39 results about "Retinal scan" patented technology

The present disclosure relates to diagnosis of ocular diseases by leveraging an attention-based multimodal graphical neural network (AM-GNN) based on multimodal data that includes one or more image-text pairs. Each image of the image-text pair may depict a retinal scan of a subject that may correspond to one or more imaging modalities. A multimodal dataset may be collected from different imaging modalities by data crawling to automatically collect relevant data for the targeted diseases from identified sources. In some aspects, a concept bank, comprising concepts or terminologies associated with the targeted diseases, may be provided to the AM-GNN to generate a multimodal graph. The AM-GNN may be configured to find an association between a set of nodes including image-text pairs, a set of diseases, and the concepts using an attention mechanism. Based on the multimodal graph, an output may be generated, including a predicted disease and related concept(s).

A private identity system is configured to invoke an embedding layer that constructs encoded embeddings from plaintext identification inputs using pre-trained models. Identification inputs can be captured (e.g., biometric, face image, retinal scan, fingerprint, voice, health data, behavioral, etc.) in plaintext versions and transformed into encoded embeddings. The embeddings can be produced to be homomorphic one-way encryptions of the plaintext identification instances. According to some embodiments, the encoded embeddings can be used to identify an entity. The encoded embeddings can be stored in a vector database during enrollment, retrieved by querying the vector database during prediction, and used to establish a match to an identifier, identity, and / or entity. In further embodiments, vector indexes can be used to speed queries executed on the vector database, and achieve improved computation, reduced query speed, and improved identification latency relative to multiple AI model architectures (e.g., embedding and classifier architectures).

A retinal scanning display device including an optical system emits laser light to a pupil of a user, an optical multiplexer individually receives types of laser light having different wavelengths emitted from a plurality of semiconductor lasers, which combines the types of laser light coaxially to output the combined laser light. On the optical multiplexer, a beam shaper shapes the combined laser light into a light beam having a predetermined divergence or convergence angle, and a direction-of-emission changer changes image signal corresponding to an image to be formed, the direction of emission of the light beam output through the beam shaper within at least a plane intersecting the direction of emission. Then, an optical engine formed with the semiconductor lasers, the optical multiplexer, the beam shaper and the direction-of-emission changer guides the laser light emitted from the direction-of-emission changer to the position of the pupil of the user.

Predicting cardiovascular diseases (CVD) may be preceded by subtle changes in various health indicators that may make early predictions useful for timely intervention. The present disclosure relates to prediction of one or more diseases, particularly CVD, for a given subject by leveraging one or more machine-learning models based on longitudinal multimodal data. The techniques, as disclosed herein, may utilize the one or more machine-learning models including one or more feature generators, a temporal aggregator and a prediction model to process the longitudinal multimodal data, generating relevant features, a temporal feature vector and one or more metrics corresponding to one or more diseases associated with the subject. The generated metrics based on the current time and previous time points may be analyzed to assess whether the subject currently has or is at risk of developing one or more diseases.

The invention relates to a method for data glasses (10) with a virtual retinal display (12) (Retinal Scan Display), comprising at least one optical system (14) which projects an image onto a retina (16) of an eye (18) of a wearer of the data glasses (10), and comprising at least one laser feedback interferometry (LFI) sensor (20) which emits an infrared laser beam (22) which passes through at least part of the optical system (14). It is proposed that the LFI sensor (20) detects reflection signals (24, 24') of the infrared laser beam (22) and detects contaminations of components (26) of the virtual retinal display (12), in particular of the optical system (14), from the reflection signals (24).

