5946 results about "Radiology" patented technology

Provided herein are systems for determining a location of a gross target volume of a patient. In some examples, systems can include one or more processors that are configured to obtain image data associated with a plurality of images of a lesion of a patient. For each image, the one or more processors can be configured to backproject points representing the lesion into the 3D space to determine a plurality of distribution confidence values for a subset of voxels within the three-dimensional space. The one or more processors can be configured to determine a three-dimensional confidence distribution based on confidence values from the plurality of distribution confidence values corresponding to each voxel of the 3D space and determine a position of the lesion within the 3D space based on the 3D confidence distribution.

Optical flow information is determined and used to identify a video clip from among a sequence of frames. The video clip may be identified based on motion features derived in part from the optical flow information. In some embodiments, semantic information is concatenated with the motion features derived from the optical flow information.

This application relates to a method, apparatus, computer device, and storage medium for displaying medical images. In response to an instruction to acquire a first image sequence, a target type label for the first image sequence is determined. From a sequence of images to be identified containing various type labels, at least one second image sequence associated with the target type label is acquired. The first image sequence and at least one second image sequence are then displayed. Each type label is obtained based on a classification result determined according to each image sequence to be identified and a preset classification model. The embodiments of this application employ a preset classification model to automatically identify the image sequences to be identified, obtaining the classification results of each image sequence to be identified, thereby obtaining the type label for the image sequences to be identified. Compared to relying on manually entering tag information for medical images, the method proposed in this application can significantly reduce the error probability of acquiring the second image sequence based on the target type label, thereby greatly improving the accuracy of image sequence acquisition.

In some implementations, a computing device may receive one or more reference images respectively having an indication of a reference object of interest within the one or more reference images. The computing device may generate a probability map associated with a target image based at least in part on matching scores that are associated with the reference object. The computing device may identify an object of interest within the target image based at least in part on the probability map.

This application relates to the field of medical image processing technology, and provides a method, apparatus, device, and storage medium for reconstructing ECT images. This allows for more thorough capture of subtle changes in the concentration distribution of radioactive tracers at high temporal resolution, helping to reduce image noise levels and improve image quality. This application obtains a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device; the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence; using the estimated second ECT image sequence as a priori image, a target second ECT image sequence is reconstructed; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

Systems and techniques are described herein for foveated imaging. For instance, a method for foveated imaging is provided. The method may include determining a first region of interest within a first image based on a gaze of a user relative to the first image; clustering pixels of the first image into a one or more clusters of pixels based on a comparison of one or more characteristics of pixels of the first image; identifying a cluster of pixels from among the one or more clusters of pixels based on an intersection over union of the first region of interest and the cluster of pixels; generating a second region of interest based on the first region of interest and the cluster of pixels; and capturing a second image based on the second region of interest.

The application discloses a cell detection method and system based on deep learning and machine vision, which comprises the following steps: cutting a cervical liquid-based cell pathology digital image into a plurality of sub-images to form a sub-image set; inputting the sub-image set into a normal cell detection model and a lesion cell detection model respectively, and identifying normal cells and lesion cells and their categories; selecting an overlapping area of the normal cells and the lesion cells, and performing secondary classification on the overlapping area through a classification model; performing threshold segmentation on granulocytes in the cervical liquid-based cell pathology digital image to obtain all possible granulocyte targets, filtering to determine final granulocyte targets, and counting and determining the number of the granulocytes. The application solves the problem that the detection model in the prior art has poor detection effect due to the imbalance of normal cell and lesion cell categories, improves the accuracy of cell identification, and realizes accurate detection and identification of cervical liquid-based cells in combination with accurately identified granulocytes.

