60 results about "Pulmonary tumor" patented technology

The invention provides a lung tumor recognition system and method based on a convolutional neural network. The lung tumor recognition method based on the convolutional neural network includes the steps that multiple cascading images collected along a preset dimensionality are received; the cascading images are blocked according to the rule that moving is the same in the same window, the image blocks in the cascading images corresponding to the same window position are combined and convolved to obtain feature image blocks, and down-sampling is conducted on each feature image block; the feature image blocks subjected to down-sampling are subjected to up-sampling; a heat degree prediction map is drawn on the basis of probability that each pixel in the heat degree prediction map corresponds to at least one feature image block subjected to up-sampling may be true or false. The system and method effectively increase the recognition rate of local areas of the cascading images with spatial association.

The invention relates to a dry powder inhalation with low-density or ultralow-density inhalable medicinal powder, a method for preparing the dry powder inhalation and application of the dry powder inhalation in preparing medicament for treating tumor, in particular lung tumor. The dry powder inhalation can deliver the tumor medicament to the lung directly so as to improve the medicament effect andreduce side effect; and the 'foaming' technology adopted for preparing the dry powder inhalation optimizes the micromeritic properties of the medicinal powder, and greatly improves the deposition performance of the tumor medicinal dry powder inhalation. The dry powder inhalation, the preparation method and the application thereof are also suitable for preparing non-vector dry powder inhalation toovercome the defects of uneven mixture between medicament and vector easily occurring in the dry powder inhalation.

The invention relates to a novel lung tumor locating needle. The novel lung tumor locating needle comprises a puncture needle, anchoring locating needles, a locating line and a pushing device, wherein the puncture needle is composed of a puncture needle cannula and a puncture needle handle, the pushing device comprises a pushing tube and a pushing tube handle, the pushing tube is arranged in the puncture needle cannula, the far end of the pushing tube abuts against the near ends of the anchoring locating needles, the width of the near ends of the anchoring locating needles is larger than or equal to the outer diameter of the pushing tube, the far end of the locating line is connected with the near ends of the anchoring locating needles, the locating line extends out of the exterior of the pushing device from the interior of a tube cavity of the pushing tube, the far end of the locating line is provided with multiple colored bands with different colors, and the length of each colored band ranges from 3 mm to 15 mm. According to the novel lung tumor locating needle, the functions of determining lesion localization and measuring the depths of lesions are achieved simultaneously, the lung or parts of the lung can be lifted up, excision of the lesions is convenient for a surgeon, and the depths of the lesions can be measured during an operation.

The present invention provides molecules that selectively home to various normal organs or tissues, including to lung, pancreas, skin, retina, prostate, ovary, lymph node, adrenal gland, liver or gut; and provides molecules that selectively home to tumor bearing organs or tissues, including to pancreas bearing a pancreatic tumor or to lung bearing a lung tumor. The invention also provides conjugates, comprising an organ or tissue homing molecule linked to a moiety. Such a moiety can be, for example, a therapeutic agent or a detectable agent. In addition, the invention provides methods of using an organ homing molecule of the invention to identify a particular organ or tissue by contacting the organ or tissue with a molecule of the invention. The invention also provides methods to diagnose or treat a pathology of the lung, pancreas, skin, retina, prostate, ovary, lymph node, adrenal gland, liver or gut by administering to a subject having or suspected of having a pathology a molecule that homes to the selected organ or tissue. The invention further provides methods of identifying a target molecule in lung, pancreas, skin, retina, prostate, ovary, lymph node, adrenal gland, liver or gut.

The invention relates to a diagnosis model constructed based on artificial intelligence fusion multi-modal information and used for various pathology types of benign and malignant pulmonary nodules. The method comprises the steps: constructing a multi-resolution 3D multi-classification deep learning network model; constructing a machine learning multi-classification model; training the constructedmulti-resolution 3D deep learning model by using CT image, and obtaining a weight; training the constructed machine learning multi-classification model by using lung tumor marker information, and obtaining a weight; and fusing the lung CT imaging information and the lung tumor marker information at the tail end of the model by migrating the weights of the deep learning network and the machine learning network model through migration learning. According to the invention, a deep learning network model and a machine learning model are adopted to respectively mine deep features related to pathological type classification in a lung CT image and a lung tumor marker, and the fusion of the CT image and the lung tumor marker multi-modal information is realized by fusing the two network models to rapidly diagnose the specific pathology type of the pulmonary nodule.

