199 results about "Mr images" patented technology

The invention discloses a renal clear cell carcinoma prognosis prediction method based on multi-mode MRI (Magnetic Resonance Imaging) image and digital pathological omics fusion. The method comprises the following steps: S1, collecting a training data set based on an MR image and a pathological image; s2, feature extraction of MR radiomics; s3, deep learning feature extraction of the pathological image; s4, an MR-pathological feature fusion module; and S5, deploying the network. According to the method, depth features with prognosis information are obtained from two scales of pre-treatment images and post-operation pathology, effective features are extracted by adopting image omics and a convolutional neural network mode according to data characteristics of MR images and pathology images, and depth fusion of the two types of features is completed in a hidden space through a multi-task guiding mode, so that the accuracy of the MR image and the pathology image is improved. A precise prognosis model with multi-scale information is provided, and the method has a relatively strong clinical application prospect and is of great significance for realizing precise immunotherapy and improving prognosis of a patient.

The invention relates to a method of MR imaging of an object positioned in the examination volume of an MR system (1). It is an object of the invention to provide a deep learning-based denoising approach that overcomes the Rician bias problem. As a solution, the invention proposes a method comprising the following steps: a) subjecting the object to an imaging sequence comprising RF pulses and switched magnetic field gradients, whereby MR signals are generated, b) acquiring the MR signals, c) reconstructing a complex-valued MR image from the acquired MR signals, d) denoising the MR image using a deep learning algorithm that operates on the real and the imaginary parts of the MR image, and c) computing a magnitude MR image from the denoised complex-valued MR image. According to an aspect of the invention, the deep learning algorithm uses a set of denoising models that are trained using different loss functions. In this way, a trade-off between noise removal and preservation of small image details can be controlled. Moreover, the invention relates to an MR system (1) and to a computer program for an MR system (1).

The present application provides a cardiac fibrosis diagnosis model based on multi-task attentional feature fusion. The cardiac fibrosis diagnosis model is established by the following steps: S01: image collection and labeling: obtaining cardiac magnetic resonance (MR) images as sample data, and performing manual labeling to obtain heart labels corresponding to the MR images; S02: image preprocessing, including normalization processing, data enhancement, and data clipping; S03: model establishment, including establishment of an image recovery network and establishment of an image segmentation and classification network, and executing an image recovery task; S04: model pre-training: training the image recovery network such that the encoder of the image recovery network fully learns the feature of the cardiac fibrosis image; and S05: model training. An objective of the present application is to improve the segmentation precision and diagnosis accuracy of a network model for a cardiac fibrosis image.

The invention provides a multi-mode nuclear magnetic resonance image glioma segmentation method and application thereof, and belongs to the field of medical image processing. The invention provides an M2ES-UNet network aiming at the problems of insufficient utilization of spatial features, single multi-modal fusion mechanism and cross-level semantic loss of an existing segmentation method. According to the method, multi-view anatomical information is extracted through a multi-plane feature collaboration module; utilizing an orthogonal dimension fusion convolution and modal introspection-collaboration module to respectively realize differential fusion of shallow and deep features; and the progressive jump transmission of the features is realized through a coding information smooth transmission module. According to the method, multi-plane and multi-mode complementary information can be effectively mined, the boundary precision and robustness of brain glioma segmentation are remarkably improved, and clinical diagnosis is assisted.

Techniques for generating magnetic resonance (MR) images of a subject from MR data obtained by a magnetic resonance imaging (MRI) system, the techniques including: obtaining input MR data obtained by imaging the subject using the MRI system; generating a plurality of transformed input MR data instances by applying a respective first plurality of transformations to the input MR data; generating a plurality of MR images from the plurality of transformed input MR data instances and the input MR data using a non-linear MR image reconstruction technique; generating an ensembled MR image from the plurality of MR images at least in part by: applying a second plurality of transformations to the plurality of MR images to obtain a plurality of transformed MR images; and combining the plurality of transformed MR images to obtain the ensembled MR image; and outputting the ensembled MR image.

A system and method for synthesis of PET images from multimodal MR images comprising a user (21), input / output device(s) (23) and processing unit (24). The method comprises three stages namely, model training stage, validation stage and model deployment stage. The model training stage comprises the steps of creation of the dataset in the form of pairs of images comprising a single or plurality of multimodal MR images (2) and corresponding single or plurality of PET images (16) and model training using a DCGAN workflow (1) that consists of a generator network (3) and a discriminator network (17). The validation stage is performed to assess similarity between synthetic (15) and real PET images (16) and clinical utility of synthesized PET images. In the model deployment stage, the trained model is loaded and deployed to obtain synthetic PET images (15) from multimodal MR images (2) as input (22).

