93results about How to "Reduce missed diagnosis" patented technology

The invention discloses a vision field definition enhancement system for gastrointestinal endoscope diagnosis and treatment (ESCGV). A liquid delivery pipe (2) is led out from a washing bottle (1), and passes by a first peristaltic pump (3); and the washing bottle (1) is also provided with a stirring and heating sub-system for stirring and heating the washing liquor therein. The invention also discloses a vision field definition enhancement method for gastrointestinal endoscope diagnosis and treatment. The method comprises the following steps: 1) adding the washing liquor into the washing bottle; 2) magnetically stirring the washing liquor in the washing bottle, and simultaneously heating the washing liquor; 3) setting the temperature of the washing liquor at 25-37 DEG C; and 4) sending the washing liquor into an observation window of a gastrointestinal endoscope. The system and method can obviously enhance the vision field definition under the gastrointestinal endoscope, make the operation convenient and simple, greatly reduce the missed diagnosis and erroneous diagnosis, and obviously increase the treatment quality of the gastrointestinal endoscope.

The invention provides a capsule endoscope system capable of obtaining a real-time position and posture and a working method of the capsule endoscope system. The working method comprises the following steps: S1, collecting data of an inertial sensor from radio frequency communication equipment of a capsule endoscope, and carrying out operation processing on the obtained data on a remote server; carrying out data noise reduction through the remote server, and removing irregular sensor data; S2, carrying out position and posture solving according to accelerated velocity and angular velocity data obtained by an MEMS (Micro-electromechanical Systems) inertial sensor of the capsule endoscope, so as to obtain real-time position and posture information of the capsule endoscope; S3, carrying out error compensation on the obtained real-time position and posture information, and adjusting the obtained position and posture information values through operation, so as to enable positioning and posture determination of the capsule endoscope to be more accurate.

The invention discloses a circuit breaker voiceprint and vibration entropy feature fusion fault diagnosis method. The method comprises the steps of carrying out band-pass filtering on a sound signal,and obtaining the roughness, box dimension, direction degree, contrast ratio, linearity and regularity which reflect time-frequency texture features through generalized S transformation; constructingvoiceprint features based on the roughness, the box dimension, the direction degree, the contrast ratio, the linearity and the regularity which reflect the time-frequency texture features; selecting parameters for a vibration signal, and obtaining permutation entropy through variational mode decomposition; constructing vibration entropy features based on the permutation entropy; performing deep fusion based on the voiceprint features and the vibration entropy features to obtain a joint feature vector; performing learning training of an SVM model on the joint feature vector; and based on the GWO-SVM diagnosis model, optimizing a penalty factor and a mixing coefficient in the SVM model by adopting the GWO model. In the embodiment of the invention, the influence of sensor frequency response and environmental noise on the fault diagnosis of the circuit breaker is reduced.

The invention discloses a thyroid nodule diagnosis method based on a deep learning network, and belongs to the field of image processing and artificial intelligence aided disease diagnosis. The methodcomprises the following steps: searching an ultrasonic original image and a pathological report of a thyroid nodule of a thyroid patient, and constructing a thyroid nodule database; preprocessing theultrasonic image; performing semantic segmentation on the ultrasonic image preprocessed in the step 2 through a Deeplab v3+ method based on Xception-JFT, and forming a semantic segmentation result graph; judging benign and malignant thyroid nodules based on a deep learning network; and forming a thyroid nodule diagnosis information report. According to the method, the Deeplab v3+ algorithm basedon Xeption-JFT is adopted to establish the thyroid ultrasound image segmentation network model, the optimal segmentation effect is achieved by continuously improving the backbone network Xception, nodule information can be automatically and rapidly recognized under high accuracy and high robustness, image features are automatically extracted for accurate segmentation to obtain a better diagnosis result, and an objective reference is provided for clinical diagnosis.

