Carl Zeiss Meditec Gmbh

Carl Zeiss Meditec Ag

Mary K Durbin

Utkarsh Sharma

Sophie Kubach

Lin An

Matthew J Everett

Paul Francis Stetson

Scott A Meyer

Jonathan David Oakley

Robert W Knighton

Lars Omlor

Conor Leahy

Bagrat Amirbekian

Niranchana Manivannan

Keith E O'Hara

Homayoun Bagherinia

Harihar Narasimha-Iyer

Tilman Schmoll

Nathan Shemonski

Luis De Sisternes

Vincent Michael Patella

Qienyuan Zhou

Siddharth Srivastava

Jennifer Y Luu

David Nolan

Keith Brock

Priya Kulkarni

Gregory G Anderson

Warren Lewis

Kenneth Bailey Freund

Xiaoyu Xu

Jeffrey Schmidt

Michael Chen

Ali Fard

Shamika Gune

Kevin Meng

Katherine Makedonsky

Arindam Bhattacharya

An OCT system includes a machine learning (ML) model trained to receive a single OCT scan/image and provide an image translation and/or denoise function. The ML model may be based on a neural network (NN) architecture including a series of encoding modules in a contracting path followed by a series of decoding modules in an expanding path leading to an output convolution module. An intermediate error module determines a deep error measure, e.g., between a training output image and at least one encoding module and/or decoding module, and an error from the output convolution module is combined with the deep error measure. The NN may be trained using true averaged images as ground truth, training outputs. Alternatively, the NN may be trained using randomly selected, individual OCT images/scans as training outputs.

System for OCT image translation, ophthalmic image denoising, and neural network therefor

An ophthalmic imaging system provides an automatic focus mechanism based on the difference of consecutive scan lines. The system also provides of user selection of a focus point within a fundus image. A neural network automatically identifies the optic nerve head in an FA or ICGA image, which may be used to determine fixation angle. The system also provides additional scan tables for multiple imaging modalities to accommodate photophobia patients and multi-spectrum imaging options.

Patient tuned ophthalmic imaging system with single exposure multi-type imaging, improved focusing, and improved angiography image sequence display

An optical coherence tomography (OCT) unit provides a mechanism/method for automatic artery and vein identification. The method may be embodied by specialized algorithm or by a machine learning model, such as a support vector machine or neural network. The present method may identify a vascular configuration that forms a vortex structure in the deep capillary plexus (DCP) of an OCT-based (e.g., OCTA) image of a retina. The identified vortex may then be designated a vein, and other vascular structures from other plexuses, such as from the superficial vascular plexus (SVP), that connect to the identified vortex structure are likewise designated veins. Additional criteria, some based on the vein designations, may be used to identify arteries.

OCT-based retinal artery/vein classification

An automated segmentation and identification system/method for identifying geographic atrophy (GA) phenotypic patterns in fundus autofluorescence images. A hybrid process combines a supervised pixel classifier with an active contour algorithm. A trained, machine learning model (e.g., SVM or U-Net) provides initial GA segmentation/classification, and this is followed by Chan-Vese active contour algorithm. The junctional zones of the GA segmented area are then analyzed for geometric regularity and light intensity regularity. A determination of GA phenotype is made, at least in part, from these parameters.

Segmentation and classification of geographic atrophy patterns in patients with age related macular degeneration in widefield autofluorescence images

An ophthalmic imaging system provides a user interface to facilitate the montaging of scan images collected with various imaging modalities, such as images collected with a fundus imaging system or an optical coherence tomography (OCT) system. The amount of each constituent image used in the montage is dependent upon its respective quality. During the collecting of scans (constituent images) for montaging, any scan may be designated for rescanning, such as if its current quality is deemed less than sufficient. In the case of using an OCT system to collect constituent images (e.g., cube scans), the scanned region of a constituent image may be modified based on physical characteristics of the eye being scanned.

Systems and methods for improved montaging of ophthalmic imaging data

Methods for improved acquisition and processing of optical coherence tomography (OCT) angiography data are presented. One embodiment involves improving the acquisition of the data by evaluating the quality of different portions of the data to identify sections having non-uniform acquisition parameters or non-uniformities due to opacities in the eye such as floaters. The identified sections can then be brought to the attention of the user or automatically reacquired. In another embodiment, segmentation of layers in the retina includes both structural and flow information derived from motion contrast processing. In a further embodiment, the health of the eye is evaluating by comparing a metric reflecting the density of vessels at a particular location in the eye determined by OCT angiography to a database of values calculated on normal eyes.

