A retinal imaging system includes an eyepiece lens assembly, an image sensor, a dynamic fixation target viewable through the eyepiece lens assembly, and a controller coupled to the image sensor and the dynamic fixation target. The controller includes logic that causes the retinal imaging system to perform operations including: acquiring a first image of the eye, analyzing the first image to determine whether a lateral misalignment between the eye and the eyepiece lens assembly is greater than a threshold misalignment, in response to determining the lateral misalignment is greater than the threshold misalignment, adjusting a visual position of the dynamic fixation target to encourage the eye to rotate in a direction that compensates for the lateral misalignment, and acquiring the retinal image of the eye while the eye is encouraged to rotate towards the direction that compensates for the lateral misalignment.

An image processing device, an image processing method, and an image processing program make it possible to accurately recognize property changes between medical images having different imaging times. A property classification unit classifies each pixel included in a target region of each of a first medical image and a second medical image of which an imaging time is later than an imaging time of the first medical image for the same subject into any one of a plurality of types of properties. A quantitative value derivation unit specifies property pairs each representing a property change between corresponding pixels in at least partial regions within the target regions, and derives a quantitative value in the at least partial regions for each of the property pairs.

For generating a breathing alert is disclosed, a method receives a video stream of a subject. The method further estimates a breathing signal from the video stream. The method determines one of a large-scale motion and/or a breathing event of the subject based on the breathing signal. The method generates an alert if both no breathing event is identified and no large-scale motion of the subject is identified within an event time interval.

The present disclosure provides a method, a device, and a non-transitory computer-readable storage medium for detecting a medical condition of an organ. The method includes obtaining 2D image sequences of the organ in a plurality of different directions and applying a plurality of classification branches to the 2D image sequences. Each classification branch receives a 2D image sequence of one direction and provides a classification result with respect to that direction. Each classification branch includes a convolutional neural network configured to extract first image features from the corresponding 2D image sequence and a recurrent neural network configured to extract second image features from the first image features. The method further includes fusing the classification results provided by the plurality of classification branches for detecting the medical condition.

A digital detection method and system for predicting drug resistance of transgenic maize are disclosed. The method includes acquiring an RGB image, three-dimensional point cloud data and chlorophyll relative content of a maize plant after medicament spraying at a current moment; calculating a pixel ratio and morphological feature according to the RGB image and three-dimensional point cloud data; inputting a detection parameter of the maize plant at the current moment into a series model to predict the detection parameter of the maize plant at a next moment to obtain a graph of change in the detection parameter in a next period; estimating a drug resistance characteristic according to the graph of the change in the detection parameter of the maize plant; and inputting the detection parameter of the maize plant at the current moment into a parallel model to predict the variety of the maize plant.

Systems and methods for processing electronic images from a medical device comprise receiving a first image frame and a second image frame from a medical device, and determining a region of interest by subtracting the first image frame from the second image frame, the region of interest corresponding to a visual obstruction in the first image frame and/or second image frame. Image processing may be applied to the first image frame and/or second image frame based on a comparison between a first area of the first image frame corresponding to the region of interest and a second area of the second image frame corresponding to the region of interest, and the first image frame and/or second image frame may be provided for display to a user.

According some aspects, a secondary device may display secondary audiovisual content along with playback of audiovisual content on a primary device. For example, the secondary device may display an augmented reality application synchronized with the video. Aspects may predetermine a set of frame transition ranges for the video, where each respective frame transition is determined based on frames of the video that are determined to be substantially identical by a frame reference function and frames that are determined to be different. Two frames may be substantially identical even if they are different in the source video. This may be due to shortcomings in the frame reference function, or encoding/compression losses in transmission and playback of the video. Playback may be synchronized based on a first detected frame, but synchronization may be refined upon detecting a frame transition to a second frame that is no longer substantially identical to prior frames.

Systems and methods for performing a medical imaging analysis task using a machine learning based motion model are provided. One or more medical images of an anatomical structure are received. One or more feature vectors are determined. The one or more feature vectors are mapped to one or more motion vectors using the machine learning based motion model. One or more deformation fields representing motion of the anatomical structure are determined based on the one or more motion vectors and at least one of the one or more medical images. A medical imaging analysis task is performed using the one or more deformation fields.

A method for precisely applying chemicals targeted by digital maps developed from remotely sensed data, including: obtaining EOS data through a growing season of a crop growing in a field; processing the EOS data to reflectance values; removing error-inducing effects of atmospheric alteration from the processed EOS data; calculating from the processed EOS data a crop performance index that indicates one or more poor performing areas of the field; generating one or more maps of the crop performance index to allow a user to determine whether each of the one or more poor performing areas of the field are due to biological pests instead of topographic or soil constraints in discrete locations of the field; guiding the user to the one or more poor performing areas of the field using the one or more maps to allow the user to scout the one or more poor performing areas of the field to confirm and identify the biological pests; and providing guidance for a chemical application at the one or more poor performing areas that were confirmed as having the biological pests. Other embodiments are provided.

A device for identifying a size doubling number of a lesion, having: a planar surface defining a data portion that includes data zones, including: a first zone that defines first zone arcs distributed in a concentric configuration, the first zone arcs defining first mutually unique radii, ranging from a first minimum radius to a first maximum radius; the first zone includes first zone size markers, respectively positioned adjacent ones of the first zone arcs, and the first zone size markers respectively identify the size doubling number of the clinical lesion having a size that corresponds to a respective one of the first zone arcs, the size doubling number being one or more of a volume doubling number, an area doubling number and a radius doubling number.