminate pixels outside the lung fields;

wherein the computer instructions for said substep c) comprise;

constructing intermediate lung segmentation contours representing boundaries of groups of contiguous pixels having said first logic level,

determining the centroid for each of said intermediate lung segmentation contours;

detecting centroids outside the lung fields based on a gray-level profile, and

prohibiting regions of the binary image defined by intermediate lung segmentation contours having centroids outside the lung fields from having said first logic level during subsequent iterations of said generating a binary image.

17. The medium of claim 16, wherein the computer instructions for said performing global threshold analysis further comprise:

smoothing the intermediate lung segmentation contours constructed during the last iteration of said generating a binary image to provide said first initial lung segmentation contours.

18. A computer-readable medium storing computer instructions for identification of lung fields within a chest region based on posteroanterior chest radiographic images, by performing the steps of:

generating first image data representative of a posteroanterior chest image inclusive of lung fields;

performing global threshold analysis of the posteroanterior chest image;

constructing, based on said lung fields identified in said processed image global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

delineating costophrenic angle margins for each of said lung fields; and

constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins;

wherein the computer instructions for said delineating costophrenic angle margins comprise;

a) placing costophrenic regions of interest (ROIs) over approximate locations of each costophrenic angle as determined based on the first initial lung segmentation contours;

b) identifying diaphragm border points within each of said ROIs;

c) identifying costal border points within each of said ROIs;

d) checking the placement of said ROIs at least once to determine whether the ROIs were positioned accurately in substep a); and

e) repeating substeps a) through d) if it is determined in substep d) that the ROIs were not accurately positioned in substep a).

19. A system for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

an image acquisition device configured to generate first image data representative of a posteroanterior chest image inclusive of lung fields;

a global gray-level thresholding unit configured to perform global threshold analysis of the posteroanterior image;

means for constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

a local gray-level thresholding unit configured to perform, based on said first initial lung segmentation contours, local threshold analysis to construct second initial lung segmentation contours for said posteroanterior chest image; and

a rolling ball filter unit configured to apply a rolling ball filter to said second initial lung segmentation contours to smooth the shape of said second initial lung segmentation contours.

20. The system of claim 19, further comprising:

a gray-level histogram analysis unit configured to perform a global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level which designate a gray-level threshold range;

wherein said global gray-level thresholding unit comprises;

a) means for generating a binary image with pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range;

b) means for iteratively using said a) means a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range; and

c) means for processing said binary image to eliminate pixels outside the lung fields.

21. The system of claim 19, wherein said local gray-level thresholding unit further comprises:

means for positioning regions-of-interest (ROIs) along said first initial lung segmentation contours;

means for performing local gray-level thresholding within each of said ROIs;

means for generating a composite binary image based on an output of said means for performing local gray-level thresholding; and

means for constructing based on said composite binary image said second initial lung segmentation contours for said posteroanterior chest image.

22. The system of claim 19, further comprising:

means for delineating costophrenic angle margins for each of said lung fields; and

means for constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins.

23. The system of claim 22, wherein said means for constructing final lung segmentation contours comprises:

a splicing unit configured to splice said costophrenic angle margins to said second initial lung segmentation contours to construct said final lung segmentation contours.

24. The system of claim 22, wherein said means for delineating costophrenic angle margins comprises:

means for determining whether said posteroanterior chest image exhibits an abnormal hemithorax.

25. A system for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

an image acquisition device configured to generate first image data representative of a posteroanterior chest image inclusive of lung fields;

a global gray-level thresholding unit configured to perform global threshold analysis of the posteroanterior image;

means for constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

a gray-level histogram analysis unit configured to perform a global gray-level histogram analysis of said posteroanterior chest radiographic image to identify a maximum gray-level and a minimum gray-level which designate a gray-level threshold range;

wherein said global gray-level thresholding unit comprises;

a) means for generating a binary image with pixels having either a first logic level or a second logic level based on a gray-level threshold within said gray-level threshold range,

b) means for iteratively using said a) means a predetermined number of times at progressively larger gray-level thresholds within the gray-level threshold range, and

c) means for processing said binary image to eliminate pixels outside the lung fields;

wherein said c) means comprises;

means for constructing intermediate lung segmentation contours representing boundaries of groups of contiguous pixels having said first logic level,

means for determining the centroid for each of said intermediate lung segmentation contours,

means for detecting centroids outside the lung fields based on a gray-level profile, and

means for prohibiting regions of the binary image defined by intermediate lung segmentation contours with centroids outside the lung fields from having said first logic level during subsequent iterations of using said a) means.

26. The system of claim 25 wherein said global gray-level thresholding analysis unit further comprises:

means for smoothing the intermediate lung segmentation contours constructed during the last iteration of using said a) means to provide said first initial lung segmentation contours.

27. A system for identifying lung fields within a chest region based on posteroanterior chest radiographic images, comprising:

an image acquisition device configured to generate first image data representative of a posteroanterior chest image inclusive of lung fields;

a global gray-level thresholding unit configured to perform global threshold analysis of the posteroanterior image;

means for constructing, based on said global threshold analysis, first initial lung segmentation contours for said posteroanterior chest image;

means for delineating costophrenic angle margins for each of said lung fields; and

means for constructing final lung segmentation contours based on said first initial lung segmentation contours and said costophrenic angle margins;

wherein said means for delineating costophrenic angle margins comprises;

a) means for placing costophrenic regions of interest (ROIs) over approximate locations of each costophrenic angle as determined based on the first initial lung segmentation contours;

b) means for identifying diaphragm border points within each of said ROIs;

c) means for identifying costal border points within each of said ROIs;

d) means for checking the placement of said ROIs at least once to determine whether the ROIs were positioned accurately by said a) means; and

e) means for iteratively using means a) through d) if it is determined by d) means that the ROIs were not accurately positioned by said a) means.