84 results about "Pulmonary parenchyma" patented technology

The invention discloses a generative adversarial network improved CT medical image pulmonary nodule detection method. The method includes: 1), acquiring a section of a pulmonary CT image; 2), separating according to image morphological properties to acquire a ROI pulmonary parenchyma area; 3), acquiring different suspected pulmonary nodule candidate sets according to a connected domain formed by abinarized image; 4), building a model of an assistant classifier generative adversarial network to generate positive samples overcome the circumstance that positive-negative sample number is unbalanced; 5), building a convolution neural network to classify suspected pulmonary nodule parts to acquire pulmonary nodule areas; 6), using a non-maximum suppression algorithm to acquire a final area of pulmonary nodule. By the method, efficient processing performance of a computer can be fully utilized, certain expandability is provided, and data processing efficiency is improved; through a convolution neural network algorithm, classifying accuracy is improved, CT image data processing performance is improved, and pulmonary nodule images can be built and analyzed more efficiently.

The invention discloses a CT image pulmonary parenchyma three-dimensional semantic segmentation method based on a deep neural network. The segmentation method comprises an offline part and an online part. The offline part comprises four steps: preprocessing a data set; Constructing a full convolutional neural network framework; Constructing a GRU cyclic convolutional neural network framework; Network training. The online part comprises the following five steps: preprocessing an image; Extracting and fusing pixel features; Extracting and fusing voxel characteristics; Performing segmented output; Segmentation result evaluation. The feature fusion comprises feature fusion between an encoding layer and a decoding layer and feature fusion between anatomical structure layers. A deep neural network model with a gating circulation unit is designed, and spatial features are extracted by using anatomical structure prior information of a lung to effectively represent an apparent evolution relationship between fault sequences, so that accurate three-dimensional semantic segmentation is carried out on lung parenchyma in a CT image.

A pulmonary nodule detection device and method based on a shape template matching and combining classifier comprises an input unit, a pulmonary parenchyma region processing unit, a ROI (region of interest) extraction unit, a coarse screening unit, a feature extraction unit and a secondary detection unit. The input unit is used for inputting pulmonary CT sectional sequence images in format DICOM; the pulmonary parenchyma region processing unit is used for segmenting pulmonary parenchyma regions from the CT sectional sequence images, repairing the segmented pulmonary parenchyma regions by the boundary encoding algorithm and reconstructing the pulmonary parenchyma regions by the surface rendering algorithm after the three-dimensional observation and repairing; the ROI extraction unit is used for setting a gray level threshold and extracting the ROI according to the repaired pulmonary parenchyma regions; the coarse screening unit is used for performing coarse screening on the ROI by the pulmonary nodule morphological feature design template matching algorithm and acquiring selective pulmonary nodule regions; the feature extraction unit is used for extracting various feature parameters as sample sets for the post detection according to selective nodule gray levels and morphological features; the secondary detection unit is used for performing secondary detection on the selective nodule regions through a vector machine classifier and acquiring the final detection result.

In a method and apparatus for image processing proceeding from a computed tomography (CT) image of a lung as an original image that is registered using a contrast agent, pulmonary parenchyma pixels are determined, the pulmonary parenchyma pixels are presented in false colors, and the remaining image regions are presented in the gray scale values of the original image.

The invention provides an automatic division method for pulmonary parenchyma of a CT image. According to the automatic division method, the CT is divided through carrying out a random migration algorithm for two times to obtain the accurate pulmonary parenchyma; in the first time, the random migration algorithm is used for dividing to obtain a similar pulmonary parenchyma mask; and in the second time, the random migration algorithm is used for repairing defects of the periphery of a lung and dividing to obtain an accurate pulmonary parenchyma result. Seed points, which are set by adopting the random migration algorithm to divide, are rapidly and automatically obtained through methods including an Otsu threshold value, mathematical morphology and the like; and manual calibration is not needed so that the working amount and operation time of doctors are greatly reduced. According to the automatic division method provided by the invention, a process of 'selecting the seed points for two times and dividing for two times' is provided and is an automatic dividing process from a coarse size to a fine size; and finally, the dependence on the selection of the initial seed points by a dividing result is reduced, so that the accuracy, integrity, instantaneity and robustness of the dividing result are ensured. The automatic division method provided by the invention is funded by Natural Science Foundation of China (NO: 61375075).

