The present disclosure is directed to systems and methods for measuring and analyzing pupillary psychosensory responses. An electronic device is configured to receive video data with at least two frames. The electronic device then locates one or more eye objects in the video data and determine pupil and iris sizes of the one or more eye objects. The electronic device determines the pupillary psychosensory responses of the one or more eye objects by tracking a ratio of pupil diameter to iris diameter throughout the video. Several metrics for the pupillary psychosensory responses can be determined (e.g., velocity of change of the ratio, peak to peak amplitude of the change in ratio over time, etc.). These metrics can be used as measures of an individual's cognitive ability and mental health in a single session or tracked throughout multiple sessions.

An ophthalmic information processing apparatus includes a specifying unit and an image deforming unit. The specifying unit is configured to specify a three-dimensional position of each pixel in a two-dimensional front image depicting a predetermined site of a subject's eye, based on OCT data obtained by performing optical coherence tomography on the predetermined site. The image deforming unit is configured to deform the two-dimensional front image, by changing position of at least one pixel in the two-dimensional front image based on the three-dimensional position, to generate a three-dimensional front image.

An ophthalmic information processing apparatus includes a specifying unit and an image deforming unit. The specifying unit is configured to specify a three-dimensional position of each pixel in a two-dimensional front image depicting a predetermined site of a subject's eye, based on OCT data obtained by performing optical coherence tomography on the predetermined site. The image deforming unit is configured to deform the two-dimensional front image, by changing position of at least one pixel in the two-dimensional front image based on the three-dimensional position, to generate a three-dimensional front image.

Deep learning methods and systems for detecting biomarkers within optical coherence tomography volumes using such deep learning methods and systems are provided. Embodiments predict the presence or absence of clinically useful biomarkers in OCT images using deep neural networks. The lack of available training data for canonical deep learning approaches is overcome in embodiments by leveraging a large external dataset consisting of foveal scans using transfer learning. Embodiments represent the three-dimensional OCT volume by “tiling” each slice into a single two dimensional image, and adding an additional component to encourage the network to consider local spatial structure. Methods and systems, according to embodiments are able to identify the presence or absence of AMD-related biomarkers on par with clinicians. Beyond identifying biomarkers, additional models could be trained, according to embodiments, to predict the progression of these biomarkers over time.

Provided are apparatus and method for predicting biometrics using a fundus image. The method for predicting biometrics using a fundus image includes steps of preparation of a plurality of learning fundus images, generation of a learning model for predicting corresponding biometrics using the prepared data based on at least one characteristic of the fundus reflected in the prepared plurality of learning fundus images, reception of a prediction target of fundus image, and prediction of the biometrics of the subject of the prediction target of fundus image by using the generated learning model.

A hand-held sized ocular and neurological screening device, system and method, the screening device comprising an eyepiece and a hand-held housing, the housing comprising a tubular stimulus chamber defining a light stimulus channel, wherein an illumination source is configured to provide light stimulus towards an opening through the light stimulus channel and an operational chamber comprising an infrared camera positioned outside the stimulus channel and inclined towards the opening, the infrared camera is configured to capture images of the pupils and eye movements through the opening without interfering with the light stimulus and a controller configured to receive the captured images from the infrared camera. The hand-held sized device can include a clip-on fixture for fixing the device onto a table, a desktop, or any portable ophthalmic apparatus.

A hand-held sized ocular and neurological screening device, system and method, the screening device comprising an eyepiece and a hand-held housing, the housing comprising a tubular stimulus chamber defining a light stimulus channel, wherein an illumination source is configured to provide light stimulus towards an opening through the light stimulus channel and an operational chamber comprising an infrared camera positioned outside the stimulus channel and inclined towards the opening, the infrared camera is configured to capture images of the pupils and eye movements through the opening without interfering with the light stimulus and a controller configured to receive the captured images from the infrared camera. The hand-held sized device can include a clip-on fixture for fixing the device onto a table, a desktop, or any portable ophthalmic apparatus.

An implementation cost of a line-scan optical coherence tomography (OCT) apparatus is reduced by miniaturizing a scanning mirror and using a light source with relaxed requirement in intensity uniformity. The mirror reflects a probe light beam to different parts of a sample for line-scanning the sample. A line-compressing lens compresses the probe light beam's cross-sectional length before the beam reaches the mirror, allowing the mirror to be miniaturized to reflect only the compressed beam. In generating a linear light beam that gives the probe light beam, a cascade of collimating lens, Powell lens and focusing lens generates the linear light beam from a raw light beam of a point source. A slit further filters the linear light beam to remove a peripheral portion thereof such that the linear light beam is substantially uniform in intensity even if an asymmetrical divergent light source is used.

The invention relates to a computing system and a computer-implemented method for classifying images of retina of eyes of subjects. A captured image of a retina is processed to obtain a plurality of different segmented images each having different selected portions of the captured image using different selection rules. The multiple segmented images are provided to respective dedicated machine learning models to output an image classification based on the respective segmented images provided as input. An ensemble classification is determined based on the multiple classifications obtained by means of the multiple trained machine learning models.

A device has a light guide portion and a light collector portion. The light guide portion is cup shaped. The wall of the cup has a cross section defined by sections of two ellipses disposed in a predefined manner with each other. The light collector portion is also cup-shaped, inverted with reference to the light guide portion, by a section of an ellipse and straight lines defined with reference to the light guide portion. The device radiates an annular illumination at the rim of the cup through total internal reflection of light from an LED, collected by the light collector portion. The device is made of a clear, colourless, substantially transparent material by injection moulding, one example being Polycarbonate. An annular light source system and a fundus camera using such a system are also disclosed.