The invention discloses a pericardium adipose tissue segmentation method, device and equipment based on an MR image and a medium. The method comprises the following steps: acquiring a target heart magnetic resonance image containing pericardium adipose tissue of a target object; segmenting pericardium adipose tissue in the target heart magnetic resonance image based on a tissue segmentation model to obtain target pericardium adipose tissue of the target object; the tissue segmentation model comprises a three-branch cross-domain feature collaborative encoder, a double attention feature fusion module and a dynamic boundary perception decoder, and the three-branch cross-domain feature collaborative encoder performs feature extraction from global, local and frequency domains based on a three-branch structure; the double attention feature fusion module is established based on a space attention mechanism and a parallel channel attention mechanism, and the dynamic boundary perception decoder is used for enhancing the global and boundary perception capability. According to the scheme, the pericardium adipose tissue in the heart MR image can be accurately and efficiently segmented by using the tissue segmentation model based on multi-feature collaboration.

The invention discloses a reference driving undersampling magnetic resonance (MR) image reconstruction method based on denoising regularization (RED) and depth image prior. The method does not depend on a large-scale clinical data set, a constraint reconstruction model based on denoising regularization and depth image prior is constructed, and the reconstruction of the target MR image under the under-sampling data can be realized only through the driving of a high-resolution reference image similar to the structure of the target image to be reconstructed. The method comprises the specific steps of selection of a single magnetic resonance reference image, denoising engine training of reference driving, depth image prior network construction based on U-Net, construction of a constraint optimization model combining k-space data fidelity and RED regularization, sub-problem decomposition and iterative updating based on an ADMM algorithm, and data correction and reconstruction output. According to the method, the structure prior is introduced into the deep network by using the reference image, the learning efficiency is improved, the dependence on the training data is reduced, the high-quality reconstruction of the MR image under the under-sampling data can be realized, the detail structure and the texture feature are effectively reserved, and the reconstruction precision and the visual quality are remarkably improved. The method is suitable for clinical MRI rapid reconstruction and other imaging applications, and has high practical value.

The invention provides a multi-modal weak supervision medical image segmentation method, and belongs to the field of medical image segmentation, and the method comprises the steps: carrying out the feature extraction of CT and MR images, and obtaining CT and MR modal feature representations; fusing same-layer features represented by CT and MR modal features under each scale; performing down-sampling and multi-layer convolution processing on the fusion feature representation; respectively decoding the enhanced feature representation, the CT modal feature representation and the MR modal feature representation to obtain a multi-modal prediction map, a CT prediction map and an MR prediction map, calculating cross entropy loss, respectively calculating multi-view CRF loss of the CT image and the MR image, obtaining intra-modal regular loss, calculating inter-modal consistency loss, adding the cross entropy, the intra-modal regular loss and the inter-modal consistency loss, and when the total loss is minimum, determining that the total loss is minimum. Obtaining a trained image segmentation model; inputting CT and MR images to be segmented into the trained image segmentation model, and outputting segmented images; the invention also provides an image segmentation system. And multi-modal image information is effectively fused and over-fitting is inhibited under a weak supervision condition.

The present invention is applicable to the field of computers and provides a method for registering MR images and CT images, including: obtaining the MR image and the CT image to be registered; generating a simulated CT image according to the MR image; determining a current displacement field according to the simulated CT image and the CT image to be registered; processing the current displacement field and the previous displacement field based on a diffeomorphic algorithm to generate a superposed displacement field; performing regularization processing on the superposed displacement field to generate a smooth deformation field; generating a registration pending confirmation image according to the smooth deformation field and the simulated CT image; until a preset condition is satisfied, confirming the registration pending confirmation image as the registered image, otherwise returning to the step of determining the current displacement field according to the simulated CT image and the CT image to be registered for iterative processing. The registration method provided by the embodiments of the present invention has a good registration effect on images with large deformations by processing the MR image as a simulated CT image and then performing iterative optimization registration.