The invention provides a lung image segmentation method and device and lung lesion area identification equipment, and the method comprises the steps: converting a to-be-segmented lung image into a first binary image according to a preset gray threshold; performing negation processing on the first binarized image to obtain a second binarized image; performing hole filling on the second binarized image, and performing negation processing to obtain a third binarized image; removing an interference region in the third binarized image to obtain a first mask; filling the first mask to obtain a second mask; performing subtraction operation on the second mask and the first mask to obtain a lung region mask; multiplying the lung region mask and the to-be-segmented lung image to obtain a lung parenchyma image. According to the invention, the lung parenchyma image can be extracted rapidly and accurately and lung lesion area identification is carried out by using the three-dimensional image, so that doctors are assisted well and the working efficiency is improved.

The present invention discloses a pulmonary nodule segmentation method based on the Hession matrix and three-dimensional shape indexes. According to the method, medical CT images are fully utilized; sequential pulmonary parenchymas are segmented through using an optimal threshold according to the gray values of sequential CT images, and the volume data of the three-dimensional pulmonary parenchymas are constructed; the Hession matrix feature values of each voxel point in the volume data of the three-dimensional pulmonary parenchymas are calculated; the three-dimensional shape indexes are constructed according to the shape features of a three-dimensional nodule model and on the basis of the Hession matrix feature values and two-dimensional shape indexes; and a three-dimensional sphere-like filter, namely, a 3D shape index nodule detection function is constructed finally and is adopted to perform nodule detection on the three-dimensional volume data of the pulmonary parenchymas, and with detected nodule regions adopted as a plurality of seed points of region growth, three-dimensional segmentation is performed on nodules on the basis of a confidence-based region growing algorithm. The method of the invention is simple in operation, can automatically detect and segment different types of suspected pulmonary nodules and has high stability and high accuracy.

The invention discloses a CT image processing system for pneumoconiosis, and the system comprises an image processing server which comprises a CPU, a graphic processor and a DICOM reading-writing unit connected to the graphic processor, wherein the DICOM reading-writing unit reads a CT image from the graphic processor and analyzes the CT image; a CT image preprocessing unit which carries out the gray scale inhomogeneity correction, denoising and artifact removing of the CT image analyzed by the DICOM reading-writing unit; an image segmenting unit which carries out the pulmonary parenchyma segmentation, pulmonary nodule segmentation and pulmonary nodule false positive target segmentation of the CT image after preprocessing; a depth network learning unit which extracts the high-dimensional features of a sub-image block where a pulmonary nodule region is located after the segmentation; and an SVM classification unit which receives the high-dimensional features for classification. The system is high in classification precision of the data of the CT image, and is stable in robustness.

The invention discloses a method for reconstruction of a super-resolution coronary sagittal plane image of a lung 4D-CT image based on motion estimation. The method for reconstruction of the super-resolution coronary sagittal plane image of the lung 4D-CT image based on the motion estimation comprises the sequential steps of (1) reading data of the lung 4D-CT image which is formed by a plurality of lung 3D images, wherein the phase positions of the lung 3D images are different; (2) extracting coronary sagittal plane images, corresponding to the same position of the lung, from all the phase positions according to the data of the lung 4D-CT image; (3) estimating motion vector fields between the lung coronary sagittal plane images with different frames based on the full search block matching algorithm; (4) reconstructing the super-resolution lung 4D-CT coronary sagittal plane image by means of the iteration back projection method and based on the motion vector fields obtained in the step (3). According to the method for reconstruction of the super-resolution coronary sagittal plane image of the lung 4D-CT image, the resolution ratio of the reconstructed super-resolution lung 4D-CT coronary sagittal plane image obtained with the method is improved obviously, the brightness and definition of blood vessels and peripheral tissue in the lung parenchyma are improved obviously in a partial enlarged image, the limitation of low resolution caused by the collection time and radiological dose is eliminated, and accurate radiotherapy of lung cancer can be effectively guided.

