System and methods for automatic processing of digital retinal images in conjunction with an imaging device

• US 9,462,945 B1
• Filed: 04/22/2014
• Issued: 10/11/2016
• Est. Priority Date: 04/22/2013
• Status: Active Grant

• Current Assignee: VisionQuest Biomedical, LLC
• Original Assignee: VisionQuest Biomedical, LLC
• Inventors: Barriga, Simon, Agurto, Carla, Yu, Honggang, Zamora, Gilberto, Joshi, Vinayak, Soliz, Peter
• Primary Examiner(s): Dinh, Jack
• Application Number:
  US14/259,014
• Time in Patent Office: 903 Days
• Field of Search: 351/208, 351/206, 351/246
• US Class Current: 1/1
• CPC Class Codes: A61B 3/0025 characterised by electronic signal processing, e.g. eye models
• CPC Class Codes: A61B 3/0033 characterised by user input arrangements
• CPC Class Codes: A61B 3/12 for looking at the eye fundus, e.g. ophthalmoscopes A61B3/13 takes precedence
• CPC Class Codes: A61B 3/14 Arrangements specially adapted for eye photography apparatus or arrangements for taking photographs per se G03B
• CPC Class Codes: A61B 3/152 for aligning

Abstract

Systems and methods of obtaining and recording fundus images by minimally trained persons, which includes a camera for obtaining images of a fundus of a subject'"'"'s eye, in combination with mathematical methods to assign real time image quality classification to the images obtained based upon a set of criteria. The classified images will be further processed if the classified images are of sufficient image quality for clinical interpretation by machine-coded and/or human-based methods. Such systems and methods can thus automatically determine whether the quality of a retinal image is sufficient for computer-based eye disease screening. The system integrates global histogram features, textural features, and vessel density, as well as a local non-reference perceptual sharpness metric. A partial least square (PLS) classifier is trained to distinguish low quality images from normal quality images.

Claims

1. A method to perform automatic retinal screening, the method comprising the steps of:

illuminating a retina using an illumination source;

capturing a digital retinal image with a retinal camera supported by a platform;

transmitting the digital retinal image to a processor;

performing via the processor an assessment in real time of the image quality wherein image quality is determined by one or more of the image quality steps of;

determining a position of the Optical Disc or Macula relative to the digital retinal image obtained;

determining percent of total pixels obscured by shadowing; and

determining the value of the third quartile of a histogram of a gradient of pixel values from the digital retinal image;

adjusting the digital retinal camera if the image quality does not meet predetermined image quality requirements; and

executing real time analysis of presence or absence of disease in the digital retinal image

wherein the digital retinal image is inadequate to meet the predetermined image quality requirement when one or more of the following is present;

i) the number of pixels in the third quartile of a gradient image of the digital retinal image occupies more than 15 percent of the image area, ii) the total number of pixels within 10 percent of saturation is larger than 25 percent of the total number of pixels in the image after inverting values of a thresholded image, iii) a macula position is at an angle greater than between 0 degree and 20 degrees below a position of an optic disc with respect to a horizontal axis of the digital retinal image, iv) a distance between an optic disc center and a macula is more than two optic disc diameters and less than three optic disc diameters, v) an optic disc and a macula are positioned within one-third and two-thirds sections of a vertical digital image, and vi) a macula is positioned within a vertical region limited by temporal vessel arcades.

2. The method of claim 1 wherein the performing step determines alignment of the image according to one or more imaging protocols.

3. The method of claim 2 additionally comprising assigning a descriptive label to the image as to its adequacy according to one or more imaging protocols.

4. The method of claim 1 wherein the performing step determines presence and extent of crescents and shadows in the image.

5. The method of claim 4 additionally comprising assigning a descriptive label to the image as to presence and extent of crescents and shadows.

6. The method of claim 1 wherein the performing step determines quantitative image quality of the image via a classification process trained using examples of visual perception by human experts.

7. The method of claim 6 additionally comprising assigning a descriptive label to the image as to its quantitative image quality.

8. The method of claim 1 wherein the performing step classifies the image according to a set of image quality labels.

9. The method of claim 8 additionally comprising assigning a descriptive label to the image as to its image quality.

10. The method of claim 1 wherein the adjusting step employs a user interface to indicate a user quality of the image and suggested actions to take with respect to the camera.

11. The method of claim 1 additionally comprising assigning a label to the image indicative of the likelihood.

12. The method of claim 1 wherein the executing step employs a decomposition phase using AM-FM methods to decompose the image into magnitude, frequency, and angle of each image pixel across a plurality of spatial scales.

13. The method of claim 12 wherein the executing step additionally employs a feature extraction phase to group decomposition phase components into feature vectors.

14. The method of claim 13 wherein the executing step additionally employs a feature reduction phase to reduce number of features.

15. The method of claim 1 wherein the executing step additionally comprises tuning one or more parameters to vary assignment of labels and/or performance of a classification model in terms of sensitivity and specificity.

16. The method of claim 15 wherein the executing step employs a two-tier threshold to assign labels to the image in two steps.

17. The method of claim 16 where the first step comprises a high sensitivity classification model.

18. The method of claim 17 where the second stop comprises a high specificity classification model.

19. A system to perform automatic retinal screening, said system comprising:

an illumination source illuminating a retina;

a retinal camera capturing a retinal image wherein the retinal camera is supported by a platform;

a processor receiving the image and performing an assessment in real time of the image quality; and

adjusting a setting of the camera if the image quality does not meet predetermined quality requirements.

20. A method to automatically determine image quality of a digital retinal image, the method comprising the steps of:

illuminating a retina using an illumination source;

capturing the digital retinal image with a retinal camera supported by a platform;

transmitting the image to a processor;

performing via the processor an assessment in real time of the image quality, comprising;

determining alignment of the image according to one or more imaging protocols;

determining presence and extent of crescents and shadows in the image;

determining quantitative image quality of the image via a classification process trained using examples of visual perception by human experts; and

adjusting the camera if the image quality does not meet a predetermined quality requirement.

21. The method of claim 20 wherein the digital retinal image is inadequate to meet the predetermined image quality requirement for shadowing when a number of pixels in a third quartile of a gradient image of the digital retinal image occupies more than 15 percent of an image area of the digital retinal image.

22. The method of claim 20 wherein the digital retinal image is inadequate to meet the predetermined image quality requirement for crescents when a total number of pixels within 10 percent of saturation is larger than 25 percent of a total number of pixels in the digital retinal image after inverting values of a thresholded image.

23. The method of claim 20 wherein the digital retinal image is inadequate to meet the predetermined image quality requirement for alignment when a macula position is at an angle greater than between 0 degree and 20 degrees below a position of an optic disc with respect to a horizontal axis of the digital retinal image.

24. The method of claim 20 wherein the digital retinal image meets the predetermined image quality requirement for alignment when a distance between an optic disc center and a macula is more than two optic disc diameters and less than three optic disc diameters.

25. The method of claim 20 wherein the digital retinal image meets the predetermined image quality requirement for alignment when an optic disc and a macula are positioned within one-third and two-thirds sections of a vertical digital image.

26. The method of claim 20 wherein the digital retinal image is adequate to meet the predetermined image quality requirement when a macula is positioned within a vertical region limited by temporal vessel arcades.