50 results about "Liver ct" patented technology

Disclosed is a three-dimensional liver CT image automatically segmenting method based on hyper voxels and the graph cut algorithm. The method comprises analyzing a volume data histogram to adaptively enhance the image contrast; performing primary liver contour segmentation layer by layer through an adaptive threshold and an morphological method, selecting the largest liver segment to compute and extract a liver interest area; selecting seed points on the largest liver segment according to a primary liver contour, modelling foreground and background colors through a Gaussian mixed model; generating hyper voxels on the contrast-enhanced liver interest area through an SLIC (simple linear iterative clustering) algorithm, structuring an undirected weighted graph by taking the hyper voxels as the vertex, and segmenting the undirected weighted graph through the graph cut algorithm; performing postprocessing on segmentation results through a morphological algorithm. The three-dimensional liver CT image automatically segmenting method based on the hyper voxels and the graph cut algorithm can achieve rapid and accurate automatic segmentation of the liver in a three-dimensional abdominal CT image.

The invention discloses a liver CT image multi-lesion classification method based on sample generation and transfer learning. The method mainly solves the problem that an existing method is not high in liver multi-lesion detection performance. The implementation scheme is as follows: dividing a data set; respectively constructing a liver organ segmentation network and a liver lesion detection network; based on the deep convolution generative adversarial network, constructing a liver cyst sample generation network and a liver hemangioma sample generation network, and respectively generating newliver cyst and liver hemangioma samples; constructing a liver lesion classification network; subjecting a liver CT image to be detected firstly to organ segmentation by using a liver organ segmentation network, then subjecting a segmentation result to lesion detection by using a liver lesion detection network, and finally classifying detected lesions by using a liver lesion classification network. According to the invention, imbalance of different types of sample sizes is relieved, the lesion classification performance is improved, and the method can be used for positioning and qualifying various lesions such as liver cancer, liver cyst and hepatic hemangioma existing in the liver CT image.

The invention discloses a tumor region segmentation method and system for a liver CT image based on a cascaded full convolutional network. According to the method, automatic segmentation of a liver tumor region is realized by using a full convolutional network model. The method comprises the following steps: performing preprocessing operations such as filtering and sharpening enhancement on a CT image; training a cascade network by using the preprocessed CT data set; achieving liver region segmentation by using two stages of FCN networks and segmenting a tumor region from a liver region of interest. A first-stage FCN network uses a variable pooling method so that more liver features are reserved and liver segmentation precision is improved; and a second-stage FCN uses hole convolution replaces an original convolution layer and a pooling layer, so that the position information of the target can be reserved while a image is reduced and the feeling domain is increased. A large number of features do not need to be manually extracted, and the tumor region can be effectively and accurately segmented from the liver CT image.

The invention provides a liver image segmentation quality evaluation method and device and computer equipment. Inputting the acquired liver CT sequence diagram into an image segmentation model; obtaining a blood vessel mask and a liver mask contained in each liver CT image; carrying out connected domain analysis on each blood vessel mask; utilizing the obtained blood vessel connected domain; veinregions contained in the blood vessel masks are determined; and obtaining and outputting a liver blood vessel segmentation quality evaluation result of the liver CT sequence diagram acquired this timeby utilizing the determined liver mask and vein region. The medical personnel are informed of the liver blood vessel segmentation quality of the liver CT sequence diagram; therefore, it is avoided that under the condition that the liver blood vessel segmentation quality of the liver CT sequence diagram is poor, medical staff continue to refer to the liver CT sequence diagram to diagnose and treata patient, and the auxiliary effect of the liver auxiliary diagnosis system on liver disease diagnosis and surgical treatment is greatly improved.

The invention discloses a three-way-decision-based liver tumor CT image classification method. A classifier is trained; image preprocessing and feature extraction are carried out on a marked liver CT image training sample to form a feature vector; rule extraction is carried out by using an attribute reduction method to form a knowledge base, thereby providing a basis for follow-up case classification diagnoses. T-be-classified cases are processed image pretreatment to extract liver regions, hepatic vessels, and liver tumors; 14 case features are calculated; the case feature values are inputted into a three-way-decision-based classifier; and then cases are classified into three types: a benign tumor, a malignant tumor, and an uncertain tumor. Therefore, a doctor can make different therapeutic regimens for different tumor types.

