63 results about "Abdominal ct" patented technology

A method Is disclosed for fully automated segmentation of human vertebral body images in a CT (computerized tomography) study with no user interaction and no phantoms, which has resiliency to, anatomical abnormalities, and protocol and scanner variations. The method was developed to enable automated detection of osteoporosis in CT studies performed for other clinical reasons. Testing with 1.044 abdominal CTs from multiple sites, resulted in detection of 96.3% of the vertebral, bodies and 1% false positives. Of the detected vertebral bodies, 83.3% were segmented adequately for sagittal plane quantitative evaluation of vertebral, fractures indicative of osteoporosis. Improved results were observed when selecting the best sagittal plane of 3 for each vertebra, yielding a segmentation success rate of 85.4%. The method is preferably implemented in software as a building block in a system for automated osteoporosis detection.

The embodiment of the invention provides a liver tumor segmentation method and device based on multi-stage CT image guidance, and the method comprises the steps: obtaining an abdomen CT image with enhanced contrast, inputting the abdomen CT image with enhanced contrast into a preset single-channel full convolutional neural network, and obtaining a liver region-of-interest image; inputting the liver region-of-interest image into a preset tumor segmentation network to obtain a tumor segmentation result corresponding to the contrast-enhanced abdominal CT image; wherein the tumor segmentation network is obtained by performing multi-channel fusion training according to liver region-of-interest image samples in different periods and image samples with tumor region marks corresponding to the abdominal CT image samples. According to the embodiment of the invention, the liver tumor CT images in different periods are effectively subjected to feature mining by adopting a multi-channel fusion training network, so that the trained network has higher segmentation precision and better robustness on liver tumors.

The invention discloses an abdominal organ segmentation method based on secondary three-dimensional region growth, which belongs to the field of medical image processing. The method comprises the following steps: firstly, combining with apriori knowledge such as anatomical position and gray value distribution of an interested organ to extract original seed points automatically, and combining with image edge extracted by a Canny edge detection algorithm to carry out first three-dimensional region growth of an image; then, extracting the three-dimensional morphologic edge of a segmentation result graph obtained after the first growth; and finally, combining with the extracted three-dimensional morphologic edge and the Canny edge of the original image to carry out second three-dimensional region growth of the original image, and carrying out three-dimensional morphologic expansion of the segmentation result obtained after the second three-dimensional region growth to obtain a final segmentation result of the interested abdominal organ. The abdominal organ segmentation method effectively restrains the phenomenon of oversegmentation existing in the prior three-dimensional region growth method, and can accurately extract an interested organ from an abdominal CT image; therefore, the method can be used for assisting clinical diagnosis.

The embodiment of the invention provides a liver tumor segmentation method and device based on deep learning, and the method comprises the steps: obtaining an abdominal CT/MR image of a patient, inputting the abdominal CT/MR image of the patient into a preset dense connection type full convolutional neural network, and obtaining a liver region-of-interest image; and inputting the liver region-of-interest image into a preset deep convolutional generative adversarial network generator to obtain a tumor segmentation result. According to the embodiment of the invention, through adoption of the densely connected full convolutional neural network and the deep convolutional generative adversarial network, the robustness and segmentation accuracy of liver tumor segmentation are improved, and the features extracted in the two-dimensional plane and the structure features extracted in the three-dimensional space are fused, so that the tumor segmentation precision is higher.

The invention discloses a method for automatically segmenting liver tumors in abdominal CT sequence images. The method comprises the following steps of: preprocessing: preprocessing an abdominal CT sequence image so as to obtain a liver area in the image; liver enhancement: improving a contrast ratio of normal liver parenchyma to tumor tissue by adoption of a segmented nonlinear enhancement operation and an iterative convolution operation according to a grey level distribution characteristic of the liver area; automatic segmentation: constructing an image segmentation energy function for multi-target segmentation by utilizing the enhancement result and combining image boundary information, minimizing the energy function by adoption of an optimal algorithm and obtaining a liver tumor preliminary automatic segmentation result; and post-processing: optimizing the preliminary segmentation result by adoption of a three-dimensional mathematic morphological open operation, and removing mis-segmented areas to improve the segmentation precision. The method is beneficial for helping radiological experts and surgeons to timely and effectively obtaining overall information and three-dimensional display of liver tumors and providing technical support for computer-aided diagnosis and treatment of liver diseases.