The invention discloses a hepatic echinococcosis focus segmentation method and a hepatic echinococcosis focus segmentation system based on a neural network. The method comprises the following steps: S1, training and verifying a cystic echinococcosis focus segmentation model; S2, training and verifying a follicular echinococcosis focus segmentation model; S3, obtaining a segmented liver region fromthe one-pack worm CT image, and inputting the liver region into the focus recognition model to obtain a recognition result; S4, when it is determined that the recognition result is the cystic echinococcosis focus, inputting the VOI region into the cystic echinococcosis focus segmentation model to obtain a first segmentation result; And S5, when it is determined that the recognition result is thefollicular echinococcosis lesion, performing blood vessel recognition and segmentation on the VOI region, and inputting the blood vessel segmentation result and the VOI region into the follicular echinococcosis lesion segmentation model to obtain a second segmentation result. According to the method and the system provided by the invention, fusion recognition and feature extraction are carried outon the multi-modal medical image through various models, a doctor is assisted to carry out echinococcosis screening work, and the diagnosis efficiency and accuracy are improved.

The invention discloses an oral panoramic film decayed tooth depth recognition method based on deep learning. The method comprises the following 5 steps: preprocessing data; segmenting the image; constructing a model; performing feature extraction; and performing classification and identification. According to the method, the convolutional neural network is adopted to identify the decayed teeth, automatic identification of the decayed teeth in the oral cavity panoramic film is realized, and result export of automatic decayed tooth depth identification can be realized. Visual fatigue easily caused by long-term viewing of image data by medical staff is solved, diagnosis and treatment capacity and working efficiency are improved in an assisted mode, medical resource requirements are reduced,and medical efficiency is improved.

The invention is suitable for the technical field of computers, and provides an endoscope system, an endoscope image recognition method and equipment, and a storable medium, and the endoscope system comprises an image receiving unit which is used for receiving a to-be-recognized endoscope image, an image processing unit which is used for extracting biological characteristic information from the endoscopic image to be identified and generating a lesion part recognition result according to the biological feature information and a biological feature-lesion image recognition model generated by training a plurality of pre-collected endoscope image samples carrying lesion part information through a convolutional neural network, and a display unit which is used for identifying and displaying thelesion part in the to-be-identified endoscope image according to the lesion part identification result. Compared with an existing system which directly carries out lesion information analysis according to image apparent characteristics, the system is higher in recognition precision and efficiency, can be suitable for lesion recognition of endoscope images with different illumination and shooting angles, and greatly reduces the probability of missed diagnosis or misdiagnosis.

The invention provides an image enhancement method and device. The image enhancement method comprises the steps that an image is decomposed into L layers based on first transformation; the image is decomposed into L+N layers based on second transformation, and N>=1; and the Lth layer low-frequency image obtained through decomposition based on second transformation and the first to the Lth layers of high-frequency images obtained through decomposition based on first transformation are reconstructed. According to the technical scheme, the contrast of the image is enhanced to a large extent so that the quality of the image is enhanced.

The invention discloses a detecting device for Brucella infection. The detecting device comprises a test strip; a Brucella antigen detection area and a Brucella antibody detection area are arranged on a chromatographic membrane of the test strip. According to the detection device, a Brucella antigen and a Brucella antibody in a sample can be detected simultaneously; various outer membrane proteins are applied in a combined manner, so that the occurrence of false positive results can be reduced effectively; a double-antigen sandwich method and a double-antibody sandwich method are adopted, so that the sensitivity and the specificity of detection can be higher than those obtained according to the conventional indirect method, and the rapid detection of brucellosis is realized; the problems of high interference, and high false positive result or false negative result occurrence rate caused by long detection period, poor specificity and low sensitivity in the conventional brucellosis detection process are solved.

The invention a tuberculosis detection model construction method and applications. The method includes the following steps: A, obtaining chest X-ray images labeled with tuberculosis focus area labeling frames and with assigned quantities; B, performing image preprocessing on the chest X-ray images so that preprocessed image data can be obtained; and C, selecting a basic convolutional neural network model, and inputting the preprocessed image data into the basic convolutional neural network model to perform training so that a trained tuberculosis detection model can be obtained, wherein convolution kernels with different shapes can be adopted according to different anchor sizes in the basic convolutional neural network model. Thus, tuberculosis focus areas can be accurately predicted through the model, so that missed diagnosis and misdiagnosis rates can be reduced.