The invention provides a liver tumor image segmentation method based on a two-way three-dimensional convolutional neural network. The method comprises the following steps: preparing a data set; performing filtering and standardized preprocessing operation on the original CT image in the data set; training a two-way three-dimensional convolutional neural network with a parallel path structure; segmenting a liver tumor in the CT image by using the trained two-way three-dimensional convolutional neural network, and generating a probability graph of a tumor segmentation result; based on the two-way three-dimensional convolutional neural network, tumor segmentation of the liver CT image is fully automatically realized, the accuracy is high, and the speed is high.

The invention discloses a liver CT automatic segmentation method based on deep shape learning, and the method comprises the steps: firstly building a liver segmentation data set, carrying out the preprocessing, and carrying out the coarse segmentation of a liver CT through the liver segmentation; secondly, establishing a liver shape set, learning a liver shape by using a variational auto-encoder, constructing a geometrical shape regularization module, and then adding the geometrical shape regularization module into liver segmentation to obtain a liver segmentation model constrained by geometrical shape consistency for automatic segmentation of liver CT. According to the method, the expressed shape features are creatively added into the existing deep segmentation network through the regularization module, and shape prior information is introduced in the training process of the convolutional neural network, so that the regularity and generalization ability of the segmentation model can be improved, and the segmentation result is enabled to better conform to the medical anatomy characteristics of the standard liver. The method has the advantages of being automatic, high in precision and capable of being migrated and expanded, and automatic and accurate segmentation of the abdominal large organs, such as the liver, can be achieved.

The invention relates to the technical field of computer vision, in particular to a priori shape constraint-based BCA-UNet liver segmentation method, which comprises the following steps of: inputtinga liver CT image, preprocessing the liver CT image to obtain a preprocessed liver CT image, inputting the preprocessed liver CT image into a trained liver segmentation model, and obtaining a BCAUNet liver segmentation model; obtaining liver segmentation results. According to the method, the optimized active contour loss function is adopted to calculate the loss of the high-dimensional features, the features between the two networks are fused to serve as the attention signal of the next layer, the attention signal is used for constraining the segmentation network (BCAUNet, error back propagation is optimized layer by layer, and the loss of edge contours is avoided. Besides, the liver segmentation model is sensitive to the edge contour of the image so that the segmentation precision can be enhanced and the surface distance error can be reduced.

The invention provides a tumor three-dimensional positioning system. The tumor three-dimensional positioning system comprises an image acquisition module, a preprocessing module, an image segmentationmodule, a three-dimensional reconstruction module, a tumor information acquisition module and an information output module. The tumor three-dimensional positioning system sequentially carries out preprocessing and image segmentation on the liver CT image to obtain a gray value matrix of the liver CT image sequence and a gray value matrix of an area where a tumor is located, performs interpolationoperation between layers of the liver two-dimensional CT image by utilizing the gray value information of two adjacent layers according to the difference of gray values of specific coordinates in thetwo layers of images and the gray relationship to obtain a new CT image, converts the gray value and the pixel points to obtain a three-dimensional image of the liver and the tumor, and finally, obtains tumor center coordinates and edge coordinates of a tumor and liver part interface through a tumor information obtaining module, thereby qualitatively and quantitatively presenting a tumor orientation. The tumor three-dimensional positioning method has the advantages of being high in precision and convenient to observe, diagnose, treat and position.

