Digital Diagnostics Inc.

University of Iowa Research Foundation

The United States Government As Represented by the Department of Veterans Affairs

Idx, LLC

Meindert Niemeijer

Michael D Abramoff

Warren Clarida

Joseph M Reinhardt

Ryan Earl Rohret Amelon

Abhay Shah

Jacob Patrick Suther

Milan Sonka

Xiayu Xu

Ben Clark

Eric Talmage

Ryan Amelon

Tay Stutts

Timothy Dinolfo

John Casko

Michael D Abràmoff

Warrent Claride

Sangyeol Lee

Bram Van Ginneken

The methods and systems provided can automatically determine an Arteriolar-to-Venular diameter Ratio, AVR, in blood vessels, such as retinal blood vessels and other blood vessels in vertebrates. The AVR is an important predictor of increases in the risk for stroke, cerebral atrophy, cognitive decline, and myocardial infarct.

Automated determination of arteriovenous ratio in images of blood vessels

Systems and methods are provided herein for minimizing retinal exposure to flash during image gathering for diagnosis. In an embodiment, a system captures a plurality of retinal images of different retinal regions. The system determines that a first portion of a first image does not meet a criterion while a second portion of the first image does meet the criterion, identifies a portion of the retina depicted in the first portion that does not meet the criterion, and determines whether the portion of the retina is depicted in a third portion of a second image and whether the third portion meets the criterion. Responsive to determining that the third portion meets the criterion, the system performs the diagnosis. Responsive to determining that the portion of the retina is not depicted in the second image, the system captures an additional image of the retinal region.

Image retention and stitching for minimal-flash eye disease diagnosis

Disclosed is a system for qualifying medical images submitted by user for diagnostic analysis comprising: an image input module, configured to receive one or more image input by a user; an image protocol conformation module, configured to receive the one or more images from the image input module, and further configured to analyze each of the one or more images for conformity with a predefined protocol and wherein images that do not conform to the predefined protocol are flagged as non-conforming images; an image output module, configured to identify to the user each of the one or more images flagged as non-conforming and prompting the user to resubmit a new image for each of the non-conforming images; and an image resubmission module, configured to receive the user resubmitted image and provide the resubmitted image to the image protocol conformation module.

System and methods for qualifying medical images

Systems and methods are disclosed herein for detecting eye alignment during retinal imaging. In an embodiment, the system receives an infrared stream from an imaging device, the infrared stream showing characteristics of an eye of a patient. The system determines, based on the infrared stream, that the eye is improperly aligned at a first time, and outputs sensory feedback indicative of the improper alignment. The system detects, based on the infrared stream at a second time later than the first time, that the eye is properly aligned, and receives an image of a retina of the properly aligned eye from the imaging device.

Using infrared to detect proper eye alignment before capturing retinal images

Systems and methods are disclosed herein for adjusting flash intensity based on retinal pigmentation. In an embodiment, a processor determines a retinal pigmentation of a retina of an eye positioned at an imaging device. The processor commands the imaging device to adjust an intensity of a flash component from a first intensity to a second intensity based on the retinal pigmentation. The processor commands the imaging device to capture an image that is lit by the flash component at the second intensity, and receives the image from the imaging device.

Dynamic adjustment of flash intensity based on retinal pigmentation

A device receives an input image of a portion of a patient's body, and applies the input image to a feature extraction model, the feature extraction model comprising a trained machine learning model that is configured to generate an output that comprises, for each respective location of a plurality of locations in the input image, an indication that the input image contains an object of interest that is indicative of a presence of a disease state at the respective location. The device applies the output of the feature extraction model to a diagnostic model, the diagnostic model comprising a trained machine learning model that is configured to output a diagnosis of a disease condition in the patient based on the output of the feature extraction model. The device outputs the determined diagnosis of a disease condition in the patient obtained from the diagnostic model.

Autonomous diagnosis of a disorder in a patient from image analysis

Provided is a method for training a neural network to detect features in a retinal image. The method may include the steps of: combining and randomizing feature images into a Training data set; combining and randomizing the feature images into a testing dataset; training a plurality of neural networks having different architectures using a subset of the training dataset while testing on a subset of the testing dataset; identifying the best neural network based on each of the plurality of neural networks performance on the testing data set; inputting images to the best neural network and identifying a limited number of false positives and false negative and adding the false positives and false negatives to the training dataset and testing dataset; and repeating the foregoing steps until an objective performance threshold is reached.

Systems and methods for feature detection in retinal images

Methods and systems for image registration implementing a feature-based strategy that uses a retinal vessel network to identify features, uses an affine registration model estimated using feature correspondences, and corrects radial distortion to minimize the overall registration error. Also provided are methods and systems for retinal atlas generation. Further provided are methods and systems for testing registration methods.

Optimal registration of multiple deformed images using a physical model of the imaging distortion

Disclosed is an automated method which can detect images containing red lesions. First, each image is preprocessed; next, candidate objects that may represent red-lesions are extracted; and in the final stage the probability for each candidate to represent a red-lesion is estimated using a classifier and a large set of specifically designed features.

Growing community of inventors

Prairie Village, KS, United States of America

Meindert Niemeijer