The application relates to the technical field of medical intelligent diagnosis, and provides a multi-modal ultrasonic scanning method, device and related system. The method collects depth image data of a skin area of a user through a camera module, controls a mechanical arm to adjust an ultrasonic probe to a first target pose according to scanning intention information of the user and the depth image data, wherein the ultrasonic probe and a force feedback sensor are integrated at the tail end of the mechanical arm, calculates an ultrasonic quality evaluation parameter according to first ultrasonic image data of the ultrasonic probe and first contact pressure data of the force feedback sensor, obtains a first ultrasonic quality evaluation parameter, adjusts the ultrasonic probe to a second target pose through the mechanical arm according to the first ultrasonic quality evaluation parameter, and scans the user according to the depth image data, second ultrasonic image data of the ultrasonic probe and second contact pressure data of the force feedback sensor to obtain a scanning result. In this way, the stability and safety of the ultrasonic probe during scanning are improved.

The application provides a nail contour extraction method and system. The extraction method based on the system comprises the following steps: a data acquisition step: 3D scanning to obtain finger 3D data of a target finger, the finger 3D data containing complete nail 3D data; a data point cloud step: converting the finger 3D data into 3D point cloud data to obtain finger point cloud, the finger point cloud containing nail curved surface point cloud; a nail point cloud acquisition step: segmenting the finger point cloud to obtain the nail curved surface point cloud; a nail point cloud flattening step: flattening the nail curved surface point cloud with equal surface area to obtain nail plane point cloud in a plane state; and a nail contour drawing step: drawing a nail contour according to the contour of the nail plane point cloud. The nail contour extraction method and system of the application obtain nail plane point cloud formed by flattening nail curved surface point cloud of a nail with an arc, and the shape of the peripheral contour point cloud of the flattened nail plane point cloud is more consistent with the peripheral contour of the nail with the arc.

1. Name of the product in this design: Disposable steep pulse therapy ablation needle (three needles). 2. Purpose of this design: A disposable steep pulse therapy ablation needle (three needles) for the ablation treatment of tumors. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: a 3D model.

Systems and methods are provided for processing X-ray images, wherein the methods are implemented as a software program product executable on a processing unit of the systems. Generally, an X-ray image is received by the system, the X-ray image being a projection image of a first object and a second object. The first and second objects are classified, and a respective 3D model of the objects is received. At the first object, a geometrical aspect like an axis or a line is determined, and at the second object, another geometrical aspect like a point is determined. Finally, a spatial relation between the first object and the second object is determined based on a 3D model of the first object, a 3D model of the second object, and the information that the point of the second object is located on the geometrical aspect of the first object.

This invention relates to the field of image data processing technology and discloses a feature fusion method for multi-target detection and classification of text images. The method includes: acquiring a low-level spatial detail feature tensor and a high-level global semantic feature tensor and aligning the spatial axis resolution to form a feature tensor pair; calculating the channel dot product of the feature tensor pair to obtain a spatial feature modulation matrix; using the element values ​​of the spatial feature modulation matrix as feature flow control weights, combining the feature tensor pairs through a nonlinear gated mapping, and outputting the final multi-target detection and classification fusion feature. This invention establishes a gated constraint boundary through the spatial feature modulation matrix, suppresses high-level feature blurring, strengthens pixel contour edges, eliminates text target feature diffusion and crosstalk, reshapes semantic causal correspondence, and enhances detection recall and classification accuracy.

A knowledge distillation method for fracture detection includes obtaining medical images in chest X-rays, including region-level labeled images, image-level diagnosis positive images, and image-level diagnosis negative images; performing a supervised pre-training process on the region-level labeled images and the image-level diagnosis negative images to train a neural network to generate pre-training weights; and performing a semi-supervised training process on the image-level diagnosis positive images using the pre-training weights. A teacher model is employed to generate pseudo ground truth (GT) on the image-level diagnosis positive images to supervise the training of a student model, the pseudo GT is processed by an adaptive asymmetric label sharpening (AALS) operator to generate sharpened pseudo GT to provide positive detection responses on the image-level diagnosis positive images.

An image processing method includes: in response to a target trigger operation, performing recognition processing on a target image, where the target image is an image input to a target application or an image currently displayed by the target application, and the target application is an application capable of performing the recognition processing or calling a target program file to perform the recognition processing; and outputting a recognition result for the target image, where the recognition result is capable of indicating source information of at least a portion of an image content of the target image.