The invention discloses a medical image segmentation method based on interactive foreground extraction and information entropy watershed. The medical image segmentation method comprises the followingsteps: standardizing an original image; eliminating white point noise and edges existing in the image by utilizing morphological opening operation; marking the approximate position of the foreground of the image by using a rectangular frame, and removing a background area in the image; modeling the deterministic foreground and background of the image by using a Gaussian mixture model, creating newpixel distribution, and generating a complete image; finding a threshold value of complete image segmentation by utilizing the image information entropy, and converting the threshold value into a binary image; and performing image extraction on the binarized image through a watershed algorithm to obtain a required image. Image edges are filtered through an interactive foreground extraction method, images can be effectively segmented by combining information entropy and a watershed algorithm to acquire complete liver CT images. Problems that interference caused by uneven distribution of pixelvalues, mutual connection of foreground sub-images and different shapes of liver organs among individuals are overcome.

The invention discloses a big data-based liver three-dimensional image dynamic demonstration system. The system comprises a data receiving module that obtains at least ten liver CT images, in DICOM formats, of specific liver, a three-dimensional image processing module that generates three-dimensional images of the specific liver, a liver information database that stores the three-dimensional images of the specific liver in a classified manner, and a dynamic demonstration module, wherein the three-dimensional image processing module comprises an image preprocessing sub-module, a liver extraction sub-module and an image drawing sub-module; the image preprocessing sub-module performs image smoothing and image enhancement processing on each liver CT image; the liver extraction sub-module segments the liver data image to detect a contour edge of the liver and extract a contour line of the liver; and the image drawing sub-module constructs a plurality of volume data units between every two adjacent liver CT images in the segmented liver CT images according to corresponding actual spatial positions, and obtains the three-dimensional image of the specific liver by performing shear transformation and two-dimensional image deformation on each volume data unit.

The invention belongs to the technical field of hepatobiliary surgery treatment information sharing. The invention discloses a hepatobiliary surgery treatment information sharing system and sharing method based on Internet. The hepatobiliary surgery treatment information sharing system based on the Internet comprises a patient information acquisition module, an information processing module, a diagnosis module, a central control module, a treatment information generation module, a network communication module, an illness state prediction module, a prevention and control module, a database anda display module. According to the hepatobiliary surgery treatment information sharing system and sharing method based on Internet, benign and malignant lumps are identified by utilizing the classifier trained by a textural feature data set of the liver CT image through the diagnosis module, and the result is not influenced by human subjective factors, so that the human subjective factor influenceof pathological examination and other examinations is avoided, and the diagnosis accuracy is greatly improved; and meanwhile, by utilizing a prediction model based on a deep learning technology through the illness state prediction module, the problem that the small difference of the gene expression quantity is difficult to grasp due to subjectivity of people is solved, and positive significance is achieved for development of gene therapy of liver cancer.

The invention discloses a liver 3D CT reconstruction data information processing system. The liver 3D CT reconstruction data information processing system includes a data processing center, a plurality of first type user terminals and a plurality of second type user terminals. Each first type user terminal is further provided with a data input module used for inputting liver CT images into each first type user terminal, a data primary processing module used for packaging and compressing the liver CT images and source information into specific liver data packets and a data uploading module uploading the specific liver data packets to the data processing center. The data processing center is further provided with a data receiving module used for receiving the specific liver data packets, a 3D image processing module generating specific liver 3D images based on the liver CT images and liver information database storing the specific liver 3D images according to the liver source information. Each second type user terminal is further provided with a log-in retrieval module used for logging in the data processing center for retrieving and browsing the specific liver 3D images.

The invention discloses a liver CT tumor segmentation and classification method based on deep learning. The method comprises the following steps of preprocessing a liver CT image, performing interpolation according to resolution ratios in X and Y directions, performing resampling, searching data, constructing training data, performing liver and tumor segmentation through 2D Dense U-Net, extracting corresponding three-dimensional features through 3D Dense U-Net, and forming a generative adversarial network. The invention belongs to the technical field of liver CT tumor segmentation and classification methods, and particularly relates to a method which can assist doctors in early screening and diagnosis of liver cancer by using a deep learning technology, can greatly reduce the workload of the doctors, improves the accuracy of liver cancer diagnosis, and improves the accuracy of liver cancer diagnosis. The liver CT tumor segmentation and classification method based on deep learning has a great application prospect.