The invention discloses a pulmonary nodule detection method based on a three-dimensional region generation network, and belongs to the field of medical image detection. The method comprises the following steps of firstly, dividing and preprocessing an image data set containing pulmonary parenchyma; secondly, on the basis of the structural design of the pulmonary nodule detection network, constructing an SRI module and an SI module and using for image encoding and decoding operations; and on the training data set, adopting the cross entropy and Smart L1 function loss, and using a random gradient descent method to optimize the network; and in the test stage, inputting the preprocessed test data set into the optimized network to detect candidate pulmonary nodules, and then further determiningthe pulmonary nodules by using non-maximum suppression. According to the method, aiming at the characteristic that the pulmonary nodule size difference is large, space and channel information is fully utilized in the aspects of network construction and training, the pulmonary nodule detection capability of the network is improved, and the experiments show that good pulmonary nodule detection precision and detection effectiveness can be obtained.

The invention provides a segmentation method of pulmonary nodules in lung CT images, comprising the following steps: (1) extracting pulmonary parenchyma contour lines including pulmonary nodules fromlung images to obtain pulmonary parenchyma images; (2) adopting clustering algorithm to segment the suspected pulmonary nodule region from the pulmonary parenchyma image in the step (1); (3) removingthe region of suspected pulmonary nodules by image roundness method, and eliminating the interference of blood vessel or trachea; (4) according to the size of the suspected pulmonary nodules in step (3) Gaussian template is constructed, and the region with high correlation coefficient is selected to extract the pulmonary nodules. The invention considers the shape characteristic of the solitary pulmonary nodule and the distribution characteristic of the image gray value formed by the change of the CT value of the lung, and reduces the error of the segmentation result of the pulmonary nodule.

The invention discloses a pulmonary nodule detection method based on machine learning so that pulmonary nodule detection can be carried out automatically and the high precision can be kept. The methodcomprises: acquiring a CT image of a lung; segmenting the CT image to obtain a pulmonary parenchyma image; segmenting the pulmonary parenchyma image to obtain a plurality of pulmonary nodule candidates; extracting grayscale, shape and texture features of the pulmonary nodule candidates; and carrying out dimension reduction on multi-dimensional mixed features and carrying out classification by using a classifier with rules and support vector machine mixture to realize a pulmonary nodule detection effect. With the novel segmentation method and classification method, the false positive phenomenon is reduced and the detection precision of the pulmonary nodule of the medical image is improved. The pulmonary nodule detection method can be applied to the computer aided detection system.

The invention provides a three-dimensional pulmonary lobe segmentation method based on a CT image, and belongs to the field of computer aided medicine. According to the method, marker-based watershedtransformation is mainly carried out in a CT image, and the lung parenchyma is subdivided into five lung lobes. The method comprises the following steps: firstly, segmenting pulmonary parenchyma, trachea and blood vessels by adopting an automatic segmentation algorithm, and generating cost images for watershed segmentation and mark points belonging to different lung lobes by utilizing the information of pulmonary parenchyma, trachea and blood vessels; And then carrying out watershed transformation on the basis of the obtained cost image and the mark point to obtain an initial lung fissure area; And finally, combining the initial pulmonary fissure area information with information of pulmonary parenchyma, trachea and blood vessels to generate a new cost image, and performing watershed transformation again to obtain a final pulmonary lobe segmentation result.

The embodiment of the invention provides a medical treatment path navigation method and a medical treatment path navigation system. The medical treatment path navigation method comprises the following steps: at the path planning stage, determining a puncture initial point and a puncture target point leading to a lesion area path, and at the path navigation stage, by combining electromagnetic navigation, puncture and balloon dilatation, breaking through a passage leading to the lesion area via pulmonary parenchyma, thus navigating a guide catheter to the lesion area, therefore, the problem that the lesion position having no communicating obvious natural air passage can not be reached is solved, and the guide catheter can be navigated to any lesion area in the lung.

The invention discloses a pulmonary tuberculosis detection method and system based on a chest X-ray image and a storage medium, and the method comprises the following steps: obtaining the chest X-rayimage, and carrying out the preprocessing of the chest X-ray image; inputting the preprocessed chest X-ray image into a trained target detection model, and obtaining a lung parenchyma segmentation region of the preprocessed chest X-ray image; performing image segmentation on the preprocessed chest X-ray image according to the lung parenchyma segmentation region to obtain a lung region image; inputting the lung region image into a plurality of trained feature detection networks to obtain a plurality of feature images; and performing weighted combination on the feature image to obtain the category of the chest X-ray image. According to the invention, the chest radiography image is processed through the target detection model and the feature detection network, the pulmonary tuberculosis probability of the chest radiography image is predicted, and the workload of medical staff in pulmonary tuberculosis diagnosis is greatly reduced. The invention can be widely applied to the field of medical image processing.

