Automatically detecting eye type in retinal fundus images

Patent metadata

Field	Value
Publication number	US-10966602-B2
Application number	US-201916564524-A
Country	US
Kind code	B2
Filing date	Sep 9, 2019
Priority date	Oct 23, 2015
Publication date	Apr 6, 2021
Grant date	Apr 6, 2021

Abstract

A computer-implemented method includes obtaining an image of a retinal fundus. A plurality of features is extracted from the image of the retinal fundus. The plurality of features includes at least one feature based on anatomical domain knowledge of the retinal fundus and at least one response of a pre-trained deep convolutional neural network to at least a portion of the image of the retinal fundus. The retinal fundus is determined to belong to a left eye or a right eye, based on an analysis of the plurality of features.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
obtaining a region of interest from an image of a retinal fundus;
passing the region of interest through a stack of N learned convolutional layers of a pre-trained deep convolutional neural network;
discarding a response of an Nth layer of the N learned convolutional layers to the region of interest;
applying principal component analysis on responses of an (N−1)th layer of the N learned convolutional layers to the region of interest to generate a set of principal components,
wherein the Nth layer is trained for one thousand classes of a training dataset, and the (N−1)th layer generates more than four thousand responses as potential unsupervised features;
extracting an unsupervised feature of the retinal fundus from the set of principal components; and
determining whether the retinal fundus belongs to a left eye or a right eye, based on features of the retinal fundus.

2. The computer-implemented method of claim 1 , wherein the region of interest is centered on an optic disc depicted in the image of the retinal fundus.

3. The computer-implemented method of claim 1 , wherein the determining is performed using a support vector machine that has been trained using a plurality of features including the unsupervised feature and a supervised feature of the retinal fundus.

4. The computer-implemented method of claim 3 , wherein the supervised feature comprises a cross-sectional vote-based probability score for the image of the retinal fundus.

5. The computer-implemented method of claim 3 , wherein the supervised feature comprises relative blood vessel densities on the left side and the right side of an optic disc depicted in the image of the retinal fundus.

6. The computer-implemented method of claim 3 , wherein the supervised feature comprises an orientation of a major retinal blood vessel with respect to an optic disc depicted in the image of the retinal fundus.

7. The computer-implemented method of claim 3 , wherein the supervised feature is extracted from an image of retinal blood vessels that has been segmented from the image of the retinal fundus.

8. The computer-implemented method of claim 3 , wherein parameters of the support vector machine comprise C, degree, gamma, kernel, and tolerance, and the parameters are computed using a five-fold cross validation accuracy-based grid search over a validation dataset.

9. A non-transitory machine-readable storage medium encoded with instructions executable by a processor, wherein the instructions cause the processor to perform operations comprising:
obtaining a region of interest from an image of a retinal fundus;
passing the region of interest through a stack of N learned convolutional layers of a pre-trained deep convolutional neural network;
discarding a response of an Nth layer of the N learned convolutional layers to the region of interest;
applying principal component analysis on responses of an (N−1)th layer of the N learned convolutional layers to the region of interest to generate a set of principal components;
extracting an unsupervised feature of the retinal fundus from the set of principal components;
determining whether the retinal fundus belongs to a left eye or a right eye, based on features of the retinal fundus,
wherein the determining is performed using a support vector machine that has been trained using a plurality of features including the unsupervised feature and a supervised feature of the retinal fundus,
wherein the supervised feature comprises an orientation of a major retinal blood vessel with respect to an optic disc depicted in the image of the retinal fundus; and
identifying the orientation by computing an angle between a center of the optic disc and a centroid of the major retinal blood vessel.

10. The non-transitory machine-readable storage medium of claim 9 , wherein the region of interest is centered on an optic disc depicted in the image of the retinal fundus.

11. The non-transitory machine-readable storage medium of claim 10 , wherein the Nth layer is trained for one thousand classes of a training dataset, and the (N−1)th layer generates more than four thousand responses as potential unsupervised features.

12. The non-transitory machine-readable storage medium of claim 9 , wherein the supervised feature comprises a cross-sectional vote-based probability score for the image of the retinal fundus.

13. The non-transitory machine-readable storage medium of claim 9 , wherein the supervised feature comprises relative blood vessel densities on the left side and the right side of an optic disc depicted in the image of the retinal fundus.

14. The non-transitory machine-readable storage medium of claim 9 , wherein the supervised feature is extracted from an image of retinal blood vessels that has been segmented from the image of the retinal fundus.

15. The non-transitory machine-readable storage medium of claim 9 , wherein parameters of the support vector machine comprise C, degree, gamma, kernel, and tolerance, and the parameters are computed using a five-fold cross validation accuracy-based grid search over a validation dataset.

16. A system, comprising:
a processor; and
a non-transitory machine-readable storage medium encoded with instructions executable by the processor, wherein the instructions cause the processor to perform operations comprising:
obtaining a region of interest from an image of a retinal fundus;
passing the region of interest through a stack of N learned convolutional layers of a pre-trained deep convolutional neural network;
discarding a response of an Nth layer of the N learned convolutional layers to the region of interest;
applying principal component analysis on responses of an (N−1)th layer of the N learned convolutional layers to the region of interest to generate a set of principal components,
wherein the Nth layer is trained for one thousand classes of a training dataset, and the (N−1)th layer generates more than four thousand responses as potential unsupervised features;
extracting an unsupervised feature of the retinal fundus from the set of principal components; and
determining whether the retinal fundus belongs to a left eye or a right eye, based on features of the retinal fundus.

Assignees

• IBM

Inventors

• Chakravorty Rajib
• Garnavi Rahil
• Roy Pallab

Classifications

• A61B3/12 Primary

  for looking at the eye fundus, e.g. ophthalmoscopes (A61B3/13 takes precedence) · CPC title

• A61B3/0025 Primary

  characterised by electronic signal processing, e.g. eye models · CPC title

• G06V40/197

  Matching; Classification · CPC title

• G06V40/193

  Preprocessing; Feature extraction · CPC title

• G06T7/0012

  Biomedical image inspection · CPC title