Systems and Methods for Machine Learning Based Hanging Protocols

• US 20100080427A1
• Filed: 09/29/2008
• Published: 04/01/2010
• Est. Priority Date: 09/29/2008
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Yeluri, Vijaykalyan, Lau, Denny
• Granted Patent: US 8,165,368 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• CPC Class Codes: G16H 30/20 for handling medical images, e.g. DICOM, HL7 or PACS
• CPC Class Codes: G16H 30/40 for processing medical images, e.g. editing

Abstract

Certain embodiments of the present invention provide methods and systems for determining a hanging protocol for display of clinical images in a study. Certain embodiments provide a method for determining a hanging protocol for clinical image display. The method includes capturing one or more high level characteristics for an image study based on relationships between images in the study. The method also includes identifying one or more low level characteristics for individual images in the image study. The method further includes classifying the image study based on a combination of the high level characteristics and the low level characteristics. The method additionally includes determining a hanging protocol for the image study based on the classifying. The method includes providing the hanging protocol for display of images in the image study.

Claims

1. A method for determining a hanging protocol for clinical image display, said method comprising:

capturing one or more high level characteristics for an image study based on relationships between images in the study;

identifying one or more low level characteristics for individual images in the image study;

classifying the image study based on a combination of the high level characteristics and the low level characteristics;

determining a hanging protocol for the image study based on the classifying; and

providing the hanging protocol for display of images in the image study.

2. The method of claim 1, wherein the one or more high level characteristics comprise at least one of detection method, number of relevant images, and a comparison of image resolution versus monitor resolution.

3. The method of claim 1, wherein the one or more low level characteristics comprise at least one of header data elements and order information.

4. The method of claim 1, further comprising adapting the hanging protocol based on a stored user layout of images.

5. The method of claim 1, wherein determining a hanging protocol further comprises determining a hanging protocol for the image study by applying a machine learning algorithm to the combination of the high level characteristics and the low level characteristics.

6. The method of claim 1, further comprising allowing a user to modify a layout automatically generated based on the hanging protocol.

7. The method of claim 1, wherein the determining a hanging protocol further comprises determining a methodology to detect abnormalities in the image study, the method further comprising:

displaying images with different views from the image study;

displaying relevant images to support the methodology from one or more additional image studies; and

limiting a number of images displayed.

8. A system for displaying clinical images for review, said system comprising:

an input receiving one or more images in an image study for display;

a machine learning engine selecting a hanging protocol for display of the one or more images, said machine learning engine capturing one or more high level characteristics for the image study based on relationships between images in the study, identifying one or more low level characteristics for individual images in the image study, classifying the image study based on a combination of the high level characteristics and the low level characteristics, and determining a hanging protocol for the image study based on the classifying; and

an output providing the hanging protocol for use in displaying the one or more images.

9. The system of claim 8, wherein the one or more high level characteristics comprise at least one of detection method, number of relevant images, and a comparison of image resolution versus monitor resolution.

10. The system of claim 8, wherein the one or more low level characteristics comprise at least one of header data elements and order information.

11. The system of claim 8, wherein the machine learning engine adapts the hanging protocol based on a stored user layout of images.

12. The system of claim 8, wherein said machine learning engine determines a hanging protocol for the image study by applying a machine learning algorithm to the combination of the high level characteristics and the low level characteristics.

13. The system of claim 8, wherein the machine learning engine accepts user input to modify a layout automatically generated based on the hanging protocol.

14. The system of claim 8, wherein the machine learning engine determines a hanging protocol by determining a methodology to detect abnormalities in the image study, the methodology being determined by:

displaying images with different views from the image study;

displaying relevant images to support the methodology from one or more additional image studies; and

limiting a number of images displayed.

15. The system of claim 8, wherein the machine learning engine utilizes a probabilistic model to apply one or more high level characteristics for the image study based on image study information and relationships between images in the study with one or more low level characteristics for individual images in the image study.

16. A machine-readable storage medium including a set of instructions for execution on a processor, the set of instructions comprising:

an input routine receiving one or more images in an image study for display;

a machine learning routine selecting a hanging protocol for display of the one or more images, said machine learning routine capturing one or more high level characteristics for the image study based on relationships between images in the study, identifying one or more low level characteristics for individual images in the image study, classifying the image study based on a combination of the high level characteristics and the low level characteristics, and determining a hanging protocol for the image study based on the classifying; and

an output routine providing the hanging protocol for use in displaying the one or more images.

17. The machine-readable storage medium of claim 16, wherein the one or more high level characteristics comprise at least one of detection method, number of relevant images, and a comparison of image resolution versus monitor resolution.

18. The machine-readable storage medium of claim 16, wherein the one or more low level characteristics comprise at least one of header data elements and order information.

19. The machine-readable storage medium of claim 16, wherein the machine learning routine accepts user input to modify a layout automatically generated based on the hanging protocol.

20. The machine-readable storage medium of claim 16, wherein the machine learning routine determines a hanging protocol by determining a methodology to detect abnormalities in the image study, the methodology being determined by:

displaying images with different views from the image study;

displaying relevant images to support the methodology from one or more additional image studies; and

limiting a number of images displayed.