The invention discloses a lung parenchyma CT image segmentation method based on a weighted full convolutional neural network, and belongs to the field of medical image processing. The method comprisesthe following steps: selecting a public lung data set for preprocessing, and extracting a lung parenchyma boundary in a labeled image as a semantic category; designing an improved network structure based on a standard full convolutional neural network framework, and establishing an overall structure framework of the pulmonary parenchyma segmentation convolutional neural network by using a principle that a standard path structure for encoding and decoding simultaneously comprises jump connection, expansion convolution and batch normalization; adopting a weighted loss function layer; dividing the data set; carrying out offline model training out to acquire model weight parameters; inputting a test image and outputting a segmentation result by an output layer through layer-by-layer feedforward of a network. According to an existing lung parenchyma segmentation method, a segmentation missing phenomenon is prone to occurring in a focus area in lung parenchyma, and correct segmentation of the focus area in lung parenchyma segmentation can be effectively improved through enhancement processing on important pixels.

The invention discloses a fully automatic segmentation method for lung parenchyma CT images, which is characterized in that the method comprises the following steps: 1) preliminary extraction of lungparenchyma template; 2) full automatic separation of left and right lung; 3) smooth mediastinal surface; 4) Rib surface smoothing. The invention designs a full automatic segmentation method for CT images of lung parenchyma, which can realize full automatic and accurate segmentation of CT images of lung, is helpful for assisting clinicians in diagnosing lung diseases, greatly reduces workload and operation time of doctors, and more quickly and accurately detects lung diseases. The invention also discloses an automatic segmentation method for CT images of lung parenchyma. After obtaining the preliminary pulmonary parenchyma template, the invention carries out the full automatic separation of the left and right lungs, smoothes the mediastinal surface and the costal surface, ensures the accuracy, integrity and robustness of the segmentation result, and assists the clinician in diagnosing lung diseases; as that whole process of the invention does not need manual calibration and parameter setting, the universality and universality of the lung parenchyma CT image segmentation method are ensured.

A pulmonary nodule auxiliary diagnosis system based on an adversarial network and a Faster R-CNN belongs to the technical field of medical instruments and comprises a data acquisition module used foracquiring CT image data of the lung of a patient, a pulmonary parenchyma segmentation module used for segmenting to obtain complete pulmonary parenchyma, and a pulmonary nodule detection network constructed based on a deep learning model. The pulmonary parenchyma segmentation image is input into the pulmonary nodule detection network for nodule detection and identification; and qualitative and quantitative data of the pulmonary nodules are determined automatically, the qualitative and quantitative data is displayed to a user, and finally an image-text diagnosis report is given to assist a doctor in diagnosis. According to the invention, accurate detection and diagnosis of tiny pulmonary nodules can be carried out, the film reading problem can be more rapidly, accurately and scientificallysolved for doctors, and imaging support is provided for clinical auxiliary decision making of lung cancer.

The invention provides a pulmonary nodule image deep learning identification system based on an RGB channel overlay method and a method thereof. The system comprises a data reading module, a pulmonary parenchyma extracting module, a pulmonary nodule extracting module, an RGB channel overlay module, a sample training module and a prediction module. According to the invention, pre-processing operation, such as RGB channel overlay, is performed on original data, and a great amount of training set data are trained by using a deep learning method, and then accurate identification of images is achieved via a prediction model obtained after training.

The invention discloses a pulmonary nodule detection method based on a convolutional neural network, and belongs to the field of deep learning image detection. After the pulmonary parenchyma is extracted, the attention of the neural network to the target area can be effectively improved, and the number of parameters can be effectively reduced. After the step of preliminary candidate nodule detection is completed, the target area close to the center of the pulmonary nodule is selected for nodule detection, the overall detection range is narrowed, a relatively accurate classification effect is obtained, and the classification performance is improved.