Acquisition and analysis techniques for improved outcomes in optical coherence tomography angiography

An ophthalmic image diagnostic tool and method submits a test image to a neural network trained to identify abnormal regions of an ophthalmic image, to distinguish between multiple types of abnormalities, and to associate an abnormality type with each identified potentially abnormal region. Each potentially abnormal region in the test image is highlighted, and in response to a user-selection of a highlighted region, a previously diagnosed sample (e.g., from a library of samples) of the abnormality type associated with the selected highlighted region is displayed.

Identifying suspicious areas in ophthalmic data

Methods and systems in ophthalmic imaging are presented that increase the sensitivity of automated diagnoses by the use of a combination of both functional and structural information derived from a variety of ophthalmic imaging modalities. An example method to analyze image data of an eye of a patient includes processing a first image dataset to obtain one or more functional metrics; processing a second image dataset to obtain one or more structural metrics; comparing the one or more structural metrics to the one or more functional metrics; and processing the results of said comparison to derive the probability of a disease or normality of the eye.

Methods and systems for combined morphological and angiographic analyses of retinal features

Various systems and methods for improved OCT angiography imaging are described. An example method of identifying intraretinal fluid in optical coherence tomography (OCT) image data of an eye includes collecting OCT image data using an OCT system. The data includes at least one cluster scan containing OCT image data collected at approximately same set of locations on the sample. A first motion contrast image is generated by applying a first OCT angiography processing technique to the cluster scan to highlight motion contrast in the sample. A second motion contrast image is generated by applying a second OCT angiography processing technique to the cluster scan to highlight motion contrast in the sample. An image displaying intraretinal fluid in the eye is generated using the first and second motion contrast images and then displayed or stored or a further analysis thereof.

Methods for high sensitivity flow visualization

Systems and methods to improve the detection and classification of inflammation in the eye are presented. The inflammatory markers in image data can be graded by comparing identified and extracted characteristics from the images with characteristics derived from a set of images of eyes from a general population of subjects. The image data can be divided into sub-regions for analysis to better isolate the true inflammatory markers from the impacts of cataracts or other opacities. In another embodiment, the location of an imaging beam can be controlled to minimize the impact of lens opacities from the collected data used to analyze the inflammation state of an eye.

Methods and systems for automatic detection and classification of ocular inflammation

Systems and methods for improving the assessment of disruption or abnormalities to retinal layers are presented. The disruptions are detected by analyzing at least one segmented boundary of optical coherence tomography data. Several different types of analysis can be used alone or in combination to make an assessment of the level of disruption to the particular boundary or layer defined by the boundary. The results can be presented as an end face image and quantified to report an amount of disruption. In one embodiment, a method for determining the disruption to the photoreceptor outer segment is described.

Assessment of retinal disruption

Methods for analyzing optical coherence tomography (OCT) images of the macula to reduce variance and improve disease diagnosis are presented. One embodiment of the invention is directed towards selecting analysis locations and data segmentation techniques to take advantage of structural homogeneities. Another embodiment is directed towards reducing the variance in a collection of normative data by transforming the individual members of the database to correspond to a Standard Macula. Variations in foveal size are corrected by radial transformation. Variations in layer thickness are corrected by axial shifting. Diagnosis is performed by comparing OCT images from a patient to the improved normative database.

Methods to reduce variance in OCT analysis of the macula

Acquistion and analysis techniques for improved outcomes in optical coherence tomography angiography

Methods for analyzing and visualizing OCT angiography data are presented. In one embodiment, an automated method for identifying the foveal avascular zone in a two dimensional en face image generated from motion contrast data is presented. Several 3D visualization techniques are presented including one in which a particular vessel is selected in a motion contrast image and all connected vessels are highlighted. A further embodiment includes a stereoscopic visualization method. In addition, a variety of metrics for characterizing OCT angiography image data are described.

Analysis and visualization of OCT angiography data

Growing community of inventors

San Francisco, CA, United States of America

Mary K Durbin