The invention provides a liver CT diagnosis system based on computer assistance. The liver CT diagnosis system comprises an image acquisition module, an image three-dimensional reconstruction module,an image enhancement module and an image segmentation module, wherein the image acquisition module comprises a CT machine and a computer, the CT machine imports acquired liver CT slice images into thecomputer, and the computer processes the liver CT slice images to obtain two-dimensional liver images; and the liver three-dimensional image reconstruction module is used for drawing and reconstructing the two-dimensional liver image to obtain a three-dimensional liver image. The image enhancement module is used for carrying out blood vessel enhancement processing on the three-dimensional liver image to obtain a blood vessel enhanced image; the image segmentation module is used for segmenting the blood vessel enhanced image; according to the CT diagnosis method for the liver based on computerassistance, scientific research personnel can conveniently research and know the position of the blood vessel in the liver area, and liver operation can be well assisted.

The invention provides a liver blood vessel segmentation system based on a Hessian matrix and a gray scale method. The liver blood vessel segmentation system comprises an image acquisition and preprocessing unit, a blood vessel image enhancement unit, a portal vein center line extraction unit, a liver blood vessel segmentation unit and a data storage unit. The image acquisition and preprocessing unit is used for converting the liver CT image into an image file conforming to the gray level display range of the display device and preprocessing the image file; the blood vessel image enhancement unit is used for performing image enhancement on the preprocessed liver image to obtain a liver blood vessel enhanced image; the portal vein center line extraction unit is used for extracting and verifying a portal vein center line; the liver blood vessel segmentation unit is used for segmenting a liver blood vessel image according to the portal vein center line extracted by the portal vein centerline extraction unit; according to the liver blood vessel three-dimensional reconstruction system and method, liver blood vessels are accurately and reliably segmented and subjected to three-dimensional reconstruction, and clinicians are helped to improve the diagnosis efficiency.

The invention discloses a method and a device for segmenting a liver tumor under a CT image. The method comprises the steps: cutting each original CT image of a first preset size in a training set into three adjacent slices of a second preset size, taking a segmentation result of a middle slice as a label, and enabling the corresponding three adjacent slices to form a training sample; inputting one or more training samples into a preset network for training to obtain a trained segmentation model; wherein the preset network comprises a preprocessing module and a full-convolution point cutting network; and for any to-be-detected CT image, inputting three adjacent slices of a first preset size into the trained segmentation model each time, and obtaining a segmentation result of a middle slice based on a joint loss function. According to the scheme, the problems existing in traditional liver tumor image segmentation can be effectively solved, the problems of large imaging difference of liver CT images, different tumor shapes and sizes and the like are solved, and the performance and effect of the automatic liver tumor segmentation method are improved.

The invention relates to a liver tumor segmentation method based on an improved U-net network, and the method comprises the steps: 1, obtaining an abdominal CT data set, and carrying out the preprocessing operation; 2, before the liver tumor is segmented, segmenting a liver region, building a neural network for liver segmentation based on a Keras deep learning framework, and selecting tensor flowat the rear end; 3, training the liver segmentation network based on the improved U-net; 4, constructing a liver tumor segmentation network based on the improved U-net based on a Keras deep learning framework, and training the network; and 5, segmenting a liver region from the abdominal liver CT image by adopting a liver segmentation network based on improved U-net, and segmenting tumors and normal liver tissues from the liver region. The method not only can eliminate a large amount of wrong segmentation, but also can reduce the complexity of a network model.

The invention discloses a differential curvature-based liver surface smoothness measuring method, and mainly relates to a digital image processing technology. According to the technical scheme, the method comprises the following steps: (1) preprocessing a liver CT image by using a medical image window modulation algorithm to enhance the contrast of an abdominal liver area; (2) obtaining a segmentation result of the liver by using a deep network model; (3) obtaining the contour of the liver segmentation result by using an edge detection algorithm; (4) selecting a to-be-evaluated liver area by utilizing man-machine interaction to obtain a to-be-evaluated liver profile curve; and (5) measuring the smoothness of the liver shape by utilizing curvature calculation of the curve. The invention effectively quantifies the smoothness of the liver surface, achieves the accurate and objective measurement of the smoothness of the liver contour, and further assists a doctor in the quantitative evaluation of the liver function.

