19 results about "Intensity normalization" patented technology

The invention relates to a medical image segmentation method based on wavelet boundary enhancement and multi-scale perception, and the method comprises the steps: firstly carrying out the preprocessing of an input medical image, including size standardization, intensity normalization and data enhancement; then, inputting the processed image into a deep fusion segmentation network, extracting high-frequency boundary features through wavelet transform and generating a boundary attention map, and capturing global context information in combination with a multi-scale dynamic sparse attention mechanism; and finally, fusing the multi-scale features through a boundary enhancement up-sampling module in a decoder stage, and optimizing a segmentation result by adopting multi-scale supervision and a mixed loss function. According to the method, the boundary precision and the detail retention capability of medical image segmentation are effectively improved, and the segmentation performance under a fuzzy boundary, a multi-scale structure and a complex background is remarkably improved.

The invention discloses a brain tumor MRI image segmentation method based on a soft clustering KAN network, and belongs to the technical field of medical information processing, and the method comprises the steps: carrying out the voxel alignment and intensity normalization of a multi-modal three-dimensional brain tumor MRI image, and constructing a unified input sample; layer-by-layer downsampling and multi-scale feature extraction are realized through a KAN residual encoder formed by alternately cascading double-layer KAN attention subnets and three-dimensional maximum pooling layers; three-dimensional position coding is added to the highest-layer features, and bottleneck features containing global structure priori are obtained through soft K-means clustering bottleneck layer modeling; fusing the jump connection feature and the up-sampling feature by using an attention gating mechanism to complete decoding; and through soft clustering regularization and a multi-scale depth supervision constraint optimization model, finally through fusion of a main segmentation branch and a refined branch, outputting a multi-subarea three-dimensional segmentation result of the whole tumor, the tumor core and the enhanced tumor.

The invention provides a motor defect acoustic feature enhancement and visualization method and system based on normal sample spectrum normalization, and a storage medium. The method comprises the following steps: step 1, establishing a standard qualified sample database and calculating a normalization vector; the method comprises the following steps: collecting audio signals of qualified motors of the same model when running under a standard working condition, and performing statistical analysis after Fourier transform to obtain a frequency domain intensity normalized vector; 2, processing a to-be-measured motor signal and generating a normalized time-frequency diagram; the method comprises the following steps: acquiring an audio signal of a to-be-detected motor with the same model during operation under a standard working condition, carrying out Fourier transform processing to obtain an original time-frequency diagram, and then carrying out spectrum normalization processing to generate a one-dimensional frequency domain intensity normalization vector; and step 3, generating a visual time-frequency grey-scale map and a color map. The method has the beneficial effects that the convenience and accuracy of labeling work are improved, better-quality input data are provided for a subsequent AI model, and the development efficiency and universality of a detection scheme are remarkably improved.

The invention discloses a radiology report generation method based on hierarchical interactive fusion, and relates to the technical field of medical image intelligent processing, and the method comprises the steps: collecting original image data, carrying out the intensity normalization, obtaining normalized image data, carrying out the multi-level visual feature extraction and feature dimension unification of the normalized image data, and carrying out the multi-level visual feature extraction and feature dimension unification of the normalized image data; forming a hierarchical visual feature set; coding processing is carried out on the hierarchical visual feature set, cross-layer attention relations between a shallow layer and a deep layer and between a middle layer and the deep layer are constructed in the coding process, a shallow layer two-dimensional biased field and a middle layer two-dimensional biased field are formed, and a migration smoothness index and a consistency index are obtained after nonlinear resampling; and in the decoding stage, a multi-path cross attention structure is constructed based on the coded hierarchical visual feature set, an abnormal priori graph is constructed according to the shallow two-dimensional biased field and the middle two-dimensional biased field, attention bias is formed, and a cross-hierarchical cross attention aggregation vector is generated. According to the invention, association expression of multi-level visual features is realized.

This invention discloses a method and system for following up on magnetic resonance imaging (MRI) images of amyloid-related radiological abnormalities, belonging to the field of medical image processing. The method includes: acquiring multimodal MRI image data; performing bias field correction and intensity normalization based on a bias field fitting model to generate a first image set; performing global alignment and edge detection on the first image set to extract anatomical physical boundaries, generating brain region mask data; performing branch deformation field registration on the first image set within the brain parenchyma defined by the brain region mask data to generate a spatially aligned second image set; calculating the average intensity of normal brain parenchyma regions in the second image set to obtain a scaling factor, acquiring a three-dimensional residual image; performing anisotropic diffusion filtering denoising on the three-dimensional residual image, and performing differential denoising and morphological screening processing, outputting the detection results. This invention eliminates systematic artifacts caused by insufficient bias field correction.