The invention discloses a human echinococcosis colloidal gold immunochromatographic assay urine testing quick diagnosis test paper card. The test paper card comprises a plastic card box and a detection module in the card box; the plastic card box is provided a sample loading hole and an observation window; the detection module comprises sample filtering paper, a nitric acid fibrous membrane and absorbent paper; echinococcosis resistance fluid antigen humanized single chain variable region antibody which is marked by colloidal gold is absorbed on the absorbent paper and nitric acid fibrous membrane; a quality control line and a detection line are marked on the echinococcosis resistance fluid antigen humanized single chain variable region antibody marked by the colloidal gold on nitric acid fibrous membrane; the sample filtering paper is opposite to the sample loading hole; and the nitric acid fibrous membrane is opposite to the observation window. The test paper card has the advantages of simplicity and convenience, fastness, no invasiveness, reliability and stability.

The invention discloses an artificial intelligence assisted prostate tumor early diagnosis method based on surface enhanced Raman spectroscopy. The method comprises the following steps of: collectinga serum sample; adopting a Raman spectrometer for detection; performing graphic analysis and data processing on an acquired Raman spectrum signal; selecting a biological tissue Raman spectrum concentration range of 400-1800cm<-1> for analysis, and preprocessing an original spectrum acquired by a spectrograph through Origin Pro 8 software, i.e. performing background reduction and spectrum area homogenization, drawing a serum average Raman spectrum, and then comparing changes and differences of related Raman characteristic peaks, thus finally obtaining a diagnosis result of prostate tumor earlydiagnosis. The method provided by the invention is higher in sensitivity and specificity, and unnecessary needle biopsy and missed diagnosis caused by an existing screening diagnosis method are reduced.

The invention provides a puncture forceps holder biopsy needle used beneath a bronchoscope. The puncture forceps holder biopsy needle is composed of an outer sleeve tube, a puncture needle, biopsy forceps, a tensioning steel wire, a connecting steel wire, and a sliding block, a helical casing and a helical connector which correspondingly control the puncture needle and the biopsy forceps, a sliding block and a handle which control the tensioning steel wire and the like. In the outer sleeve tube of the puncture forceps holder biopsy needle, the puncture needle is connected with a puncture needle sliding block through the connecting steel wire, the biopsy forceps are located in the puncture needle and connected with the biopsy forceps sliding block through the tensioning steel wire. According to the puncture forceps holder biopsy needle, transbronchial needle aspiration and forcep biopsy can be achieved on the same puncture part, the quantity of obtained samples can be increased, the diagnostic rate of mediastinal lesions is increased, meanwhile the discomfort of a patient is greatly reduced, and the safety of the patient is improved.

The invention discloses a dynamic mammary gland ultrasonic video whole focus real-time detection and segmentation system based on artificial intelligence. The system at least comprises at least one ultrasonic machine and an AI server. Wherein an AI-based dynamic mammary gland ultrasonic video detection and segmentation device is arranged in the AI server, and the device at least comprises (1) a system robustness design module, (2) a data preprocessing module, (3) a data amplification module, (4) a focus detection module and (5) a focus segmentation module. According to the system, on the premise that an ultrasonic machine and an existing diagnosis process are not changed, automatic real-time detection on all focuses involved in the dynamic breast ultrasonic video image can be achieved, thedetected focuses are intelligently segmented and measured, and a doctor can be effectively helped to reduce missed diagnosis while the efficiency and the accuracy are improved.