The invention discloses a pig brain deep brain stimulation DBS electrode planning and reconstruction system and method, and belongs to the technical field of medical image processing and neurosurgery path planning. An automatic registration module based on a magnetic resonance (MR) image and a pig brain standard map is used for accurately identifying a target brain region; the in-target optimal electrode trajectory planning module based on principal component analysis (PCA) and the electric field simulation module based on a Hodgkin-Huxley model are used for optimizing the treatment effect and predicting the tissue activation volume (VTA); and a multi-modal hybrid registration module based on postoperative computed tomography (CT) and preoperative MR images, which is used for accurately reconstructing the actual position of the electrode. According to the invention, through combination of automatic and data-driven planning and accurate postoperative verification, the accuracy, reliability and efficiency of electrode implantation in pig brain DBS research are significantly improved.

Systems and methods for image reconstruction and quantitative MRI. Generative models such as diffusion models are used to reconstruct MR images and generative models and constrained mathematical models fit to estimate quantitative maps from the reconstructed MR images.

The present disclosure relates to systems, methods, and apparatuses associated with reconstructing MR images based on undersampled magnetic resonance (MR) data. The MR data includes 2D or 3D information and may include multiple contrasts and multiple coils. The MR images are reconstructed using deep learning (DL) methods, which can accelerate the scanning and / or image generation process. By reconstructing the MR images based on respective subsets of contrasts, coils, and / or readout segments and then combining the reconstructed MR images to obtain a desired multi-contrast result, the challenges imposed by large amounts of MR data and hardware limitations are overcome.

The application discloses a kind of fast magnetic resonance multi-sequence joint imaging method, system, equipment and medium, it is related to deep learning technical field, including the following steps: acquisition different sequence multiple undersampling MR images, wherein multiple undersampling MR images share the same public feature, each undersampling MR image enjoys respective corresponding specific feature;The mapping relationship between multiple undersampling MR images and latent variable is modeled based on latent variable model, and the optimization objective of reconstruction task in fast magnetic resonance multi-sequence joint imaging is constructed based on mapping relationship;Optimization objective is solved by depth development network.The features of multiple undersampling MR images are subdivided into intermodal shared public features and modality unique specific features, the fine decoupling of the shared and specific features of MR image, solve the problem that intermodal correlation modeling is weak, and redundant information can be effectively avoided or sequence-specific detail information is lost.

The invention relates to a magnetic resonance imaging system and a method for suppressing electromagnetic interference in magnetic resonance data obtained by the magnetic resonance imaging system. Systems and methods for suppressing electromagnetic interference (EMI) in magnetic resonance (MR) data are also provided. These systems and methods include identifying a first subset of MR data that is affected by EMI and suppressing EMI in the first subset to obtain a second subset of MR data. Suppression of EMI in the first subset is performed by applying filtering to the first subset of MR data to suppress contributions of MR spin echo signals in the first subset of MR data, thereby obtaining signal suppressed MR data; suppressing EMI in the signal suppressed MR data to obtain EMI suppressed MR data; and applying the filtered inverse to the EMI suppressed MR data to obtain a second subset of MR data. These systems and methods include generating MR images using the second subset of MR data and outputting the generated MR images.

A first artificial intelligence (Al) engine receives a plurality of incomplete magnetic resonance (MR) K-space data matrices of an object scanned by an MR device. Each of the incomplete MR K-space data matrices comprises complex values and is the result of a corresponding san of the object by the MR device using a sparse-sampled MR scan acquisition sequence. Each sparse- sample MR scan acquisition sequence employs a unique sampling pattern. The first Al engine reconstructs a complete MR K-space data matrix of the scanned object, corresponding to a complete MR K-space acquisition. The reconstruction is based on the data in the plurality of incomplete MR K-space data matrices.

This application provides a method, system, device, and computer-readable storage medium for planning the healing of spinal compression fractures based on multimodal three-dimensional medical images. The method includes: acquiring spinal CT and MR images of the patient; identifying and segmenting the spinal CT and MR images to obtain CT and MR images of the fractured vertebral body at the site of the compression fracture; inputting the CT and MR images of the fractured vertebral body into a preset bone cement anti-leakage planning model to output a bone cement injection anti-leakage planning scheme; wherein the bone cement injection anti-leakage planning scheme includes at least: the viscosity of the bone cement, the injection volume, the injection location, the injection timing, and the injection rate. According to the embodiments of this application, bone cement injection anti-leakage planning can be performed quickly and accurately.

