US 12,614,280 B2Granted

System for estimating primary open-angle glaucoma likelihood

AssigneeThe Regents of the University of California

InventorsLinda Zangwill, Mark Christopher, Christopher Bowd, Akram Belghith

Life of this patent

Estimating the term…

Loading patent…

Abstract

This application describes a system to estimate the likelihood of glaucoma from fundus images. The system may include receiving an optic nerve head region of an original image. A variation to the optic nerve head region are applied. A first primary open angle glaucoma likelihood estimate based on the first varied optic nerve head region is generated using a trained convolutional neural network. A second primary open angle glaucoma likelihood estimate based on the second varied optic nerve head region is generated using a trained convolutional neural network. The first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate are combined into a final primary open angle glaucoma likelihood estimate. The final primary open angle glaucoma likelihood estimate is presented to a user.

Claims18

Dependent claims

A system comprising:

at least one processor;
and
at least one memory storing instructions which, when executed by the at least one processor, cause operations comprising:
training a convolutional neural network with a plurality of varied images to generate a trained convolutional neural network able to detect minor features that improve a predictability of estimating primary open angle glaucoma likelihoods, wherein the training performs an iterative optimization for selecting convolutional neural network weights that optimizes an ability of the trained convolutional neural network to distinguish between healthy and diseased eyes;
receiving an optic nerve head region of an original image;
applying a first variation to the optic nerve head region, wherein the first variation comprises a horizontally mirrored version of the optic nerve head region to imitate a variation expected to occur biologically;
applying a second variation to the optic nerve head region, wherein the second variation comprises noise, a rotation, a color adjustment, and/or a nonlinear geometric transformation;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a first primary open angle glaucoma likelihood estimate based on the first varied optic nerve head region;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a second primary open angle glaucoma likelihood estimate based on the second varied optic nerve head region;
combining, into a final primary open angle glaucoma likelihood estimate, the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate by calculating a combination of the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate using learned weights, wherein the learned weights are associated with a subset of quantitative values generated by the trained convolutional neural network;
and
presenting the final primary open angle glaucoma likelihood estimate on a user interface on a display device.

9 more elements

A method comprising:

training a convolutional neural network with a plurality of varied images to generate a trained convolutional neural network able to detect minor features that improve a predictability of estimating primary open angle glaucoma likelihoods, wherein the training performs an iterative optimization for selecting convolutional neural network weights that optimizes an ability of the trained convolutional neural network to distinguish between healthy and diseased eyes;
receiving an optic nerve head region of an original image;
applying a first variation to the optic nerve head region, wherein the first variation comprises a horizontally mirrored version of the optic nerve head region to imitate a variation expected to occur biologically;
applying a second variation to the optic nerve head region, wherein the second variation comprises noise, a rotation, a color adjustment, and/or a nonlinear geometric transformation;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a first primary open angle glaucoma likelihood estimate based on the first varied optic nerve head region;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a second primary open angle glaucoma likelihood estimate based on the second varied optic nerve head region;
combining, into a final primary open angle glaucoma likelihood estimate, the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate by calculating a combination of the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate using learned weights, wherein the learned weights are associated with a subset of quantitative values generated by the trained convolutional neural network;
and
presenting the final primary open angle glaucoma likelihood estimate on a user interface on a display device.

6 more elements

A non-transitory computer-readable medium comprising instructions which, when executed by at least one processor, cause operations comprising:

training a convolutional neural network with a plurality of varied images to generate a trained convolutional neural network able to detect minor features that improve a predictability of estimating primary open angle glaucoma likelihoods, wherein the training performs an iterative optimization for selecting convolutional neural network weights that optimizes an ability of the trained convolutional neural network to distinguish between healthy and diseased eyes;
receiving an optic nerve head region of an original image;
applying a first variation to the optic nerve head region, wherein the first variation comprises a horizontally mirrored version of the optic nerve head region to imitate a variation expected to occur biologically;
applying a second variation to the optic nerve head region, wherein the second variation comprises noise, a rotation, a color adjustment, and/or a nonlinear geometric transformation;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a first primary open angle glaucoma likelihood estimate based on the first varied optic nerve head region;
generating, using the trained convolutional neural network and based at least on part on the first variation and the second variation, a second primary open angle glaucoma likelihood estimate based on the second varied optic nerve head region;
combining, into a final primary open angle glaucoma likelihood estimate, the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate by calculating a combination of the first primary open angle glaucoma likelihood estimate and the second primary open angle glaucoma likelihood estimate using learned weights, wherein the learned weights are associated with a subset of quantitative values generated by the trained convolutional neural network;
and
presenting the final primary open angle glaucoma likelihood estimate on a user interface on a display device.

6 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/289,673Filed 30 Oct 2019 · us provisional application
Priority: 30 Oct 2018
Published: US 2021/0407096 A130 Dec 2021
Granted: 28 Apr 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification11 codes

G06T 7/11
Segmentation; Edge detection: region-based segmentation
G06T 3/20
Geometric image transformations in the plane of the image: linear translation of whole images or parts thereof, e.g. panning
G06T 3/40
Geometric image transformations in the plane of the image: scaling of whole images or parts thereof, e.g. expanding or contracting
G06T 3/60
Geometric image transformations in the plane of the image: rotation of whole images or parts thereof