The invention relates to a pulmonary nodule automatic detection method, a pulmonary nodule automatic detection device and a computer system. The method comprises the following steps: acquiring a CT image to be detected; performing filtering enhancement processing on the CT image to be detected to obtain a lung enhanced CT image sequence; segmenting the CT image sequence by adopting a threshold method to obtain an image only containing a pulmonary parenchyma region; cutting the image obtained by the pulmonary parenchyma segmentation module into a plurality of image blocks, and obtaining a region of interest through a multi-scale feature fusion UNet network model; and carrying out automatic detection and identification on the region of interest by adopting a 3D CNN model to obtain a pulmonary nodule detection result. Compared with the prior art, the method has the advantages of high detection sensitivity and precision and the like.

The invention belongs to the technical field of medical diagnosis, and discloses a processing system and an information processing method for distinguishing pulmonary tuberculosis and tumor information. The processing system for distinguishing pulmonary tuberculosis and tumor information comprises a patient information acquisition module, a lung image acquisition module, a central control module,a lung image enhancement module, an image segmentation module, an image feature extraction module, an image retrieval module, a comparison module, a disease analysis module, a treatment scheme writingmodule, a diagnosis report generation module and a display module. Interference factors such as a human body trunk and a bed board are automatically removed through the image segmentation module, sothat a pulmonary parenchyma image can be rapidly and accurately extracted to better assist a doctor; meanwhile, through the image retrieval module, the efficiency of searching for the similar lung images by the doctor user is greatly improved, corresponding retrieval feature vectors are obtained according to different focus types contained in the to-be-retrieved images, similar sample lung imagesare retrieved based on the retrieval feature vectors, and therefore the lung image retrieval accuracy is improved.

The invention provides a CT image pulmonary parenchyma template weasand elimination method based on breadth-first search. The method comprises the steps that (1) traversal starts from traversal pixels in a pulmonary parenchyma template; (2) whether the traversal is completed or not is determined, and if so, the step (8) is carried out, otherwise the step (3) is carried out; (3) a next traversal pixel is acquired; (4) whether the pixel is traversed is determined, and if so, the step (2) is carried out, otherwise the step (5) is carried out; (5) whether the pixel value is 1 is determined, and if so, the step (6) is carried out, otherwise the step (7) is carried out; (6) in a communication area with the breadth-first search pixel value of 1, the step (2) is carried out; (7) the point is marked as traversed, and then the step (2) is carried out; (8) whether the number of communication areas is three is determined, and if so, the step (9) is carried out, otherwise a final pulmonary parenchyma template is acquired; and (9) the pixel value of all pixels in the smallest communication area is set to 0 to acquire the final pulmonary parenchyma template.

The invention discloses a three-dimensional lung feature extraction method based on a CT image. The method comprises the steps of performing lung segmentation on a lung CT image to obtain lung parenchyma; performing tracheal segmentation on the pulmonary parenchyma to obtain a tracheal tree; performing bifurcation point detection on the tracheal tree to obtain a bifurcation feature point set; extracting contour points through Canny edge detection, and merging bifurcation feature points. Tracheal tree bifurcation points and lung contour points are more firmly recorded by application generated by respiration, and better data are provided for feature point set registration.

The invention relates to a pulmonary marker, used for marking the lesion position of the lung. The pulmonary marker comprises a self-expanding structure which can radially expand automatically after being released from a compression device so as to realize marking. According to the pulmonary marker, due to the adoption of the self-expanding structure, when the lesion position is found during the lung examination, the pulmonary marker can be conveyed to the lesion position through a catheter or a conveyor, then the self-expanding structure is extruded and fixed by the pulmonary parenchyma after expanding at the lesion position, therefore, during the surgical operation, a doctor can rapidly locate the lesion position, thus the surgical time is shortened, the excision extension is reduced, and the postoperative quality of life of a patient is improved.

The invention relates to an image sag repairing method based on a double-Graham scanning method. The method comprises the following steps that: utilizing a boundary tracing method for a lung mask subjected to preliminary segmentation to solve the boundary point set P of a left lung and a right lung; utilizing the Graham scanning method to obtain the salient point set Q of the left lung and the right lung; solving the lung sag point to be repaired; removing the solved lung sag point to be repaired from the boundary point set P, wherein the obtained residual point set is the edge point set of the pulmonary parenchyma; detecting whether the segmented pulmonary parenchyma contains breakage or not, and if the segmented pulmonary parenchyma contains breakage, carrying out breakage connection; and finally, obtaining an integral segmented pulmonary parenchyma segmentation result. After sag repair and breakage connection are carried out, a final integral pulmonary parenchyma segmentation resultis obtained. The sag on the outer wall of the lung can be repaired, and the method performs a good repairing and smoothening function on adjacent sags including a heart, a mediastinum and the like between two pieces of pulmonary parenchyma and can provide a correct and reliable basis for subsequent pathological study.

