Abstract: An image pre-processing method and an image processing apparatus for a fundoscopic image are provided. A region of interest (ROI) is obtained from a fundoscopic image to generate a first image. The ROI is focused on an eyeball in the fundoscopic image. A smoothing process is performed on the first image to generate a second image. A value difference between neighboring pixels in the second image is increased to generate a third image.

Abstract: A system and method for maximizing bandwidth in an uplink for a 5G communication system is disclosed. Multiple end devices generate image streams. A gateway is coupled to the end devices. The gateway includes a gateway monitor agent collecting utilization rate data of the gateway and an image inspector collecting inspection data from the received image streams. An edge server is coupled to the gateway. The edge server includes an edge server monitor agent collecting utilization rate data of the edge server. An analytics manager is coupled to the gateway and the edge server. The analytics manager is configured to determine an allocation strategy based on the collected utilization rate data from the gateway and the edge server.

Abstract: A method and a system for establishing a light source information prediction model are provided. A plurality of training images are captured for a target object. A white object is attached on the target object. True light source information of the training images is obtained according to a color of the white object in each of the training images. A neural network model is trained according to the training images and the true light source information, and a plurality of pieces of predicted light source information is generated according to the neural network model during the training. A learning rate for training the neural network model is adaptively adjusted based on the predicted light source information.

Abstract: A method for estimating a ventricular volume is provided and includes: obtaining a left ventricular mask image corresponding to a heart ultrasound image, where the left ventricular mask image is a binary image; finding 3 reference point pixels in the left ventricular mask image, where each of the reference point pixels has a first value, each of the reference point pixels is surrounded by N surrounding pixels, and the surrounding pixels of each of the reference point pixels include N1 first surrounding pixels with the first value and N2 second surrounding pixels with a second value; estimating a left ventricular volume corresponding to the heart ultrasound image based on the reference point pixels.

Abstract: A method for evaluating a movement state of a heart is provided and includes: obtaining a plurality of consecutive heart ultrasound images corresponding to a heart and accordingly estimating a plurality of left ventricular volumes corresponding to the heart ultrasound images; finding a plurality of specific extremums in the left ventricular volumes and accordingly estimating a plurality of time differences among the specific extremums; estimating a statistical characteristic value of the time differences based on the time differences; and determining that an abnormal movement state of the heart occurs in response to determining that at least one of the time differences deviates from the statistical characteristic value up to a predetermined threshold.

Abstract: An image processing apparatus for evaluating cardiac images and a ventricular status identification method are provided. In the method, a region of interest (ROI) is determined from multiple target images, a variation in grayscale values of multiple pixels in the ROIs of each target image is determined, and one or more representative images are obtained according to the variation in the grayscale values. The target image is related to the pixels within an endocardial contour of a left ventricle. A boundary of the ROI is approximately located at two sides of a bottom of the endocardial contour. The ROI corresponds to a mitral valve. The variation in the grayscale values is related to a motion of the mitral valve. The representative image is for evaluating a status of the left ventricle.

Abstract: An image processing apparatus for evaluating cardiac images and a ventricular status identification method are provided. In the method, a region of interest (ROI) is determined from multiple target images, a variation in grayscale values of multiple pixels in the ROIs of each target image is determined, and one or more representative images are obtained according to the variation in the grayscale values. The target image is related to the pixels within an endocardial contour of a left ventricle. A boundary of the ROI is approximately located at two sides of a bottom of the endocardial contour. The ROI corresponds to a mitral valve. The variation in the grayscale values is related to a motion of the mitral valve. The representative image is for evaluating a status of the left ventricle.

Abstract: The disclosure provides an eye state assessment method and an electronic device. The method includes: obtaining an optic disc image area from a first fundus photography and generating multiple optic cup-to-disc ratio assessment results by multiple first models based on the optic disc image area; obtaining a first assessment result of an eye based on the optic cup-to-disc ratio assessment results; performing multiple data augmentation operations on the first fundus photography to generate multiple second fundus photographies; generating multiple retinal nerve fiber layer (RNFL) defect assessment results by multiple second models based on the second fundus photographies; obtaining a second assessment result of the eye based on the RNFL defect assessment results; and obtaining an optic nerve assessment result of the eye based on the first assessment result and the second assessment result.