The present invention discloses an organ registration system for an integrated PET / MR based on deep learning, which includes an image acquisition module, an image denoising module, an image reconstruction module, and an organ registration module. The present invention belongs to the technical field of image processing, and specifically refers to an organ registration system for an integrated PET / MR based on deep learning. In this solution, low-dose PET / MR images with different noise intensities are defined, a noise adaptability module is introduced to improve the quality of low-dose PET / MR images, and an adaptive texture matching loss is introduced to construct a strongly adaptable denoising module to ensure the accuracy and efficiency of organ registration; combined with adaptive fusion weights, multi-scale feature fusion is performed at different levels to optimize local details; a hierarchical neural structure search is introduced to improve the registration accuracy of the registration system for different types of PET / MR images; the reconstruction structure is optimized based on the proximity contrast loss to improve the subsequent registration efficiency.

The application belongs to the technical field of image processing, and provides a fusion processing system for spine MR images and CT images, a segmentation model constructed through a Mamba structure, multi-scale feature extraction realized by using a three-way Mamba module, and the segmentation accuracy of the bony structure of the spine, intervertebral discs and soft tissues is significantly improved by combining gated spatial convolution and feature-level uncertainty estimation; at the same time, the GPU memory consumption is effectively reduced by the cascade training flow design, and under the premise of ensuring the segmentation quality, the generation of high-resolution fusion images is realized; on this basis, the fusion network based on a double-input Mamba structure has global perception ability and linear extraction ability of directional features, can efficiently fuse the high spatial resolution bone structure information of CT and the soft tissue contrast advantage of MRI, ensures the consistency of anatomical structures, fully retains the complementary features between modalities, and significantly improves the comprehensive analysis accuracy of spine images.

For autonomous MR scanning of a given medical test, a simplified MR scanner can be used without or with little input or control by a technical expert, e.g., by a physician, radiologist, or person trained in MR scanner operation. The MR scanner autonomously positions, scans, checks quality, analyzes, and / or outputs answers to diagnostic questions with or without MR images. Artificial intelligence-based scan analysis allows continuous or instant changes to the scan configuration to acquire the data desired to answer the diagnostic question. By using a simplified MR scanner, both the positioning of the patient relative to the MR scanner and the localization of the scan by the MR scanner are jointly solved. Sensors can sense the patient at the scan position, with reduced radio frequency requirements allowing a more open bore.

The present invention refers to an apparatus 130 for reconstructing MR images and / or parameter maps based on MR time-domain signals. A providing unit 133 provides time-domain signals acquired in time-domain MRI using a k-space sampling pattern and a magnetic resonance acquisition sequence. A providing unit 132 provides a subspace sampling operator that provides given k-space locations of the sampling pattern time subsamples of basis vectors of a compressed subspace. The compressed subspace can be used to compress the time-domain signals. A processing unit 133 processes the time-domain signals using the subspace sampling operator. A reconstruction unit 134 reconstructs MR images and / or parameter maps based on the processed time-domain signals. This allows for more accurate and less computationally resource-intensive reconstruction of MR images and / or parameter maps in the context of time-domain magnetic resonance imaging.

The present disclosure relates to a lumbar vertebra lesion identification method, device, medium and electronic equipment, wherein the method comprises: determining a lumbar vertebra image of a user, the lumbar vertebra image being a T2 weighted magnetic resonance image; inputting the lumbar vertebra image into a target detection model to obtain a lesion category of the lumbar vertebra of the user, wherein the target detection model comprises a down-sampling layer and a pooling layer, the down-sampling layer is used to extract feature data of the lumbar vertebra image, and the pooling layer is used to classify the feature data to obtain the lumbar vertebra lesion category of the user. Compared with the way of judging the lumbar vertebra lesion by artificial judgment and identifying the lumbar vertebra lesion by machine learning method in the related art, the error of identifying the lumbar vertebra lesion type can be reduced, and thus the accuracy of identifying the lumbar vertebra lesion can be improved.

An artificial intelligence (Al) engine is trained on a plurality of annotated magnetic resonance (MR) images of a patient's brain. A plurality of MR images of a patient's head is provided. For each MR image in the plurality of provided MR images, the Al engine detects a plurality of edges of the brain and determine a biparietal diameter (BP) value, detects a plurality of frontal horn edges, and detects a plurality of occipital horn edges. A correction module determines that at least one detected edge is associated with a non -ventricular body and updates the edge to correspond to the applicable horn. The Al engine determines a frontal horn diameter (F) value, and a occipital horn diameter (O) value. An indication module provides an indication on abnormal dilation of the patient's brain ventricles based on a maximum F value, a maximum O value, and the maximum BP value.