The invention discloses a small sample abdomen multi-organ image segmentation method based on a prototype network and cross attention. The method comprises the following steps: firstly, preprocessing an abdominal computed tomography (CT) image, and completing resampling and intensity normalization; dividing the preprocessed image into a support set and a query set, and constructing a small sample segmentation task; a support image and a query image are input into a deep learning network model, a cross attention module is introduced into a multi-layer structure of an encoder to explicitly model foreground, background and boundary regions, interaction and fusion among different level features are enhanced, and pixel-by-pixel matching and distinguishing of support prototype and query features are realized in combination with a double-branch contrast learning structure. According to the method, a cross attention mechanism is introduced into multiple layers of the encoder, the correlation and boundary expression ability between the features are effectively enhanced, the discrimination and robustness of the model under the small sample condition are further improved through a double-branch contrast learning structure, and therefore under the condition that labeling data is limited, the accuracy and robustness of the model are improved. And efficient and automatic segmentation of multiple organs of the abdomen is realized.

The invention discloses a method for detecting mouse brain nucleus activation based on manganese-enhanced magnetic resonance imaging, and belongs to the field of image processing, and the method comprises the steps: converting an acquired mouse head manganese-enhanced magnetic resonance image into an NIFTI format, and enabling the direction and voxel size of the image to be consistent with a standard mouse brain map template; performing offset field correction on the image; loading a PLKA-nnUNet model, carrying out image segmentation on the image, outputting a binary brain mask, and extracting an individual mouse brain image from the image after bias field correction; carrying out image registration and intensity normalization; for each registered individual mouse brain image, calculating relaxation rate mean values of four hippocampal subregions of the mouse brain R1 image, wherein the relaxation rate mean values are used for quantitatively comparing mouse brain activation conditions; and carrying out voxel-level statistical test on the registered and normalized individual mouse brain images to identify brain regions with intensity differences among different experimental conditions. According to the method, the dependence on professional operation is reduced.

This invention provides a three-dimensional quantitative analysis and visualization method and system for aortic diameter expansion velocity based on longitudinal imaging data, applicable to the field of data processing. This application uses baseline and follow-up longitudinal aortic CTA images as the data source. Preprocessing includes spatial resampling and intensity normalization. The aortic lumen is extracted through medical image segmentation, and a voxel mask is generated. Then, the isosurface is reconstructed using the traveling cubes algorithm, and two sets of triangular surface mesh models are obtained through smoothing, cropping, repair, and remeshing. Using the baseline mesh as the source and the follow-up mesh as the target, topologically identical geometric matching is achieved through non-rigid registration. The centerline is extracted, and the point-by-point diameter is calculated. Combined with the follow-up time, three-dimensional quantitative data of the global diameter expansion velocity is obtained. Finally, multi-morphological visualization of expansion velocity and risk stratification is achieved through color mapping, 3D printing, and two-dimensional planar unfolding, constructing a complete three-dimensional quantitative analysis system for the aorta.

A system for automated tumor margin determination in radiological images, comprising: an image acquisition interface configured to receive radiological image data from one or more imaging modalities; a preprocessing unit operationally coupled to the image acquisition interface and configured to perform intensity normalization, spatial resampling, and noise reduction on the received image data to generate standardized image inputs; a feature extraction unit comprising a plurality of convolutional layers arranged in a hierarchical structure and configured to extract spatial features of different orders of magnitude from the standardized image inputs;a transformer coding unit operationally coupled to the feature extraction unit and configured to generate context-sensitive feature representations by applying self-attention operations over spatial areas of the extracted features; a fusion unit operationally coupled to both the feature extraction unit and the transformer coding unit and configured to combine features derived from convolutions and context representations derived from transformers into a unified feature map; a decoding unit operationally coupled to the fusion unit and configured to generate a segmentation map corresponding to the tumor regions by incrementally increasing and reconstructing the spatial resolution; a boundary refinement unit configured to improve the delineation of tumor margins in the segmentation map;a processing unit that is operationally coupled with the preprocessing unit, the feature extraction unit, the transformer coding unit, the fusion unit, the decoding unit, and the boundary refinement unit, wherein the processing unit executes instructions stored in a memory unit to perform automated tumor boundary delineation; and a display interface configured to overlay the delineated tumor boundaries onto the radiological image data.