The invention relates to a method for calibrating a data glasses (12) having a virtual retinal display (10) (retinal scan display), wherein scanned images (16) of the virtual retinal display (10) are generated by a laser projector (14) and projected as beamlets into an eyebox of the data glasses (12) via an optical system (18) of the data glasses (12), wherein calibration images (20) and / or calibration patterns (22) are output to a wearer of the data glasses (12) by means of the beamlet, and wherein user feedback, which is given by the wearer of the data glasses (12) based on his perception, is recorded and used for the calibration of the data glasses (12). It is proposed that, for the calibration of the data glasses (12), an automated mechanical actuator (24) for adjusting a position of the beamlet is controlled by a computing unit (26) of the data glasses (12), or an actuation of a manual mechanical actuator (28) for manually adjusting the position of the beamlet is initiated, or a manual reconfiguration of the data glasses (12) for manually adjusting the position of the beamlet is initiated, or the laser projector (14) and / or the optical system (18) for adjusting a brightness, a sharpness or a hue of the reproduced calibration images (20) and / or calibration patterns (22) are controlled by the computing unit (26) of the data glasses (12).

The invention relates to smart glasses (1a) designed for a virtual retinal scan display for a user of the smart glasses (1a). The smart glasses (1a) comprise a projector unit (14) with a time-modulated light source (2) for generating at least one first light beam (8, 11) and with a controllable deflection device (16) for the at least one first light beam (8, 11) for scanning and projecting an image content. The smart glasses (1a) also comprise a first deflection unit (10) onto which the image content is projected and which directs the projected image content onto one eye (9) of a user. Furthermore, the smart glasses (1a) comprise at least one laser feedback interferometry sensor (5) arranged in or on the smart glasses (1a), which is configured to emit a second modulated light beam (12), in particular an infrared beam. Furthermore, the data glasses (1a) have a computing unit (13).The laser feedback interferometry sensor (5) serves to detect a second light beam reflected from an object (7) located between the controllable deflection device (16) and the first deflection unit (10), or on a side of the first deflection unit (10) facing away from the eye (9) of the user of the data glasses (1a). The processing unit (13) serves to detect the object (7) depending on the reflected second light beam.

Disclosed herein are systems, methods, devices, and media for carrying out detection of ophthalmic and systemic diseases and disorders. Deep learning algorithms enable the automated analysis of ophthalmic images such as retinal scans to generate accurate detection of various diseases and disorders. Point-of-care implementations allow for rapid and efficient detection outside of the clinical setting.

A drinking vessel 2 comprises means to determine and verify the identity of a current user and access one or more accounts 14 of the user based on the determination and verification. The user account(

The invention provides a real-time verification system for operators of aircraft, vehicles, and other modes of transportation for safety and security before commencing travel. The system comprises a processing device installed in the preferred mode of transportation, and includes a processor, memory, and an AI model. The processor integrates modules for receiving biometric data (fingerprint, facial, or retinal scans), verifying identities against a cloud database of operator data and the Terrorist Screening Database (TSDB), and communicating results to Air Traffic Control (ATC). The AI model performs advanced analytics, monitoring verification data against comprehensive databases, and predicting potential issues by analyzing historical trends. It generates detailed reports and alerts for ATC and other authorities, contributing to both immediate safety and long-term improvements. This system ensures that only authorized operators can control the vehicle while fostering ongoing safety enhancements through predictive analytics and trend identification.

A multi-instances emission system for retina scanning based near eye display is disclosed. The multi-instances emission system comprises: a right light emitter and a left light emitter for respectively emitting a right light signal and a left light signal to a first eye and a second eye of a viewer; at least one optical duplicator for respectively dividing the right light signal and the left light signal into multiple light instances; and at least one light signal redirector for changing direction of a plurality of light instances of the right light signal and a plurality of light instances of the left light signal, such that the plurality of light instances of the right light signal have optical paths that converge to form a first converging point and f the plurality of light instances of the left light signal have optical paths that converge to form a second converging point.