The invention discloses an extraction method of main blood vessels from abdominal CT images based on watershed of three-dimensional region, which mainly overcomes the shortages that the existing segmentation method of blood vessels has high complexity for computation and can not utilize the third dimension information among CT images well. The extraction method comprises the following steps of: (1) reading a set of abdominal CT images, arranging the abdominal CT images according to the imaging sequence to obtain an abdominal three-dimensional data volume; (2) implementing three-dimensional watershed cutting on a plurality of small cubes in the three-dimensional data volume, and communicating each cut small cube to extract main blood vessels and the communicated organs in the abdominal CT images from the three-dimensional data volume; (3) extracting the main abdominal organs from the three-dimensional data volume; and (4) subtracting the results respectively obtained from the step (2) and step (3), and implementing post processing on the subtracted result to obtain the final main blood vessels in the abdominal CT images. The method can faster and more completely obtain the main blood vessels in the abdominal CT images and can be used for auxiliary diagnosis of the main abdominal blood vessels.

The invention discloses an abdominal CT image target organ registration method based on deep learning. Firstly, constructing an abdomen CT image database; secondly, constructing a network model basedon deep learning, and introducing a coordinate convolution layer into a convolutional neural network module of the network model so as to enhance the learning ability of the network model for target position information; then, considering that the data volume of the abdominal CT image containing the target organ bounding box is small, based on the transfer learning technology, inputting the data into a natural scene database to pre-train a network model, and then inputting the data into an abdominal CT image database to perform parameter fine adjustment on the model so as to realize abdominaltarget organ detection; and finally, constructing an abdominal target organ CT image pair, constructing a similarity measurement function according to gradient and gray level distribution characteristics between pixel points of the image pair, minimizing the function based on a gradient descent method, and realizing registration of the abdominal CT image to the target organ. According to the method, the strategy of firstly extracting the target organ area of the abdominal CT image and then registering is adopted, the influence of factors such as complex background and noise of the abdominal CTimage on target organ registration is reduced, the registration precision is high, and the robustness is high.

According to the liver tumor automatic classification method and device based on multi-stage CT image analysis, full-automatic bile duct cell carcinoma and hepatocellular carcinoma can be recognized,and a high-precision bile duct cell carcinoma and hepatocellular carcinoma recognition model is obtained. The method comprises the following steps: (1) acquiring a contrast-enhanced abdominal CT scanning image, storing the contrast-enhanced abdominal CT scanning image as an arterial phase, a portal vein phase and a delay phase, and carrying out definite diagnosis on liver cancer categories to which all data belong to serve as a model training gold standard; (2) constructing a three-dimensional full convolutional neural network segmentation model, and segmenting the intrinsic characteristics ofthe liver tissue in each stage from the abdominal CT image through model training learning; (3) constructing a three-dimensional convolutional neural network classification model; and inputting the image data obtained by segmentation into a classification model for training, so as to enable the model to perform joint learning and training on the cancer features in multiple periods, thereby predicting the category to which the cancer belongs, comparing the prediction result with a gold standard, and supervising the training process of the model in a loss value feedback mode.

The invention discloses a soft tissue minimally invasive surgical training method based on virtual reality, which can enable a user to operate a force feedback device to perform minimally invasive puncture or forceps pulling on the soft tissue, comprising the following steps: 1, reverse modeling the abdominal CT image of a patient; 2, optimize that model; 3, carry out finite element puncture simulation on that model; 4, that finite element simulation result are processed in Matlab software to verify the correctness of the model by compare with the experimental results; 5. Complete the module design of model import, scene rendering, collision detection, deformation algorithm, data input and output of force feedback device in Microsoft Visual Studio 2013 software, and realize the function ofinteraction between user and virtual scene through force feedback device. The invention can improve the system accuracy through modeling and simulation, adopt voxel octree and finite element algorithm, improve the collision and deformation accuracy and real-time, accurately perceive the virtual force through the external force feedback device, and effectively enhance the immersion feeling.