The application discloses a focused ultrasound array acoustic modeling method and system based on MRI single-mode image, and the method comprises the following steps: acquiring head MRI data and corresponding CT data of the same object; converting the CT data into a three-dimensional density matrix and a three-dimensional sound velocity matrix as an acoustic physical benchmark label; performing spatial resampling alignment, intensity normalization and three-dimensional block preprocessing on the MRI data and the benchmark label to construct a training data set; based on the training data set, an end-to-end conditional generation model, i.e., a teacher model, is constructed, and a physical constraint loss is introduced for iterative training; the teacher model parameters are frozen, a forward reconstruction network, i.e., a student model, is constructed based on the training data set, and a single-step generation distillation training is performed by using a distillation technology; the teacher model is removed, and a three-dimensional sound velocity map and a density map are obtained by using single-step inference of the student model based on the single-mode MRI image. The application realizes end-to-end mapping from an MRI image space to a high-dimensional physical attribute space.

The invention relates to the technical field of signal processing, in particular to a head side image joint enhancement algorithm based on a wavelet decomposition Unet model, which comprises the following steps: S1, converting the Raw format of an image data file to be denoised into a 32-bit floating point data format; and S2, carrying out size unification and intensity normalization processing on the input image, dividing the input image by 65535 to enable the input image to be consistent in spatial size and gray level distribution, and providing standardized feature representation for model input. Rich feature representation is provided for the U-net model, the improved U-net structure can give consideration to a global structure and local details at the same time through coding-decoding and a cross-layer jump connection mechanism, the definition of joint edges and fine structures is improved, feature information of different scales can be effectively fused by using a sampling self-attention mechanism, and the method has the advantages of being high in robustness and high in robustness. And the attention of the model to the joint region is enhanced, so that the balance of global and local features is realized, and the problems of detail loss and excessive smoothness are avoided.

The application discloses a method for detecting mouse brain nucleus activation based on manganese-enhanced magnetic resonance imaging, and belongs to the field of image processing, which comprises the following steps: converting the manganese-enhanced magnetic resonance image of the mouse head collected into NIFTI format, keeping the direction and voxel size of the image consistent with the standard mouse brain atlas template; performing bias field correction on the image; loading the PLKA- nnUNet model, performing image segmentation on the image, outputting a binary brain mask, and extracting individual mouse brain images from the image after bias field correction; performing image registration and intensity normalization; calculating the mean relaxation rate of the four hippocampal subregions of the mouse brain R1 map for each individual mouse brain image after registration, which is used for quantitatively comparing the activation of the mouse brain; and performing statistical testing at the voxel level on the registered and normalized individual mouse brain images to identify the brain regions with intensity differences between different experimental conditions. The method reduces the dependence on professional operation.

The invention discloses a forest fire smoke three-dimensional reproduction and kinetic analysis method and system based on a multi-elevation WER slice and a storage medium, the method does not depend on vertical profile data, and the method comprises the following steps: obtaining multi-elevation reflectivity slice data collected by a multi-band weather radar, performing clutter suppression, biological echo suppression and X-band attenuation elimination processing, and obtaining a multi-elevation WER slice; cross-band intensity normalization is completed; performing back projection on the preprocessed multi-elevation angle reflectivity slice data to a unified three-dimensional voxel grid according to a radar observation geometrical relationship, and performing weighted fusion on back projection data based on a preset wave band weight and an elevation angle weight to generate a smoke three-dimensional reflectivity field; based on a preset reflectivity threshold value, a smoke contour surface is extracted from the three-dimensional reflectivity field, and the smoke column top height, the maximum reflectivity height and the maximum sectional area of smoke are calculated; and outputting a smoke three-dimensional structure reproduction result and a kinetic parameter time sequence containing the smoke column top height, the maximum reflectivity height and the maximum sectional area.