The invention relates to a metasurface reflector, a retina scanning display and a manufacturing method of the metasurface reflector. Provided is a metasurface reflector or the like which has high process resistance and is not susceptible to deterioration in reflection characteristics. A nanostructure (11) is provided with a protective layer (8), a first metal layer (10) constituting a nanostructure, a dielectric layer (20), and a second metal layer (30), the protective layer (8) covers the first metal layer (10), the dielectric layer (20) is positioned between the first metal layer (10) and the second metal layer (30), and the nanostructure (11) contains Ag and Zn, or Ag and Sn.

A computer-implemented method for retinal scan image classification, the method including: receiving an input dataset including at least one retinal scan, the retinal scan acquired using one of a plurality of imaging modalities; passing the input dataset through an ensemble network, the ensemble network including a plurality of trained convolutional neural networks, each convolutional neural network configured to classify retinal scans acquired using one of a plurality of imaging modalities based on a plurality of ocular lesions, each convolutional neural network generating a presence probability for each ocular lesion in the retinal scan, the retinal scan being processed using a convolutional neural network corresponding to the imaging modality used to acquire the retinal scan; and generating an ensembled presence probability for each ocular lesion in the retinal scan.

An optical system for a virtual retinal scan display. A plurality of first exit pupils and replicated, second exit pupils are produced. A computing unit is designed, depending on second light beams, backscattered by an outer eye surface and detected using a first sensor, in a first infrared wavelength range or the modulation of the power of a second light source, and / or depending on third light beams, backscattered by the outer eye surface, in a second infrared wavelength range or a modulation of a power of a third light source and / or depending on backscattered third light beams detected using a second sensor, or the modulation of the power of the third light source, to ascertain the positions of the first exit pupils relative to a pupil center and the positions of the second exit pupils relative to the pupil center.

A device for detecting potential health issues includes a retinal scanner configured to capture retinal data of the user, a blood pressure sensor configured to capture blood pressure data of the user, an alert module configured to alert the user of a potential health risk, and at least one processor communicatively coupled to the retinal scanner, the blood pressure sensor, and the alert module.

A method for projecting image content onto the retina of a user using an optical system. Image data are captured using the image processing device. A first subset of the image data is blanked, thereby generating an active second subset of image data. The first and second subsets make up the total amount of the image data. The projector unit is controlled using the image data such that there is always only one eye box, generated on a common imaging path, with the same active image data arranged in the region of a first pupil region of the user. The first pupil region surrounds the pupil center. A portion of the eye box is arranged with a portion of the active image data within a second pupil region of the user. The second pupil region surrounds the pupil center. The second pupil region is arranged within the first pupil region.

A retina scanning display device 300 produces an image on a retina of a user’s eye 304. The device includes a light source 302, collimating optics 306, a reflective combiner 318 and a scanning element

A clinical decision support system for retinal diseases processes OCT retinal scan images to enhance certain aspects of speckle, which has been found to include relevant information that can be used to enhance the diagnosis and classification of the severity and progression of any disease found in the retina scan image. A copy of the original OCT retinal scan image is processed using a shifted window transformer block to enhance the relevant speckle information, and combine it with the original image, which is then evaluated by a trained system to classify the retinal features, including the speckle.

SOLUTION: A multi-instance light emission system for a retinal scanning near-eye display is disclosed. And at least one light direction modifier for modifying a light emission direction of the plurality of light emitting units such that optical paths of a first collimated light signal and a third collimated light signal from the plurality of collimated light signals emitted by the plurality of light emitting units have optical paths or optical path extensions that converge to form a first convergence point, and such that optical paths of a second collimated light signal and a fourth collimated light signal from the plurality of collimated light signals have optical paths or optical path extensions that converge to form a second convergence point.SELECTED DRAWING: Figure 4A

An optical system for a virtual retinal scan display. The optical system includes an image source which supplies image data, an image processing device for the image data, a projector unit having a light source modulatable in time for generating a light beam, and an actuable refraction device for the light beam for the scanning projection of the image content, a deflection unit onto which the image content is able to be projected and which is designed to guide the projected image content onto an eye of a user, and an adaptive optical element for modifying a beam divergence, which is situated in the optical path between the light source and the deflection unit. The adaptive optical element is able to be actuated so that the beam divergence of the light beam is variable as a function of the angle of incidence of the light beam on the deflection unit.

