Systems and methods for detecting retinal abnormalities

• US 8,737,703 B2
• Filed: 10/15/2010
• Issued: 05/27/2014
• Est. Priority Date: 01/16/2008
• Status: Active Grant

• Current Assignee: Charles Stark Draper Laboratory Incorporated, Joslin Diabetes Center Inc.
• Original Assignee: Charles Stark Draper Laboratory Incorporated, Joslin Diabetes Center Inc.
• Inventors: Mangoubi, Rami, Desai, Mukund, Danko, Joseph, Aiello, Lloyd Paul
• Primary Examiner(s): Chen, Wenpeng
• Application Number:
  US12/905,929
• Publication Number:
  US 20110170751A1
• Time in Patent Office: 1,320 Days
• Field of Search: None
• US Class Current: 382/128
• CPC Class Codes: G06T 2207/30041 Eye; Retina; Ophthalmic
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06V 40/197 Matching; Classification

Abstract

The systems and methods described herein relate generally to the detection of retinal abnormalities. More particularly, the systems and methods are used to analyze a fundus or other image of the retina. In some implementations, the image is analyzed by interdependently smoothing and segmenting the image. Using the smoothed and segmented image, retinal features can be identifying in the image. The systems and methods can then analyze the detected retinal features to identify retinal abnormalities.

Claims

1. A computerized method for detecting retinal abnormalities, comprising:

receiving, by a computer, an image of a retina;

smoothing and segmenting the image, by the computer, wherein the smoothing and segmenting are interdependent and smoothing and segmenting further comprise;

reducing the value of a first and a second energy function, wherein the first energy function uses a first error metric and the second energy function uses a second error metric different from the first error metric; and

combining information about a first and a second set of smoothing and segmenting parameters obtained by reducing the value of the first and the second energy functions;

identifying, in the smoothed and segmented image, by the computer, at least one retinal feature; and

analyzing, by the computer, the identified at least one retinal feature in the smoothed and segmented image to identify a retinal abnormality.

2. The method of claim 1, wherein the at least one retinal feature comprises a blood vessel.

3. The method of claim 2, wherein analyzing the identified at least one blood vessel comprises:

determining at least one path associated with the at least one blood vessel in the image;

determining a variation in at least one of;

an angle of orientation of the path at a location along the path, and a diameter of the at least one blood vessel at a location along the path;

when the variation exceeds a threshold, identifying a retinal abnormality at the location.

4. The method of claim 3, wherein the segmenting and smoothing comprises determining an edge field strength at a plurality of locations in the image, with the at least one path is determined based at least in part on the edge field strength.

5. The method of claim 3, wherein determining at least one path comprises determining at least one center line associated with the at least one blood vessel in the image.

6. The method of claim 3, wherein determining a variation in an angle of orientation comprises determining a change in the angle of orientation between the sequential plurality of locations.

7. The method of claim 3, further comprising:

filtering, by the computer, a plurality of variations.

8. The method of claim 7, wherein filtering the plurality of variations comprises applying a moving window average.

9. The method of claim 7, wherein the variation exceeds a threshold when at least one of the filtered plurality of variations exceeds the threshold.

10. The method of claim 3, further comprising:

when the variation at a location exceeds a threshold, determining, by the computer, whether a vessel bifurcation is present at the location; and

when a vessel bifurcation is present at the location, filtering out the variation by the computer.

11. The method of claim 1, wherein the abnormality is a venous caliber abnormality.

12. The method of claim 1, wherein the abnormality is a new vascularity.

13. The method of claim 1, wherein smoothing and segmenting comprises:

adaptively adjusting, by the computer, at least one of a shape and orientation defining a neighborhood associated with a plurality of locations in the image.

14. The method of claim 13, wherein adaptively adjusting at least one of a shape and orientation of a neighborhood comprises reducing the value of an energy function associated with an error metric.

15. The method of claim 1, wherein analyzing the identified at least one retinal feature comprises analyzing a texture of the image.

16. A system for detecting retinal abnormalities, comprising:

a processor configured to;

receive an image of a retina;

smooth and segment the image, wherein the smoothing and segmenting are interdependent and smoothing and segmenting further comprise;

reducing the value of a first and a second energy function, wherein the first energy function uses a first error metric and the second energy function uses a second error metric different from the first error metric; and

combining information about a first and a second set of smoothing and segmenting parameters obtained by reducing the value of the first and the second energy functions;

identify, in the smoothed and segmented image, at least one retinal feature;

analyze the identified at least one retinal feature in the smoothed and segmented image to identify a retinal abnormality.

17. The system of claim 16, wherein the at least one retinal feature comprises a blood vessel.

18. The system of claim 17, wherein analyzing the identified at least one blood vessel comprises:

determining at least one path associated with the at least one blood vessel in the image;

determining a variation in at least one of;

an angle of orientation of the path at a location along the path, and a diameter of the at least one blood vessel at a location along the path;

when the variation exceeds a threshold, identifying a retinal abnormality at the location.

19. The system of claim 18, wherein the segmenting and smoothing comprises determining an edge field strength at a plurality of locations in the image, with the at least one path associated with the at least one blood vessel in the image is determined based at least in part on the edge field strength.

20. The system of claim 18, wherein determining at least one path comprises determining at least one center line associated with the at least one blood vessel in the image.

21. The system of claim 18, wherein determining a variation in an angle of orientation comprises determining a change in the angle of orientation between the sequential plurality of locations.

22. The system of claim 18, the processor further configured to:

filter a plurality of variations.

23. The system of claim 22, wherein filtering the plurality of variations comprises applying a moving window average.

24. The system of claim 22, wherein the variation exceeds a threshold when at least one of the filtered plurality of variations exceeds the threshold.

25. The system of claim 18, wherein the processor is further configured to:

when the variation at a location exceeds a threshold, determine whether a vessel bifurcation is present at the location; and

when a vessel bifurcation is present at the location, filter out the variation by the computer.

26. The system of claim 16, wherein the abnormality is a venous caliber abnormality.

27. The system of claim 16, wherein the abnormality is a new vascularity.

28. The system of claim 16, wherein smoothing and segmenting comprises:

adaptively adjusting, by the computer, at least one of a shape and orientation defining a neighborhood associated with a plurality of locations in the image.

29. The system of claim 28, wherein adaptively adjusting at least one of a shape and orientation of a neighborhood comprises reducing the value of an energy function associated with an error metric.

30. The system of claim 16, wherein analyzing the identified at least one blood vessel comprises analyzing a texture of the image.

31. Non-transitory computer readable media storing computer executable instructions, which when executed by such a computer, cause the computer to carry out a method comprising:

receiving an image of a retina;

smoothing and segmenting the image wherein the smoothing and segmenting are interdependent and smoothing and segmenting further comprise;

reducing the value of a first and a second energy function, wherein the first energy function uses a first error metric and the second energy function uses a second error metric different from the first error metric; and

combining information about a first and a second set of smoothing and segmenting parameters obtained by reducing the value of the first and the second energy functions;

identifying, in the smoothed and segmented image, at least one retinal feature;

analyzing the identified at least one retinal feature in the smoothed and segmented image to identify a retinal abnormality.

32. The non-transitory computer readable media of claim 31, wherein the at least one retinal feature comprises a blood vessel.