Methods and apparatus for computer-aided radiological detection and imaging

Patent metadata

Field	Value
Publication number	US-9471973-B2
Application number	US-201313953543-A
Country	US
Kind code	B2
Filing date	Jul 29, 2013
Priority date	Feb 4, 2011
Publication date	Oct 18, 2016
Grant date	Oct 18, 2016

Abstract

A computer aided detection (CAD) method and system is configured to input chest radiographs and generate overlay layer for labeling and tracing tubes detected within the image. The input image is first preprocessed and then segmented according to anatomy. Seeds are generated from the segmented image and then used to grow the region. Tubes are selected from the grown region and data is overlayed on the image based on the grown seed path.

First claim

What is claimed is:

1. A computer automated detection method for detection of one or more implantable man-made devices (IMD's) within the body of a patient, comprising:
receiving an image of the patient;
segmenting the image into one or more regions of interest, the regions of interest corresponding to anatomical features of the patient;
generating one or more seed points from a designated region of interest from the one or more regions of interest in the segmented image;
projecting one or more grown seed points along a path based on the one or more generated seeds; and
overlaying an IMD position based on the grown seed path over the patient image;
wherein the overlaid IMD position comprises an IMD path that follows the grown seed path along the length of the IMD within the image;
wherein the one or more IMD comprises a catheter comprising one or more of a NG tube or ET tube;
wherein generating one or more seed points comprises:
detecting a pair of parallel lines;
calculating a midpoint between the pair of parallel lines; and
assigning a location of one of said one or more seed points at said midpoint;

wherein segmenting the image into one or more regions of interest comprises automatically detecting a region of the neck of the patient; and
wherein the one or more seeds are generated within the neck region of interest.

2. The method recited in claim 1 , wherein the image comprises a radiograph.

3. The method recited in claim 1 , wherein projecting one or more grown seed points comprises:
matching one or more templates of the catheter with the one or more seed points; and
growing the path of the one or more seed points as a function of a profile of the catheter template;
wherein the catheter template profile is dynamically applied at each seed point based on at least one prior seed point location in the grown path.

4. The method recited in claim 1 , wherein the template comprises orientation, translation and tube profiles as parameters used in growing the path of the one or more seed points; and
wherein the orientation of the template is updated adaptively based on at least one prior seed point location in the grown path.

5. The method recited in claim 4 , further comprising:
calculating a template match score;
the template match score corresponding to confidence that a new point lies within an expected path of the generated seeds;
wherein the growth of the path of the one or more seed points is stopped upon the template match score falling below a threshold value.

6. The method recited in claim 4 , further comprising:
selecting a path from a plurality of grown paths, each of said grown paths being from a distinct starting position;
said selected path corresponding to a predetermined catheter profile; and
overlaying a catheter profile at the position within the image based on the selected path.

7. The method recited in claim 1 :
wherein catheter comprises a CVC tube;
wherein the one or more regions of interest comprises a heart region of interest and a lung region of interest; and
wherein the one or more seeds are generated within the heart region of interest and grown toward the lung region of interest.

8. A computer automated detection apparatus for detection of one or more implantable man-made devices (IMD's) within the body of a patient, comprising:
a processor; and
programming executable on said processor for:
receiving an image of the patient;
segmenting the image into one or more regions of interest, the regions of interest corresponding to anatomical features of the patient;
generating one or more seed points from the a designated region of interest from the one or more regions of interest in the segmented image;
projecting one or more grown seed points along a path based on the one or more generated seeds; and
overlaying an IMD position based on the grown seed path over the patient image;
wherein the overlaid IMD position comprises an IMD path that follows the grown seed path along the length of the IMD within the image;
wherein the one or more IMD comprises a catheter comprising one or more of a NG tube or ET tube;
wherein generating one or more seed points comprises:
detecting a pair of parallel lines;
calculating a midpoint between the pair of parallel lines; and
assigning a location of one of said one or more seed points at said midpoint;

wherein segmenting the image into one or more regions of interest comprises automatically detecting a region of the neck of the patient; and
wherein the one or more seeds are generated within the neck region of interest.

9. The apparatus recited in claim 8 , wherein the image comprises a radiograph.

10. The apparatus recited in claim 8 , wherein projecting one or more grown seed points comprises:
matching one or more templates of the catheter with the one or more seed points; and
growing the path of the one or more seed points as a function of a profile of the catheter template;
wherein the catheter template profile is dynamically applied at each seed point based on at least one prior seed point location in the grown path.

11. The apparatus recited in claim 10 , wherein the template comprises orientation, translation and tube profiles as parameters used in growing the path of the one or more seed points; and
wherein the orientation of the template is updated adaptively based on at least one prior seed point location in the grown path.

12. The apparatus recited in claim 11 , further comprising:
calculating a template match score;
the template match score corresponding to confidence that a new point lies within an expected path of the generated seeds;
wherein the growth of the path of the one or more seed points is stopped upon the template match score falling below a threshold value.

13. The apparatus recited in claim 11 , further comprising:
selecting a path from a plurality of grown paths, each of said grown paths being from a distinct starting position;
said selected path corresponding to a predetermined catheter profile; and
overlaying a catheter profile at the position within the image based on the selected path.

14. The apparatus recited in claim 8 :
wherein catheter comprises a CVC tube;
wherein the one or more regions of interest comprises a heart region of interest and a lung region of interest; and
wherein the one or more seeds are generated within the heart region of interest and grown toward the lung region of interest.

15. A detection apparatus for automatically detecting one or more implantable man-made devices (IMD's) within the body of a patient, comprising:
a scanner;
the scanner configured for generating an image of the patient; and
a detection module coupled to the scanner, the module configured for:
receiving the image of the patient;
segmenting the image into one or more regions of interest, the regions of interest corresponding to anatomical features of the patient;
generating one or more seed points from the designated region of interest in the segmented image;
projecting one or more grown seed points along a path based on the one or more generated seeds; and
overlaying an IMD position based on the grown seed path over the patient image;
wherein the overlaid IMD position comprises an IMD path that follows the grown seed path along the length of the IMD within the image;
wherein the one or more IMD comprises a catheter;

wherein generating one or more seed points comprises:
detecting a pair of parallel lines;
calculating a midpoint between the pair of parallel lines; and
assigning a location of one of said one or more seed points at sa

Assignees

• Univ California

Inventors

• Enzmann Dieter R
• Brown Matthew S
• Goldin Jonathan
• Ramakrishna Bharath

Classifications

• G06T2207/20156

  Automatic seed setting · CPC title

• G06T2207/10132

  Ultrasound image · CPC title

• A61B6/463

  characterised by displaying multiple images or images and diagnostic data on one display · CPC title

• G06T2207/30172

  Centreline of tubular or elongated structure · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title