Method of tracking the movements of the retina based on images of the retina acquired through optical coherence tomography. The tracking method includes the following steps: - obtaining at least one pair of B-scan images of the retina, wherein each B-scan image of the retina is acquired for a corresponding optical scan of the retina during which an illumination beam is projected onto the retina and is moved along an optical scanning trajectory having the shape of a closed curve; - identifying pairs of characteristic areas having a similar speckle pattern in the B-scan images of said at least one pair of images selected; - calculating tracking data indicative of the movements of the retina based on the position of said characteristic areas in the B-scan images of the retina of said at least one pair of images selected.

The present disclosure relates to leveraging one or more machine-learning models to generate a cardiovascular risk score that corresponds to a prediction of a present or future presence or progression of cardiovascular disease for a given subject based on retinal scan data of the subject. The one or more machine-learning models may include (for example) a neural network-based feature generator configured to generate features by processing the retinal scan data and a prediction model that takes feature vectors and predict cardiovascular risk-factors as auxiliary target variables along with cardiovascular risk scores for a current and / or future point in time. The feature generator and the prediction model may be trained jointly in an end-to-end manner by computing a loss function that reduces the discrepancy between actual and predicted values and may apply weights to each predicted label based on its contribution to the prediction accuracy of cardiovascular risk scores.

A private identity system is configured to invoke an embedding layer that constructs encoded embeddings from plaintext identification inputs using pre-trained models. Identification inputs can be captured (e.g., biometric, face image, retinal scan, fingerprint, voice, health data, behavioral, etc.) in plaintext versions and transformed into encoded embeddings. The embeddings can be produced to be homomorphic one-way encryptions of the plaintext identification instances. According to some embodiments, the encoded embeddings can be used to identify an entity. The encoded embeddings can be stored in a vector database during enrollment, retrieved by querying the vector database during prediction, and used to establish a match to an identifier, identity, and / or entity. In further embodiments, vector indexes can be used to speed queries executed on the vector database, and achieve improved computation, reduced query speed, and improved identification latency relative to multiple Al model architectures (e.g., embedding and classifier architectures).

The invention relates to smart glasses (1a) designed for a virtual retinal scan display of a user of the smart glasses (1a). The smart glasses (1a) comprise a projector unit (14) having a light source (2), which can be modulated over time, for generating at least one first light beam (8, 11) and having an activatable deflection device (16) for the at least one first light beam (8, 11) for scanningly projecting image content. The smart glasses (1a) also comprise a first deflector unit (10) onto which the image content is projected and which deflects the projected image content onto an eye (9) of the user. The smart glasses (1a) additionally comprise at least one laser feedback interferometry sensor (5), which is arranged in or on the smart glasses (1a) and is designed to emit a second modulated, in particular infrared, light beam (12). Furthermore, the smart glasses (1a) comprise a computing unit (13). The laser feedback interferometry sensor (5) is intended to capture a second light beam reflected back from an object (7) which is arranged between the activatable deflection device (16) and the first deflector unit (10) or on a side of the first deflector unit (10) facing away from the eye (9) of the user of the smart glasses (1a). The computing unit (13) is intended to detect the object (7) on the basis of the reflected-back second light beam.

A system comprises a memory communicatively coupled to at least one processor. The at least one processor is configured to receive a code-based data exchange request from a user. The code-based data exchange request comprises first user information associated with a user profile. Further, the at least one processor is configured to trigger a request for a retinal scan of the user in response to receiving the code-based data exchange request and receive a result of the retinal scan of the user. The result comprises second user information. The at least one processor is configured to determine that the code-based data exchange request is authentic in response to determining that the first user information and the second user information are associated with the user profile, convert the code-based data exchange request to a light-based data exchange request, and generate a light-guided contract based on the light-based data exchange request.