The invention relates to an automatic segmentation method for an abdominal CT liver lesion image based on a three-stage cascade network. The method comprises the following steps: S1, acquiring three-dimensional abdominal liver CT image data; S2, preprocessing and standardizing the obtained three-dimensional abdominal liver CT image data; S3, inputting the three-dimensional abdominal liver CT imagedata after preprocessing and data standardization into AuxResUnet liver image segmentation model, and then taking 3D maximum connected region from the obtained three-dimensional abdominal liver CT image data segmentation result to exclude false positive region, so as to obtain liver VOI; S4, adopting the segmentation result of the three-dimensional abdominal liver CT image data obtained by S3 asa mask of the CT liver image data to cover the liver VOI obtain by S3; S5, inputting the covered liver VOI into AuxResUnet liver image lesion segmentation model for lesion segmentation, and obtainingthe liver image lesion segmentation result. The image segmentation method provided by the invention can realize fast and accurate segmentation of liver and liver pathological changes.

The invention discloses a CT image computer-aided diagnosis system and method for the liver disease based on data mining. The system comprises an input module, a texture feature extraction module, a classification diagnosis identification module and an output module. The method adopts the system and comprises the following steps of: segmenting and extracting the contour of an abdominal CT image inadvance, marking the normal liver CT, the liver cyst or the liver cancer, and importing the preprocessed image into the system; analyzing the image texture of the imported image to obtain the 13-dimensional gray-scale co-occurrence texture feature in the CT image by the image texture feature extraction module; according to a database, establishing a diagnosis model, and substituting the input liver CT image texture feature into a diagnosis model classifier for processing to obtain a diagnosis result and accuracy, wherein the database is stored in the classification diagnosis module and is composed of liver cancer and liver cyst image texture features and normal liver image texture features.

The present invention claims a detection method based on abdominal CT medical image fusion classification. The detection method comprises: an acquisition step of acquiring an abdominal CT scanning image of a patient; a pre-processing step of performing pre-processing to obtain a binarized grayscale image; a morphological corrosion and dilation step of performing morphological corrosion calculationto obtain a corroded image, and performing dilation operation on the binarized grayscale image to obtain a dilated image; an operation step of performing opening operation on the corroded image to obtain an open operation map and performing closed operation on the dilated image to obtain a closed operation map; a Fourier transform step of obtaining a Fourier transform value; an abdominal medicalimage fusion classification step for performing fusion classification distinguishing on the abdominal CT scanning image; and a lesion detection step of performing determination according to the Fourier transform value and the fusion classification distinguishing result to obtain a suspected lesion area of the patient. The invention can lower the complexity and improve the accuracy of recognition and diagnosis of abdominal lesions.

The invention relates to a liver tissue biopsy localization method under mimics three-dimensional reconstruction guidance for abdominal CT (computed tomography). The method includes the following steps: S1, utilizing mimics software to perform three-dimensional reconstruction on ribs in abdominal CT; S2, selecting a puncture point according to a three-dimensional model built in the step 1; S3, evaluating puncture effect on the CT horizontal plane, and measuring the coronal plane position of the puncture point; S4, measuring the horizontal plane position of the puncture point on the CT coronal plane with the aid of the inferior margin of angulus scapulae; S5, simulating labelling localization effect on the three-dimensional model diagram. The method has the advantages that the 3D model for the liver surrounding tissue is built by the aid of the mimics program, liver puncture is simulated in the computer, puncture effect is observed, surface skin localization parameters and puncture depth of liver puncture are calculated, liver puncture localization is performed on a human body according to the parameters, localization is accurate, simulative liver puncture is performed in the computer prior to liver puncture on the human body, relevant parameters are provided, success rate of liver puncture on the human body is increased, and complications are reduced.

The invention discloses a semi-supervised renal artery segmentation method based on a dense bias network and an auto-encoder. The method comprises the following steps: for an existing abdominal CT angiography image, segmenting a kidney region in the image to obtain a region-of-interest image, marking the region-of-interest image to obtain a real mask of a renal artery, and forming a supervised training data set and an unsupervised training data set; inputting the unsupervised training data set into a three-dimensional convolution denoising auto-encoder for image reconstruction training to obtain a trained denoising auto-encoder model; inputting the supervised training data set into a denoising auto-encoder model to obtain prior anatomical features of each image, and inputting the prior anatomical features and the corresponding images into a constructed dense bias network for segmentation training to obtain a segmentation model; inputting the new abdominal CT angiography image to be segmented into the denoising auto-encoder model to obtain prior anatomical features of the image, and inputting the prior anatomical features into the segmentation model to obtain a segmentation result.According to the invention, a high-accuracy output result can be obtained, and renal artery segmentation can be quickly realized.