To provide an abnormality diagnostic device that can appropriately diagnose the nature of abnormalities occurring in fuel cell systems. [Solution] The abnormality diagnosis device determines that an abnormality has occurred in the power conversion circuit if the first determination condition is met, which is that the output current of the cell stack is in a predetermined high-current state, which is greater than the assumed current value assumed from the power output from the power conversion circuit to the power line, or that output suppression operation is being performed, and the second determination condition is met, which is that the output of the cooling fan is in a predetermined high-output state, which is greater than the standard output range, and the third determination condition is not met, which is that the output voltage of the cell stack is in a predetermined low-voltage state.

Disclosed is an image data encoding / decoding method and apparatus. A method for decoding a 360-degree image comprises the steps of: receiving a bitstream obtained by encoding a 360-degree image; generating a prediction image by making reference to syntax information obtained from the received bitstream; combining the generated prediction image with a residual image obtained by dequantizing and inverse-transforming the bitstream, so as to obtain a decoded image; and reconstructing the decoded image into a 360-degree image according to a projection format. Here, the step of generating the prediction image comprises the steps of: obtaining, from motion information included in the syntax information, a motion vector candidate group including a motion vector of a block adjacent to a current block to be decoded; deriving a prediction motion vector from the motion vector candidate group, on the basis of selection information extracted from the motion information; and determining a prediction block for the current block to be decoded, using a fmal motion vector derived by adding the prediction motion vector to a differential motion vector extracted from the motion information.

A method, apparatus, and computer-readable medium for foreground object deletion and inpainting, including storing contextual information corresponding to an image of a scene, identifying one or more foreground objects in the scene based at least in part on the contextual information, each foreground object having a corresponding object mask, identifying at least one foreground object in the one or more foreground objects for removal from the image, generating a removal mask corresponding to the at least one foreground object based at least in part on at least one object mask corresponding to the at least one foreground object, determining an estimated geometry of the scene behind the at least one foreground object based at least in part on the contextual information, and inpainting pixels corresponding to the removal mask with a replacement texture omitting the foreground object based at least in part on the estimated geometry of the scene.

This invention relates to the field of image processing technology, and more particularly to an image processing method, apparatus, computing device, and computer-readable storage medium. The method includes: processing a first image using a convolutional neural network to obtain a global semantic information map of the first image; based on the global semantic information map, enlarging the grayscale image of the first image to obtain a first reconstructed image; based on the global semantic information map, performing color enhancement processing on the chroma image of the first image to obtain a second reconstructed image; and fusing the first reconstructed image and the second reconstructed image to obtain a second image. This allows the second image to be reconstructed synchronously with the first image in terms of super-resolution and color enhancement, resulting in a higher degree of consistency between the first and second reconstructed images, thus improving the quality of the second image. Simultaneously, reusing global semantic information also saves computational resources.

A treatment system includes a treatment tool, an imaging apparatus, and a control device. The treatment tool is configured to perform treatment on a living tissue by applying treatment energy from an end effector to the living tissue based on supplied electric power. The imaging apparatus captures an image of the living tissue in a state in which the treatment energy is applied from the end effector to the living tissue. The control device includes a processor to control the imaging apparatus by acquiring the captured image, determining whether an abnormality has occurred in the end effector based on the acquired captured image, and executing an instruction to stop supply of the electric power to the treatment tool when determining that the abnormality has occurred in the end effector.

This invention relates to a visual sequence compression method for multi-turn visual question answering in medical images, comprising: extracting an original visual token sequence based on a target medical image; generating a compression matrix using a pre-trained meta-compression module based on the image features of the original visual token sequence; applying the compression matrix to the original visual token sequence to obtain the compressed visual token sequence; concatenating the compressed visual token sequence with the language token sequence of the current question-answering turn, and inputting the concatenation into a large-scale visual language model to generate the question-answering response for the current turn. The beneficial effects of this invention are: by dynamically generating a compression matrix based on the features of the medical image itself, this invention can compress the length of the visual sequence while still retaining key image features that are low-attention but related to lesion identification, thereby overcoming the shortcomings of existing technologies that easily overlook small lesions and weak-contrast abnormal areas.