A multi-instance emission system for retina scanning based near-eye display is disclosed. The system comprises a plurality of light emitting units for respectively emitting collimated light signals to a first eye or a second eye of a viewer; and at least one light direction modifiers for modifying light emitting direction of the plurality of light emitting units such that optical path of a first collimated light signal and a third collimated light signal from a plurality of collimated light signals emitted by the plurality of light emitting units have optical paths or optical path extensions that converge to form a first converging point, and a second collimated light signal and a fourth collimated light signal from the plurality of collimated light signals have optical paths or optical path extensions that converge to form a second converging point.

A method for calibrating, in an at least largely automatic manner, a virtual retinal scan display for a pair of smart glasses including multiple exit pupils that are spaced apart from one another, in particular without overlapping, in a provided pupil plane of the virtual retinal scan display. In at least one calibration step, an eye-like camera is automatically translated and / or rotated in the provided pupil plane of the virtual retinal scan display in such a way that at least two exit pupils of the virtual retinal scan display, preferably all exit pupils of the virtual retinal scan display, are each sequentially captured from different positions and / or angles of the eye-like camera.

The invention relates to an optical system (1) for a virtual retinal scan display, which includes a lighting device (13). The lighting device (13) has at least one first light source for emitting a first light beam (14). Furthermore, the optical system (1) has a controllable deflection unit (12) as a first optical component for the first light beam (14) for scanning the image content. In addition, the optical system (1) has a holographic optical element (6) which is configured to deflect the first light beam (3a, 3b) onto a retina (4) of a user of the optical system (1). Depending on the angle of incidence (22a, 22b) of the first light beam (3a, 3b) on the holographic optical element (6), the holographic optical element (6) has areas with different diffraction efficiencies for the first light beam (3a, 3b).Furthermore, the optical system (1) has a first intensity filter (16a, 16b, 16c, 16d) which is designed to adjust the first intensity of the first light beam (3a, 3b) such that the light beam deflected from the holographic optical element (6) to the retina (4) strikes the user's retina (4) with the same intensity at time points over a first scan area (20a) as the first emitted light beam (14).

An optical apparatus uses polarized light to interrogate birefringence properties of the retina to detect the fixation condition of the eye by sensing characteristic birefringence patterns of the retinal structures. Optical components, as well as optional additional components, interfere with the polarization measurements by introducing unwanted retardance which alters the polarization state of the light entering the eye and of the light reaching the detection system. Compensating retarders are provided to nullify the effect of unwanted retardance in the forward and return light paths so the polarization states of the light entering the eye and the light reaching the detection system are not contaminated by the effects of the unwanted retardance. Mueller matrices are used to mathematically calculate the parameters for the compensating retarders for the unwanted retardance. A variable retarder system may also be provided to compensate for the corneal birefringence of the eye via feedback control.

The invention relates to an optical system (1) for a virtual retinal scan display which has an illumination device (13), said illumination device (13) having at least one first light source for emitting a first light beam (14). Furthermore, the optical system (1) comprises a first controllable deflection unit (12) as a first optical component for the first light beam (14) for the scanning projection of the image content, and the optical system (1) has a holographic optical element (6) which is designed to deflect the scanning first light beam (3a, 3b) onto a retina (4) of a user of the optical system (1). The holographic optical element (6) has regions with different diffraction efficiencies for the first light beam (3a, 3b) according to the angle of incidence (22a, 22b) of the first light beam (3a, 3b) on the holographic optical element (6). The optical system (1) additionally has at least one first intensity filter (16a, 16b, 16c, 16d) which is designed to adapt a first intensity of the first light beam (3a, 3b) in such a way that, when the intensity of the first emitted light beam (14) is constant, the light beam deflected by the holographic optical element (6) onto the retina (4) is also incident on the user's retina (4) with a constant intensity at points in time over a first scanning region (20a).