A procedure and an apparatus are described for identifying individuals at risk of pulmonary tumour and/or for diagnosing a pulmonary tumour using the study of levels of expression of miRNA in the blood or another biological fluid. Also described are a method and a compound for reducing or eliminating a risk of pulmonary tumour by rebalancing the miRNAs that are underexpressed or overexpressed.

The invention discloses a dense neural network lung tumor image recognition method fusing multi-scale features, and the method comprises the steps: collecting and preprocessing CT modal medical images, extracting lesion ROI regions of different scales, and forming a multi-scale data set; wherein the focus ROI regions of different scales are provided with clinically marked benign or malignant tumortags; training the multi-scale data set in a dense neural network, constructing a dense neural network model, extracting a full connection layer feature vector and carrying out feature serial fusion;and obtaining a lung tumor classification result in the NSCR classifier. A dense neural network model constructed by the method is superior to an AlexNet model, bottom-layer new features can be minedagain by effectively utilizing high-layer information, transmission of the features among networks is enhanced, and feature reuse is realized and enhanced; the invention is deep in network depth, strong in network generalization capability and high in lung tumor classification accuracy.

The invention provides nucleic acids, peptides, and antibodies for use in applications including diagnosis and therapy. The peptides target lung cancer and were identified by phage display. Targeting phage PC5-2 and synthetic peptide SP5-2 were both able to recognize human pulmonary tumor specimens from lung cancer patients. In SCID mice bearing NSCLC xenografts, the targeting phage was able to target tumor masses specifically. When the peptide was coupled to liposomes containing the anti-cancer drugs vinorelbine or doxorubicin, the efficacy of these drugs against human lung cancer xenografts was improved, the survival rate increased, and the drug toxicity was reduced.

The invention provides a culture medium for lung tumor cells. The culture medium is particularly suitable for in-vitro three-dimensional culture of lung tumor cells, particularly circulating tumor cells from blood of lung cancer patients. The invention also provides a method for in-vitro three-dimensional culture of the circulating tumor cells to form organoids.

The invention provides a semi-automatic segmentation method for sketching a regional growth lung tumor radiotherapy target region, and the method specifically comprises the following steps of S1, lesion automatic initial seed point selection characterized by obtaining lesion seed points from a lung parenchyma gradient image containing lesion through the step, and obtaining a lung parenchyma CT image gradient minimum value through calculation; and S2, final lesion position refinement characterized by carrying out the refinement of the final lesion position by determining a more accurate lesionboundary definition. By using the method provided by the invention, the steady, efficient and accurate lung lesion segmentation can be automatically realized.

The method is suitable for the technical field of medical image processing. The invention provides a lung tumor automatic sketching method based on deep learning, and the method comprises the steps: obtaining an input lung CT image when a sketching request of the lung CT image is received, carrying out the preprocessing and image enhancement of the obtained lung CT image, and obtaining a corresponding processed image; obtaining the position and size of a window of the lung tumor in the image, and cutting the screened image into a fixed size; inputting the processed image into a trained V-Net model so as to predict the lung tumor; performing deconvolution on the predicted tumor image to the size of the cut image, so that real prediction of the tumor can be obtained; and extracting the edgeline of the truly predicted lung tumor, that is, sketching the lung tumor, and obtaining an image sketched by the lung tumor. According to the lung tumor automatic sketching device, the accuracy of automatic sketching of lung tumors is improved, the sketching efficiency is remarkably improved on the basis of guaranteeing the sketching precision, and the operation safety process is improved.

The invention discloses a high-dimensional feature selection algorithm based on a Bayesian rough set and a cuckoo algorithm, and the algorithm comprises the steps: obtaining a lung tumor image, carrying out the target contour segmentation, and obtaining a segmented ROI image; extracting a high-dimensional feature component of the segmented ROI image, and constructing a decision information table containing feature attributes based on the feature component; and reducing the original feature space by adopting a BRSGA algorithm to obtain an optimal feature subset, optimizing a penalty factor anda kernel function parameter of the SVM by utilizing a CS algorithm, and inputting the reduced feature subset into the optimized SVM to obtain a classification and identification result. According to the method, the optimal feature subset is generated through the genetic algorithm and the BRS, the feature dimension is reduced on the premise that the classification accuracy is not reduced, the constraint of manual parameter setting is eliminated, and the time consumption is reduced. According to the method, global optimization is carried out on SVM parameters through CS, search space can be explored more effectively, population diversity is enriched, and good robustness and high global search capacity are achieved.