The invention provides a lung tissue dissimilation degree judgment method and device, and the method comprises the following steps: carrying out the pulmonary parenchyma segmentation of all lung CT images, and generating a pulmonary parenchyma segmentation image; performing binarization processing on the pulmonary parenchyma segmentation image to generate a pulmonary parenchyma mask image; carrying out feature enhancement processing on pulmonary vessels in all pulmonary parenchyma segmentation images to generate pulmonary vessel feature enhancement images; carrying out binarization processingon the pulmonary vessel feature enhancement image to generate a pulmonary vessel mask image; counting the number L of pulmonary parenchyma region pixel points in all pulmonary parenchyma mask images to serve as volume parameters of the pulmonary parenchyma region pixel points; counting the number V of pulmonary vessel region pixel points in all the pulmonary vessel mask images as volume parametersof the pulmonary vessel region pixel points; and based on the number L of the pulmonary parenchyma region pixel points in all the pulmonary parenchyma mask images and the number V of the pulmonary vessel region pixel points in all the pulmonary vessel mask images, calculating to obtain a pulmonary effective ventilation function region proportion ELVAR value, and outputting the ELVAR value. The method and the device provided by the invention provide a reliable basis for clinical related lung condition evaluation.

The invention relates to a lung parenchyma segmentation method for extracting a CT image based on a clustering key frame, and the method comprises the following steps: carrying out the preprocessing of lung CT image data, and carrying out the lung window windowing gray scale transformation of an image CT value; performing clustering analysis on the grey level histogram similarity of the patient lung CT image processed in the step 1, and extracting key frame data in the lung CT sequence; carrying out pulmonary parenchyma segmentation on the patient key frame CT image by adopting a multi-phase level set CV model; performing lung parenchyma mapping segmentation extraction in the lung CT complete sequence according to a lung parenchyma segmentation result in the key frame to obtain a lung parenchyma initial contour; and performing morphological corrosion and expansion operation on the lung parenchyma initial contour in the lung CT image to refine the contour, and finally obtaining a lung parenchyma segmentation result in the patient lung CT data. The invention also provides lung parenchyma segmentation equipment for realizing the segmentation method.

The invention discloses a pulmonary nodule detection method which comprises the following steps: acquiring an original data set, and preprocessing the original data set; constructing a neural networksegmentation model by adopting a Residual Block and a loss function; constructing a pulmonary nodule detection model by adopting a Faster RCNN algorithm; training and testing a neural network segmentation model and a pulmonary nodule detection model by using the preprocessed original data set; inputting a to-be-detected image to the trained and tested neural network segmentation model and the pulmonary nodule detection model; performing lung parenchyma segmentation on the to-be-detected image by using the neural network segmentation model to obtain a lung parenchyma segmentation image; performing nodule detection on the pulmonary parenchyma segmentation image by using a pulmonary nodule detection model, and outputting a pulmonary nodule candidate region; and eliminating a non-nodule regionof the pulmonary nodule candidate region to obtain a pulmonary nodule detection result. According to the invention, pulmonary parenchyma segmentation and pulmonary nodule detection are respectively carried out based on the neural network segmentation model and the pulmonary nodule detection model, so that the detection precision and efficiency are greatly improved.

Methods and devices to improve lung function in a patient having restricted ventilation. The device may include an implantable airway bypass device that relieves trapped air. The method may include a treatment procedure that minimizes irritation of tissue to control healing processes. Systems, methods and devices have been conceived and are disclosed herein for improving the mechanics of a diseased lung of a patient by implanting one or more natural airway bypass ventilation devices in a lung. For example, the patient may suffer from COPD, emphysema, chronic bronchitis, or asthma. A natural airway bypass device may comprise a pressure relief device connecting lung parenchyma distal to abnormally high resistance airways to the atmosphere.