The invention provides a hepatic vessel three-dimensional reconstruction and visualization method based on a vascular tubular model. The method comprises the steps of S10, obtaining a liver CT image,and adjusting the window width and window position, so as to enable the liver CT image to accord with the gray scale display range of display equipment; s20, performing preprocessing of denoising andimage enhancement on the liver CT image conforming to the gray level display range of the display device; s30, segmenting the blood vessel by using a segmentation method based on a Hessian matrix anda region growing algorithm according to the preprocessed blood vessel image; s40, filling a cavity caused by blood vessel segmentation by adopting a morphological method; s50, volume rendering is performed on the blood vessel by using a triangular mesh surface rendering and light projection algorithm according to inherent characteristics of the liver blood vessel so as to realize three-dimensionalreconstruction; and S60, visually displaying the three-phase image sequence by adopting a computer-aided liver blood vessel visualization system based on a CT dynamic enhanced image. According to themethod, segmentation and three-dimensional reconstruction are accurately and reliably carried out on the liver blood vessel, and visual display can be well carried out to assist a doctor in determining the liver condition.

The invention discloses a CT image liver artery segmentation method and system based on deep learning, and belongs to the technical field of medical image processing and artificial intelligence. The method comprises the steps that all liver CT development images are acquired, and the image size is adjusted to a fixed size; normalizing the CT value of the liver part, and inputting the normalized CTvalue into a neural network model; calculating the output of the neural network model and the loss value of a liver artery mask, and updating the parameters of the neural network model according to the loss value; traversing all the training samples, and completing training learning of the neural network model to obtain a liver artery segmentation model; and according to the liver artery segmentation model, segmenting the normalized CT value of the liver part to obtain a segmentation result of the liver artery. The system comprises an acquisition module, an adjustment module, a normalizationinput module, a calculation updating module, a training module and a segmentation module. According to the method, the actual condition of the hepatic artery can be accurately obtained, and no human participation is needed in the segmentation process.

The invention relates to a fatty liver intelligent grading evaluation method based on abdominal CT, and relates to the field of intelligent medical image diagnosis. The method comprises the followingsteps: reading an abdominal CT image of a patient, selecting nearby four layers of slices corresponding to the maximum area of liver and spleen to construct a training sample set, and carrying out thepreprocessing of the data of the training sample set; constructing a UNet segmentation network model, sending the training sample to the segmentation network model for supervised learning, and aftertraining convergence of the segmentation model, using the model to segment CT slices to segment liver tissues and spleen tissues in the slices; respectively carrying out gridding cutting on liver tissues and spleen tissues to obtain a plurality of small rectangular areas with the same area, randomly selecting five small rectangular areas in the two tissues as sampling areas, calculating respectivegray average values as a liver CT value and a spleen CT value, and finally grading the fatty liver according to a liver / spleen CT ratio. According to the method, full-automatic segmentation of the liver tissue and the spleen tissue based on the abdominal CT image is realized, so that the fatty liver is intelligently graded and evaluated.

The invention provides a one-dimensional convolutional neural network-based improved region growing algorithm for interactive segmentation of a liver CT image. Multiple kinds of information such as gray values, spatial information and different gradient values of pixels are taken into overall consideration through a neural network to serve as growth rules, so the stability of the region growth method is improved, and the segmentation capacity of the algorithm for an edge complex structure is enhanced. The method comprises the following specific steps: firstly, preprocessing an image, extracting slices containing the liver in a CT image sequence set, and converting a CT image into a grayscale image by using a window algorithm; then, carrying out image edge detection, calculating gradient values of a pixel under different edge detection operators to serve as features of the pixel in order to form a pixel feature vector; constructing a network model, extracting a training data set, and training the network model; and finally, performing segmentation, using the trained convolutional neural network model as a growth criterion of a region growth algorithm, using a mouse to click a liverregion to generate an initial segmentation result, and using a morphological method to fill holes to obtain a final result.