Abstract: A method and a system for establishing a light source information prediction model are provided. A plurality of training images are captured for a target object. A white object is attached on the target object. True light source information of the training images is obtained according to a color of the white object in each of the training images. A neural network model is trained according to the training images and the true light source information, and a plurality of pieces of predicted light source information is generated according to the neural network model during the training. A learning rate for training the neural network model is adaptively adjusted based on the predicted light source information.

Abstract: A method and a system for establishing a light source information prediction model are provided. A plurality of training images are captured for a target object. A white object is attached on the target object. True light source information of the training images is obtained according to a color of the white object in each of the training images. A neural network model is trained according to the training images and the true light source information, and a plurality of pieces of predicted light source information is generated according to the neural network model during the training. A learning rate for training the neural network model is adaptively adjusted based on the predicted light source information.

Abstract: A method for evaluating a movement state of a heart is provided and includes: obtaining a plurality of consecutive heart ultrasound images corresponding to a heart and accordingly estimating a plurality of left ventricular volumes corresponding to the heart ultrasound images; finding a plurality of specific extremums in the left ventricular volumes and accordingly estimating a plurality of time differences among the specific extremums; estimating a statistical characteristic value of the time differences based on the time differences; and determining that an abnormal movement state of the heart occurs in response to determining that at least one of the time differences deviates from the statistical characteristic value up to a predetermined threshold.

Abstract: A method for estimating a ventricular volume is provided and includes: obtaining a left ventricular mask image corresponding to a heart ultrasound image, where the left ventricular mask image is a binary image; finding 3 reference point pixels in the left ventricular mask image, where each of the reference point pixels has a first value, each of the reference point pixels is surrounded by N surrounding pixels, and the surrounding pixels of each of the reference point pixels include N1 first surrounding pixels with the first value and N2 second surrounding pixels with a second value; estimating a left ventricular volume corresponding to the heart ultrasound image based on the reference point pixels.

Abstract: An electronic device and a method for determining the intensity of a low-frequency current are provided. The method includes: individually applying a corresponding first current to a body part of a user in N consecutive time intervals, wherein the time intervals include an i-th time interval to an (i+N)-th time interval; obtaining electromyography values of the body part in each time interval; determining a second current corresponding to an (i+N+1)-th time interval based on the first current corresponding to each time interval, the body part, personal information of the user, and the electromyography values of each time interval; and applying a second current to the body part of the user in the (i+N+1)-th time interval.

Abstract: An image pre-processing method and an image processing apparatus for a fundoscopic image are provided. A region of interest (ROI) is obtained from a fundoscopic image to generate a first image. The ROI is focused on an eyeball in the fundoscopic image. A smoothing process is performed on the first image to generate a second image. A value difference between neighboring pixels in the second image is increased to generate a third image.

Abstract: An image processing apparatus for evaluating cardiac images and a ventricular status identification method are provided. In the method, a region of interest (ROI) is determined from multiple target images, a variation in grayscale values of multiple pixels in the ROIs of each target image is determined, and one or more representative images are obtained according to the variation in the grayscale values. The target image is related to the pixels within an endocardial contour of a left ventricle. A boundary of the ROI is approximately located at two sides of a bottom of the endocardial contour. The ROI corresponds to a mitral valve. The variation in the grayscale values is related to a motion of the mitral valve. The representative image is for evaluating a status of the left ventricle.

Abstract: A method and a system for establishing a light source information prediction model are provided. A plurality of training images are captured for a target object. A white object is attached on the target object. True light source information of the training images is obtained according to a color of the white object in each of the training images. A neural network model is trained according to the training images and the true light source information, and a plurality of pieces of predicted light source information is generated according to the neural network model during the training. A learning rate for training the neural network model is adaptively adjusted based on the predicted light source information.