The invention relates to a liver cirrhosis remote consultation system based on abdominal computed tomography (CT) image transmission. A CT instrument has a quite high tissue density resolution ratio and can be used for providing objective bases for imaging diagnosis of liver pathological changes. The liver cirrhosis remote consultation system based on the abdominal CT image transmission comprises a patient terminal collection system, a cloud computing processing center and a hospital background service management system, wherein the patient terminal collection system comprises an abdominal CT image shooting device, a medical image processing system and a data remote transmission system; the cloud computing processing center comprises an image sorting, encrypting and decoding processing module, a computer aided detection system and an expert outpatient service calling system; and the hospital background service management system comprises a case image query / storage database and diagnostic workstations. The liver cirrhosis remote consultation system based on the abdominal CT image transmission is used for diagnosis of the liver pathological changes, especially liver cirrhosis, which relate to an abdominal CT image examination, and can provide the most intuitive liver cirrhosis tissue structure images for doctors for consultation at different places; noise can be filtered when preliminary information is transmitted in a remote mode; geographical restrictions are broken; and the consultation at different places and multiple places can be realized.

The invention relates to a CT image based automatic extraction method of a 3D liver bounding volume. Volume measurement is carried out on liver segmentation data of N abdominal CT, the 3D CT slice anda liver mask layer in which the volume approaches the average liver volume most are selected, and a rectangular liver area frame corresponding to the CT slice is determined; a CT image of the determined rectangular liver area as a standard CT image is registered with a CT image of a non-determined liver bounding box roughly, the step and iteration frequency of an optimizer in rough registering are changed, and a transformation matrix is output; and geometric transformation is carried out on a binary image of the rectangular liver area frame, the matrix after transformation is placed in a 2D coordinate system, coordinates of all points whose grayscale values equal 1 are searched from the matrix, and a rectangular boundary is determined according to X and Y maximums and X and Y minimums, and serves as a liver area rectangle. The data bulk of image processing is reduced, rapid and accurate liver segmentation is guided, and the method has great significance in guiding segmentation of a liver tumor and provides basis for subsequent researches.

Embodiments describe an accurate and rapid method for assessing spinal bone density on chest or abdominal CT images using post-processed colored images. Post-processing of CT images for the purposes of displaying the spine is followed by color enhancement of routine unenhanced or contrast enhanced CT images to improve diagnostic accuracy, inter-observer agreement, reader confidence and/or time of interpretation as it relates to assessing bone density of the spine. CT images are post-processed (without changes to the standard-of-care CT imaging protocol and without additional cost or radiation for the patient) to straighten the spine for improved visualization of multiple segments. The color-enhanced images can be displayable simultaneously with the grayscale images. Methods and systems are provided for performing opportunistic bone density screening.

The invention relates to the technical field of abdominal lymph node partitioning, in particular to an abdominal lymph node partitioning method based on an attention mechanism neural network. The abdominal lymph node partitioning method based on the attention mechanism neural network is used for solving the problems that in the prior art, a doctor has a large difference in film reading results ofthe same abdominal CT medical image, and prediction of abdominal lymph node partitioning is inaccurate. The method comprises the following steps: step 1, preparing data; step 2, generating a mask, andpreprocessing the data; step 3, constructing an attention mechanism residual network model; step 4, repeating the step 3, and constructing and training a model of lymph node relative position partitioning; and step 5, classifying the abdominal lymph nodes automatically detected by the detection task by using the model trained in the step 3 and the step 4. According to the method, the original CTimage and the mask are overlapped to serve as input, and the attention mechanism is introduced into the deep residual neural network, so that the abdominal lymph nodes in the CT image can be accurately partitioned.

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 multi-view-angle-based multi-task liver tumor image segmentation method. After an abdominal CT image is preprocessed, liver segmentation and tumor segmentation of the abdominal CT image are obtained at the same time in a slice form through a convolutional neural network model. The input of the model is a three-dimensional CT slice with the size of 256 * 256 * 3, and the output of the model is corresponding segmentation of a middle slice. The model comprises a segmentation module and a fine trimming module, and a rough segmentation result and a fine trimming segmentation result are obtained respectively. The model is optimized through a combined loss function, and instability in the optimization process is avoided. According to the method, segmentation is carried out from three perspectives of a three-dimensional CT image, and three segmentation results are fused into one to obtain a final segmentation result. According to the method, liver and tumor segmentation of the abdominal CT image is realized, and the problems that three-dimensional space information cannot be utilized and optimization is unstable in the segmentation process are effectively solved.