The invention relates to an A beta-PET brain template construction and quantitative analysis method based on deep learning, and the method comprises the steps: obtaining a PET image preprocessed by a normal participant, namely, an A beta-PET image; performing spatial registration to MNI152, and performing intensity normalization processing, data screening and average calculation to obtain a preliminary A beta-PET brain template image; the screened A beta-PET image is registered to a preliminary A beta-PET brain template image, and a final A beta-PET brain template image is obtained through intensity normalization processing, data screening, average calculation and smoothing processing; calculating a first quantitative parameter to construct a brain amyloid protein deposition data database; and performing spatial registration on the A beta-PET image to be analyzed to the final A beta-PET brain template image, dividing a brain region, calculating a second quantitative parameter, and comparing the second quantitative parameter with the brain amyloid protein deposition data database to complete quantitative analysis. The method has the beneficial effects that a normal A beta-PET brain template data database is established, a matched quantitative amyloid protein deposition analysis process is provided for clinical use, and the registration effect and the diagnosis and treatment efficiency are improved.

This invention relates to the field of intelligent medical image analysis technology, and in particular to a deep learning-based method and system for stroke image segmentation. The method includes: acquiring diffusion-weighted, apparent diffusion coefficient, and fluid attenuation inversion recovery images to form a dataset; obtaining standardized images and quality grades through resampling and intensity normalization; training a prior model with normal samples to generate a potential distribution centroid library and quality calibration curve; alternately updating the deformation field and anomaly editing field of the images to be segmented to generate candidate probability maps; zeroing the anomaly editing field to generate counterfactual healthy images; and combining differential evidence and contralateral mirror evidence to output lesion masks and suspicious region masks, thereby improving cross-quality stability and verifiability.

The invention provides a medical image segmentation method and system for dynamically optimizing attention weight based on reinforcement learning. The method comprises the following steps: carrying out space and intensity normalization, low-quality sample screening and task-oriented enhancement on a multi-center medical image; constructing a segmentation network containing multi-head self-attention, and performing learnable weight weighting on each head output; taking a Dice and boundary IoU fusion index as an instant reward, and dynamically optimizing the weight by adopting a REINFORCE strategy gradient; performing end-to-end training on the model through a joint function lost by cross entropy, Dice and reinforcement learning strategies; in the reasoning stage, a high-precision segmentation result can be obtained by inputting a to-be-segmented image. The system comprises a preprocessing module, a segmentation network module, a reinforcement learning update module, a memory module and a processor module. According to the method, the segmentation precision and the boundary consistency are improved on the premise that the parameter quantity and the reasoning time consumption are not remarkably increased.

The application provides a quantitative analysis method based on correlation between EDS element distribution and atomic strain field, which comprises the following steps: 1) obtaining a spherical aberration corrected transmission electron microscope image and a corresponding EDS element distribution energy spectrum image, and performing spatial registration on the images; 2) performing noise suppression and filtering processing on the EDS element distribution energy spectrum image respectively, and through a preset calibration relationship, normalizing the element intensity to obtain a normalized EDS signal image; 3) converting each pixel site of the normalized EDS signal image into a relative volume fraction distribution image; 4) obtaining atomic strain field images of body expansion strain components, shear strain components and lattice rotation components; 5) generating a correlation coefficient image; and 6) outputting the correlation result between the element distribution and the atomic strain field. The application can quantitatively reveal the coupling relationship between the element distribution and the atomic scale deformation mode, and has the advantages of strong non-concentration error suppression capacity, good result repeatability, wide application range and the like.

The invention discloses a camera quality detection method and device based on image detection and a medium, and relates to the technical field of intelligent optical detection, and the method comprises the steps: collecting multiple frames of images and metamerism color block region images, and obtaining a to-be-processed image sequence; performing geometric alignment and intensity normalization on the to-be-processed image sequence by using the reference region to obtain standardized image data; based on the standardized image data, obtaining a potential reference image and a mechanism parameter vector through joint inversion, and extracting a quality parameter; calculating a metamerism separation index based on a metamerism color block area in the standardized image data, correcting a quality parameter by using the index, and obtaining a color parameter and the corrected metamerism separation index; and obtaining a detection result based on comparison between the color parameter and the corrected metamerism separation index and a threshold value. According to the method, the metamerism separation indexes are calculated and are subjected to linkage correction with the color parameters, so that the color consistency of the camera is dynamically corrected.