The present invention provides a method of determining the presence of a pulmonary tumor in a subject is provided. Also provided is a method of determining the presence of an aggressive pulmonary tumor in a subject. Additionally, a method of determining the risk of manifesting a pulmonary tumor in a subject is provided. Further provided is a method of determining the risk of manifesting an aggressive pulmonary tumor in a subject. A method for predicting the risk of developing or having a pulmonary tumor in a subject is also provided. A method of establishing lung cancer treatment options is additionally provided.

The invention discloses a lung neoplasm diagnosis and treatment device under radiography guidance. The lung neoplasm diagnosis and treatment device comprises a CT reconstruction device and a CT examining couch; the CT reconstruction device comprises a detector, a light pipe, a 360 degree rotating loading platform movement system, a detector Y direction linear movement system, a detector X direction linear movement system, a detector Z direction linear movement system, a light tube Z direction linear movement and detected objects; the CT examining couch comprises a couch body, a headrest, a head rest guide rail, a head rest telescopic rod, an arm fixator, an arm fixator telescopic rod, a motor, a castor, a castor lock, couch legs, a couch leg height adjuster, a head rest telescopic rod control button, an arm fixator telescopic rod control button, a couch leg height adjuster control button, a power switch and a power interface. The lung neoplasm diagnosis and treatment device under radiography guidance makes the CT examining couch operation more easy and convenient, achieves electric control type adjustment, is convenient to move, improves the use performance of the CT examining couch, and solves the problem that the conventional CT examining way is not suitable for examining flat components.

The invention provides a medicine for treating tumors, which is prepared from paris polyphylla, radix aconiti, radix aconiti praeparata, oldenlandia diffusa, dandelion, honeysuckle, centipede, tianlong, ginseng, ganoderma lucidum, astragalus root, coptis root, angelica and the like according to proportioning. As far as the medicine for treating the tumors is concerned, the synthetic action of each medicine can be utilized to treat the tumors, the physical sign can be improved, the autoimmunity can be enhanced, the curative effect is good, application is convenient, no toxic or side effect exists and treatment cost is low; clinically used by 300 patients, the medicine enjoys a total effective rate of 90% and can be mainly used for treating the tumors of the digestive system and the pulmonary tumors,.

The invention discloses a bronchoscope image feature comparison marking system based on deep learning, and the system comprises: an input layer which is used for receiving a bronchoscope image collected by a bronchoscope, carrying out the recognition of the received bronchoscope image, and transmitting a recognized in-vivo image to a judgment layer; a judgment layer which is used for recognizing the in-vivo image from the input layer, judging a part corresponding to the in-vivo image, judging whether a lesion part exists or not and judging whether the lesion part is benign or malignant or not, and then feeding back a recognition result to the output layer; and an output layer which is used for displaying the identification result from the judgment layer. The invention further discloses a bronchoscope image feature comparison marking method based on deep learning. According to the invention, the central lung tumor focus part under the bronchoscope can be identified by carrying out feature comparison on the bronchoscope image.

The invention relates to a lung tumor sketching method based on a 5G cloud radiotherapy private network, and the method comprises the steps: collecting a preset number of CT chest images and T1 MRI chest images, transmitting the CT chest images and the T1 MRI chest images to an edge cloud iMEP through a private network gateway iGW, and obtaining a lung tumor segmentation neural network model; after the to-be-segmented image is sent to the edge cloud iMEP through the private network gateway iGW, inputting the lung tumor segmentation neural network model to obtain a lung tumor automatic segmentation result of the CT image, and uploading the lung tumor automatic segmentation result to the 5G cloud network; patient related information corresponding to the to-be-segmented image recorded by the common terminal is sent to the 5G cloud network through a private network gateway iGW; and logging in the 5G cloud network to fuse the lung tumor automatic segmentation result and the corresponding patient related information, and carrying out post-processing and edge detection on the lung tumor automatic segmentation result to obtain a lung tumor sketching result.