The invention discloses an anti-keratin antibody detection method and an anti-keratin antibody regent box, and belongs to the method for detecting the characteristics of blood in body. The method provided by the invention can be operated as follows: the esophagus of a bandicoot can be made into a biological thin slice which can be taken as an antigen and coated in the reaction region of a piece of slide glass; blood serum to-be-detected is added in the reaction region and then added with complement after the incubation, rinsing and drying in a spanning way, then added with the fluorescent markers of anti-complement antibodies after the incubation, rinsing and drying in a spanning way, and then observed under a fluorescence microscope after the incubation, rinsing and drying in a spanning way; if the horny layer appears typical and regular line-shaped or lamellar fluorescence, the detection result is positive. The regent box provided by the utility model is filled with regents, cover glass and slide glass with biological thin slices, such as negative control serum, positive control serum, condensed phosphate buffer solution and the application solution of the fluorescent markers of complement and anti-complement antibodies. The method provided by the invention is better than the prior method in indicators such as sensitivity, specificity, negative predictive value and positive predicative value, therefore, the method has higher diagnosis value in diagnosing rheumatoid diseases.

The present invention provides an auxiliary diagnosis system for abnormal skin symptoms. The auxiliary diagnosis system for the abnormal skin symptoms comprises an image acquisition unit, an image marking unit, an image enhancement unit, an image segmentation unit, a feature extraction unit and an image classification unit. Based on the same data set, clinical symptom feature pictures and corresponding description and labels thereof are used to build a model on different diseases, according to different performance indicators, existing abnormal skin symptoms and classification methods are compared, finally, an optimal machine learning method is obtained for diagnosis of the abnormal skin symptoms, and besides, classification output probability of target images is obtained. Thus, multi-directional and multi-dimensional comparisons are performed to intelligently derive the specific skin diseases, and the auxiliary diagnosis system is high in accuracy, maximally reduces risks of missed diagnosis and misdiagnosis, reasonably allocates medical resources, and also reduces unnecessary medical procedures.

The invention discloses a system for identifying and positioning pancreatic neuroendocrine tumors under an ultrasonic endoscope. The system comprises an image acquisition module, a training set manufacturing module, an auxiliary diagnosis module and a joint judgment module; the training set manufacturing module is configured to mark pancreatic neuroendocrine tumor regions in a sample set image byusing a multi-target marking tool; the auxiliary diagnosis module is configured to construct an auxiliary diagnosis model, and is used for performing identification of pancreatic neuroendocrine tumorlesion regions on an input pretreated image after optimization training of a training set; and the joint judgment module is configured to display an output result in a form of a color bar pattern, andis used for judging accuracy of the identification result of the pancreatic neuroendocrine tumor lesion regions. Through intelligent and automatic identification of pancreatic neuroendocrine tumors (PNET) under an ultrasonic endoscope, the PNET can be accurately identified and positioned in a large number of generated ultrasonic endoscope pictures in an ultrasonic endoscope examination process, so that the detection rate of the PNET is improved, and missed diagnosis is reduced.

The present invention discloses a field rapid evaluation method for pedestrian head closed injuries based on an improved evaluation index. The method comprises: S1, screening a damage index which is used for evaluating head closed injuries in a finite element simulation, and a vehicle-pedestrian accident parameter which is used for evaluating head closed injuries in an accident field investigation; S2, calculating pedestrian head linear acceleration a, head rotational acceleration Alpha, and a skull impact force Fc in a finite element simulation orthogonal experiment; S3, according to the damage index chosen in S1, accquring a weight coefficient of an index according to a head closed injuries epidemiology investigation to form a comprehensive evaluation index of pedestrian head closed injuries; S4, according to regression analysis on head closed injuries epidemiology data, acquiring an impact factor function of damage evaluation; and S5, investigating and acquiring the vehicle-pedestrian accident parameter chosen in S1, calculating an injury degree probability of the head closed injuries, and grasping an injury severity degree of the pedestrian head closed injuries, so as to rapidly perform corresponding field emergency treatment according to pedestrian head injuries conditions.