The present invention relates to the field of medical image processing technology, and more specifically to a method, system, and storage medium for vascular tissue segmentation based on MR images. The method comprises the following steps: enhancing vascular information in MR images using a Canny operator to obtain an MR-enhanced image; constructing a first segmentation model and a second segmentation model based on the MR image and the MR-enhanced image, respectively, using a UNet network; guiding the training of the first segmentation model using the second segmentation model, and using the output of the trained first segmentation model as the vascular tissue segmentation result in the MR image. The present invention can guide the segmentation results on low-resolution images to learn from the segmentation results on high-resolution images, thereby obtaining high-precision segmentation results directly on low-resolution images without increasing the data processing process and model complexity of the segmentation process, thereby improving the efficiency of accurate vascular tissue segmentation.

The application provides a method and system for removing electro-physiological signal gradient artifacts in a magnetic resonance environment, and relates to the technical field of magnetic resonance image processing. The method comprises: collecting an electro-physiological signal of a magnetic resonance scan and a scan synchronization signal; aligning and windowing the electro-physiological signal based on the scan synchronization signal to obtain a windowed sequence; constructing an artifact template based on the windowed sequence; combining the windowed sequence and the artifact template to calculate a first residual; performing principal component analysis on the artifact template and the first residual to obtain a second residual; performing logarithmic domain conversion and polarity extraction operations on the second residual to obtain a logarithmic domain sequence and a polarity symbol layer; inputting the logarithmic domain sequence into a deep learning denoising network to output a de-artifact logarithmic domain sequence; and performing inverse logarithmic transformation and polarity recovery operations on the de-artifact logarithmic domain sequence and the polarity symbol layer, respectively, to determine a de-artifact signal.

A system and method automatically detect, in MR images, WM lesions exhibiting a central vein sign. A set of MR images of a brain lesion is acquired, using images of the set as input to different ML algorithms. A first ML algorithm classifies inputted image(s) into first or second classes. The first class includes CVS+ / − and the second class CVSe lesions. A second ML algorithm classifies inputted image(s) into third or fourth classes. The third class includes CVS+ lesions and the fourth central vein sign− lesions. For each set, probability values are used that the set belongs to classes as inputs to a final classifier performing a final classification of the set into second, third, or fourth classes. For each class, the final classifier outputs final probability that the set belongs to the class. The second, third or fourth class with highest probability value is provided through an interface.

The application discloses a CT image synthesis method and device based on MR images, equipment and medium, the application obtains an image synthesis model by inputting a first MR image and a first CT image into a network model for training; a global block embedding module is used for global feature extraction on the first MR image, so that the global feature extraction result is more global; an encoding module is used for local feature extraction on the first MR image, an RSC self-attention module is used for first feature processing according to the global feature extraction result and the local feature extraction result, which can expand the receptive field of the features, a decoding module is used for decoding processing on the first feature processing result to generate a second CT image; a discriminator is used for generating a discrimination result according to the first CT image and the second CT image for training, and the finally obtained target CT image is more detailed and clear, and the application can be widely applied to the field of image processing.

The present application relates to the technical field of internal target area delineation based on 4D magnetic resonance imaging, and discloses a method, apparatus, device, and medium for internal target area delineation based on 4D magnetic resonance imaging. The method comprises: obtaining a first abdominal 4D magnetic resonance image and each second abdominal 4D magnetic resonance image; determining a first probability distribution based on a preset organ reference position and the first abdominal 4D magnetic resonance image, and determining each second probability distribution based on the preset organ reference position and each second abdominal 4D magnetic resonance image; determining a target scan duration based on the first probability distribution and each second probability distribution; performing a 4D magnetic resonance scan on a target object based on the target scan duration, obtaining a target nuclear magnetic resonance image corresponding to the target scan duration, and performing internal target area delineation based on the target nuclear magnetic resonance image. The method can determine an appropriate target scan duration, shorten the 4D magnetic resonance scan duration of the target object, and improve the accuracy of internal target area delineation.

The present invention relates to a medical image segmentation method, comprising: collecting magnetic resonance imaging (MRI) image data of tumor patients as a dataset; performing data processing on the image data in the dataset; using the multimodal images in the dataset that meet the requirements after data processing as model input; designing a separate Transformer for each modality to extract features; designing a modality fusion Transformer to fuse the data of multiple modalities; gradually reshaping encoder outputs of different scales to the input size to obtain a segmentation result that matches the original image; constructing weakly enhanced images and strongly enhanced images for unlabeled data in the dataset; selecting positive and negative examples based on the encoder outputs for the different enhanced images and calculating contrast loss; calculating DICE loss on the labels and segmentation results; training the model to obtain and save the final model. The present invention can better locate tumors and improve segmentation results.