This invention provides a method and apparatus for correcting projected images. The method includes: a. acquiring multiple first projected images of a scanned object; b. acquiring multiple reference images, wherein each reference image corresponds to at least one of the multiple first projected images; and c. correcting the first projected images based on the grayscale relationship between the first projected images and their corresponding reference images. The method and apparatus for correcting projected images of this invention are characterized by low cost and high correction accuracy.

This application discloses a training method, apparatus, electronic device, and storage medium for a hair recognition model. The method includes: acquiring a first historical hair image for training, and adjusting the image parameters of the first historical hair image to obtain a second historical hair image; training a constructed feature extraction model using the first and second historical hair images, and extracting features from each hair image in the first historical hair image using the trained feature extraction model to obtain a first image feature, and extracting features from each hair image in the second historical hair image to obtain a second image feature; loading a constructed category recognition model, and training the category recognition model using training data to obtain a trained category recognition model; and fine-tuning the feature extraction model and category recognition model using a third historical hair image obtained from the training data to obtain a hair recognition model. This improves the accuracy of hair recognition.

The application provides a method for craniomaxillofacial CT image segmentation, a model training method and equipment, wherein the method for craniomaxillofacial CT image segmentation model training comprises the following steps: acquiring a craniomaxillofacial tumor CT image dataset, wherein the craniomaxillofacial tumor CT image dataset is composed of a labeled dataset and an unlabeled dataset; constructing a main network and a teacher network based on the same image segmentation deep neural network backbone model; the main network learns using the labeled craniomaxillofacial tumor CT image dataset in each training iteration; the teacher network learns using the labeled dataset in a preset starting stage and ending stage, simulates the learning process of the main network by averaging all weights of the current main network in the middle stage, and constrains the graphic data in the unlabeled dataset through segmentation consistency of different perspectives.

An image difference generator can receive first and second images of first and second display outputs of a computer program, such as a webpage. The images can contain two different categories of differences, one related to the differences in the webpage code, and one related to the differences unrelated to the webpage code, such as differences due to environmental, platform, or generally indeterministic sources, unrelated to the webpage code. The image difference generator can identify each category of the differences and their corresponding regions in each image, and highlight them differently in an output difference image. A developer of the software or webpage can inspect the output difference image and focus her development activity on the appropriate category.

The application relates to a method for quickly fixing a target detection frame, which comprises the following steps: performing target detection on an image in a current frame image to obtain target detection frames of all targets; for the case that the number of target detection frames is less than a target threshold, adopting a distance tracking method to track the target detection frames; for the case that the number of target detection frames is greater than the target threshold and the moving speed of the target is less than a speed threshold, adopting an intersection-over-union method to track the target detection frames; for the case that the number of target detection frames is greater than the target threshold and the moving speed of the target is greater than the speed threshold, first adopting a Kalman filter, and then adopting the intersection-over-union method to track the target detection frames; and correcting the size of the target detection frames in the current frame and subsequent frame images. The application can quickly fix the target detection frame according to the number of targets and the moving speed of the target by adopting corresponding methods, and can prevent the size of the same target detection frame from frequently changing, so that the target detection frame is more stable.

The application discloses a motor imagery electroencephalogram signal denoising method and device, medium and product, relates to the technical field of deep learning and biomedical signal processing, and the method comprises the following steps: acquiring a target electroencephalogram signal containing artifacts; inputting the target electroencephalogram signal containing artifacts into a trained electroencephalogram denoising model to obtain a final denoised electroencephalogram signal; wherein the electroencephalogram denoising model comprises an electroencephalogram denoising branch, an artifact prediction branch and an artifact representation interaction attention fusion reconstruction module; the electroencephalogram denoising branch comprises a multi-scale self-adaptive enhancement module, a frequency domain dynamic enhancement module and a feature extraction module. The application solves the problems of traditional electroencephalogram denoising methods in signal fidelity, spectral fidelity, spatial structure preservation and generalization ability.