The invention provides a gastroscope image intelligent processing method and a gastroscope image intelligent processing device. The gastroscope image intelligent processing method comprises the steps of: dynamically acquiring a first gastroscope image; segmenting a segmented region having lesion characteristics from the first gastroscope image, labeling primary characteristic information and position information corresponding to the segmented region on the first gastroscope image to serve as second gastroscope images, and outputting the second gastroscope images; receiving the second gastroscope images, performing combinatory analysis on the primary characteristic information and the position information in a plurality of the second gastroscope images to form secondary characteristic information and regional position information corresponding to the secondary characteristic information, and outputting a third gastroscope image labeled with the secondary characteristic information and the regional position information; and receiving the third gastroscope image and displaying the same. The gastroscope image intelligent processing method and the gastroscope image intelligent processing device can assist the doctor in diagnosis, obtain a pathological suggested region, and significantly reduce the probability of missed diagnosis and misdiagnosis.

The invention belongs to the technical field of medical early warning, and discloses a pulmonary tuberculosis patient early warning system and early warning method based on electronic medical record data. The pulmonary tuberculosis patient early warning system based on electronic medical record data comprises a patient disease data acquisition module, a data preprocessing module, an electronic medical record generation module, a main control module, a detection model construction module, a medical record analysis module, a disease grading module, an information screening module, an early warning information pushing module, a cloud storage module and a display module. According to the invention, the pulmonary tuberculosis focus area can be detected more accurately through the detection model construction module; besides, missed diagnosis of pulmonary tuberculosis caused by busy work or insufficient understanding and other reasons is reduced through the screening module; and automatic pulmonary tuberculosis reporting work achieves whole-process electronization and automation, pulmonary tuberculosis diagnosis work of a non-fixed-point medical institution is all well-documented, and the defect that time and labor are consumed when tuberculosis missing report and missing registration special work is conducted, is overcome.

The invention discloses a laryngeal disease diagnosis system based on a deep learning neural network, and belongs to the subject crossing field of combination of artificial intelligence and medical diagnosis. According to the invention, the problems of low diagnosis efficiency and low diagnosis accuracy of the laryngoscope image in the traditional method are solved. The laryngeal disease diagnosisnetwork model is established, and the established laryngeal disease diagnosis network model can be used for an intelligent system for laryngeal disease diagnosis, so that laryngoscope images are better diagnosed, doctors are helped to improve the diagnosis efficiency and diagnosis accuracy of diseases, and the missed diagnosis rate and the misdiagnosis rate are reduced. The laryngeal disease diagnosis system can be applied to intelligent detection of laryngoscope images.

The invention relates to an intelligent auxiliary Parkinson diagnosis method and device based on biological characteristics. The method comprises that Each result information of first to fourth dispersions and first to fourth similarities is distributed with a corresponding weight coefficient, an auxiliary Parkinson diagnosis result is obtained according to each result information and the corresponding weight coefficient, and the auxiliary Parkinson diagnosis result is output. According to the method, auxiliary diagnosis information which evaluates the severe degree of Parkinson of a tested one is generated according to the obtained biological characteristics of the tested one.

The invention relates to a method and a kit for enriching 4000 human pathogenic target genes. The method comprises the following steps: breaking DNA to obtain DNA fragments, carrying out terminal repairing, carrying out A addition, carrying out linker addition, and carrying out Pre-Capture LM-PCR; mixing Pre-Capture LM-PCR products of 4000 human pathogenic target genes to obtain a mixed library sample, hybridizing a specific probe library of the 4000 human pathogenic target genes with the mixed library sample, capturing DNA by adopting magnetic beads, and carrying out eluting; and carrying out Post-Capture LM-PCR. The kit contains a DNA terminal repairing reagent, an A addition reagent, a linker addition reagent, an LM-PCR reagent and a capturing reagent. According to the method and the kit, the 4000 human pathogenic genes are taken as target spots and are combined together, the gene sequences are enriched once through a biotin probe capturing technology and are subjected to high-throughput sequencing, therefore, multiple variation types of the genes can be simultaneously examined, and the detection sensitivity is high; the 4000 pathogenic genes can be taken as a principal diagnosis and detection means clinically, the cost is reduced, the missed diagnosis is reduced, the detection positive rate is improved, and therefore, the clinical popularization is easy.