The invention discloses a liver CT image segmentation system and algorithm based on mixed supervised learning, the image segmentation system comprises an image preprocessing unit, a feature extraction unit, a word vector segmentation unit and a single-layer convolution classification unit, and the image preprocessing unit is in data connection with the feature extraction unit. The feature extraction unit is in data connection with the word vector segmentation unit and the single-layer convolution classification unit. A multi-task framework is used for carrying out segmentation and classification tasks respectively, high segmentation precision is achieved through a large amount of weak label data and a small amount of strong labels, network parameters shared among multiple tasks are increased as much as possible for the problem that the independent parameter quantity among the multiple tasks is large, and a learnable word vector v is used for achieving the segmentation tasks; the number of independent parameters is reduced, the result consistency among multiple tasks is improved, high performance can be kept for classification and segmentation tasks, and good consistency is embodied.

The invention discloses a liver tumor image segmentation model training method, which comprises the following steps of: firstly, acquiring a liver CT (Computed Tomography) image, performing preprocessing operation on the liver CT image to obtain a sample image and a tumor plaque image, and then performing specific training on a tumor image segmentation model: at the first stage, inputting the sample image into a preset image segmentation model for training, obtaining a trained first tumor image segmentation model; in the second stage, inputting the tumor plaque image into the first tumor image segmentation model for training to acquire a second tumor image segmentation model after training is completed; and in the third stage, inputting the sample image into the second image segmentation model for training to acquire a tumor image segmentation model after training is completed. A network is trained from the two aspects of whole liver CT image input and tumor plaque image, the specific features of the tumor can be integrated, and the accuracy of liver tumor image segmentation is improved in combination with the overall background of the tumor data.

The invention discloses a CT angiography image segmentation method suitable for Lee's artificial liver treatment, and the method comprises the steps: obtaining a liver CT angiography image, carrying out the data enhancement processing of the liver CT angiography image in an upper computer, inputting an obtained image into a CT angiography image segmentation network, and obtaining a two-dimensional segmentation result image and a classification result; the CT angiography image segmentation network comprises a UNet + + image segmentation network and a ResNet-Att image classification network; the ResNet-Att image classification network comprises a prediction network based on ResNet 50-base, and an improved Attention self-attention module is added into the prediction network. The CT angiography image segmentation method suitable for the Lee's artificial liver treatment can be effectively used for segmenting and classifying the CT images of the liver of the patient.

The invention discloses a liver CT image tumor segmentation method based on information fusion. The method specifically comprises the following steps: step 1, carrying out the smoothing and denoising preprocessing of an image through a three-dimensional histogram reconstruction model to obtain a preprocessed liver CT image B; step 2, respectively extracting three different features of spectrum, texture and spatial relationship of the liver CT image B to obtain three feature images corresponding to the liver CT image B; step 3, carrying out aggregation processing on the pixels of the three different feature images of the liver CT image B, and detecting the edge of the tumor part in the original image A by using an improved region growing algorithm to obtain a fused image C1; and step 4, performing post-processing operation on the edge of the tumor part detected in the image C1 fused in the step 3 in combination with morphological operation to form a final image segmentation result graph A. The method solves the problem of fuzzy tumor edge detection in a CT image tumor segmentation method in the prior art.

The invention discloses a liver medical image registration and fusion method and system, and the method comprises the steps: obtaining an arterial-phase and venous-phase liver CT image sequence chart group for the same liver, and enabling the arterial-phase liver CT image sequence chart group to comprise at least ten arterial-phase liver CT images, the vein phase liver CT image sequence group comprises at least thirty vein phase liver CT images; sequentially registering each arterial phase liver CT image, finding out a vein phase liver CT image with the same cross section from the vein phase liver CT image sequence group, and acquiring at least ten vein phase liver CT images registered with the at least ten arterial phase liver CT images; sequentially fusing each artery-phase liver CT image and one vein-phase liver CT image registered with the artery-phase liver CT image to obtain at least ten liver CT fusion images; and performing sequence three-dimensional reconstruction on the at least ten liver CT fusion images to obtain a liver three-dimensional fusion CT image.