407 results about "Missed diagnosis" patented technology

The invention relates to a processing system of a CT image, which comprises a CT image acquiring module, an interesting region estimating module, a characteristic extracting module, an abnormal signal identifying module and a displaying module. The CT image acquiring module is used for acquiring a head CT image subjected to brain tissue segmentation; the interesting region estimating module is used for estimating an interesting region of subarachnoid space to the head CT image; the characteristic extracting module is used for extracting the characteristic to the head CT image subjected to the estimation of the interesting region to acquire a characteristic value; the abnormal signal identifying module is used for identifying whether an abnormal signal is included in the interesting region according to the characteristic value by using a method of mode identification; and the displaying module is used for displaying the identified result and the interesting region in which the abnormal signal exists. The invention also relates to a processing method of the CT image. The invention can display a position in which the abnormal signal exists, which is referred by medical staffs, so as to reduce misdiagnosis/missed diagnosis rate of the subarachnoid space hemorrhage.

The invention provides a method for constructing a fracture recognition model and application. The method comprises: step A, acquiring X-ray images of different fracture types of different fracture parts; step B, carrying out image preprocessing on the X-ray images of different fracture types of different fracture parts; step C, carrying out image feature extraction on the preprocessed images; step D, generating a candidate region according to the image features; step F, carrying out fracture type target detection and region localization on the candidate region; step G, according to a facturetype target detection score, carrying out difficult sample mining to obtain negative samples for training; step K, expanding positive samples for training based on an adversarial learning strategy; and step H, training is carried out by using the positive samples and the negative samples so as to obtain all fracture recognition models for different fracture parts. Therefore, the fracture type andthe regional position can be predicted accurately by using the model, so that the diagnosis time of the doctor is reduced substantially and the missed diagnosis and misdiagnosis rates are reduced.

The invention discloses a real-time auxiliary system for a controllable capsule endoscope operation on the basis of deep learning, and an operation method. The system comprises at least one client side and a service side, wherein at least one client side is connected with a capsule endoscope, is used for obtaining a capsule endoscope image collected by a current capsule endoscope and uploading thecapsule endoscope image to the service side through a network, and is also used for receiving and displaying an analysis result fed back from the display service side; the service side carries out capsule endoscope image processing according to the capsule endoscope image sent from the client side, judges positions and position characteristics corresponding to the capsule endoscope image in realtime, and feeds back the analysis result to the client side; and the service side comprises a sample database, a convolutional neural network model and a web service module. By use of the system, theimage collected by the controllable capsule endoscope is subjected to blind area monitoring and cancer focus identification and is displayed on the client side, an operation physician is assisted in checking the controllable capsule endoscope, detection accuracy and effectiveness is improved, and a missed diagnosis probability of occurrence is lowered.

The invention provides a method for locating and identifying a lesion of a medical image, a device, an equipment and a storage medium. The method comprises the following steps of obtaining a target image, wherein the target image is a medical image to be localized and identified; preprocessing the target image to obtain a preprocessed image; inputting the pre-processed images into the pre-established lesion location recognition model, and obtaining the corresponding lesion location indication map of the target image and the lesion classification in the target image; obtaining the lesion location recognition model by training medical images with the lesion classification labeled; The lesion location recognition model based on that present application can automatically detect a lesion from amedical image, the location of the lesion is given, which not only saves manpower and reduces the time consumed in identifying and locating the lesion, but also avoids the misdiagnosis, missed diagnosis, locating and identifying lesion caused by manual locating and locating the lesion, which can also assist doctors to carry out rapid and accurate analysis, and improves the diagnosis efficiency and accuracy of doctors.

The invention discloses an intelligent fault diagnosis method for a gear box, which comprises the steps of detecting and judging, wherein the result of judging is that the hidden trouble of fault exists or does not exist; if so, early warning; if not, finishing. The steps at least comprise case-based reasoning and judging, rule-based reasoning and judging, and SVM judging. The intelligent fault diagnosis method overcomes the defects of the single diagnosis method, such as easy missed diagnosis, difficult obtaining of diagnosis knowledge, and the lack of fault case samples in the existing intelligent diagnosis, can intelligently diagnose the hidden trouble of fault of the gear case, effectively improve fault precision rate of the gear box, realize intelligent diagnosis on the gear box, lower labor intensity of diagnosis personnel and risk of missed diagnosis.

The invention discloses a fault diagnosis method based on multi-sensor signal analysis. Pressure, temperature, flow and other parameters of a heat pump unit in the operation process are acquired by utilizing multiple sensors, and vibration signals of the unit are acquired by utilizing a vibration sensor so that the equipment state of the air source heat pump unit can be comprehensively mastered. Multiple intelligent technical methods are combined on the basis, and respective advantages of the intelligent technologies are comprehensively applied to perform state monitoring, fault diagnosis and intelligent indication on the air source heat pump unit through enhancing advantages and avoiding disadvantages so that sensitivity and accuracy of a monitoring and diagnosis system can be effectively enhanced and misdiagnosis rate and missed diagnosis rate can be reduced. Meanwhile, a use-facilitating signal processing platform is designed by adopting a GUI design method based on the MATLAB language. An accurate diagnosis decision is provided for general operation personnel without understanding system mechanism or analyzing data.

The present invention discloses a medical big data based disease automatic assistance diagnosis system and method. The system comprises: a background data storage unit; an information processing unit, which specifically comprises: a statistical classification module, used for acquiring case data in the background data storage unit, and performing statistical classification on the case data, so as to obtain a symptom set and a definitively diagnosed disease type set; a diagnosed disease set calculation module, used for calculating a definitively diagnosed symptom set of various diseases according to the symptom set and the disease type set that are obtained by the statistical classification module; and a disease automatic diagnosis module, used for acquiring disease symptom data provided by a user, generating a selection symptom set, comparing the selection symptom set with the definitively diagnosed symptom set of various diseases and performing calculation, so as to obtain a disease determination result; and a man-machine interaction unit, used for displaying an interface of selecting a disease by the user, and outputting a disease diagnosis result. The method disclosed by the present invention is simple, easy and strong in operatability, and provides a new clinic assistant diagnosis tool for the medical field, and reduces an error/miss diagnosis rate.

The invention relates to a medical diagnostic reagent, in particular to a quick detection reagent for the early diagnosis of acute myocardial infarction and a test strip. The invention provides a double-index united detection reagent containing a specific antibody resisting a human heart-type fatty acid binding protein H-FABP and a human cardiac troponin cTnI. The invention also provides a colloidal gold labeling immunological chromatographic test strip containing the specific antibody resisting the H-FABP and the cTnI, which is used for quickly detecting the acute myocardial infarction. The double-index united detection reagent can carry out early diagnosis on a patient having the acute myocardial infarction, can also prevent the missed diagnosis of a patient having long-time chronic myocardial infarction with mild symptoms and better solves the influence caused by a difference existing in the detection time. The colloidal gold labeling immunological chromatographic test strip provides a quick, convenient, cheap and practical detection tool for the early diagnosis of the acute myocardial infarction, is hopefully used for hospitals at all levels and can also be used for the self monitoring of the patient.

The invention discloses a computer aided detection method for microcalcification in mammograms, which comprises the following steps of: performing grayscale correction and transformation on a mammogram image to obtain an image after grayscale correction; acquiring microcalcification enhanced mammogram image by adopting dual-structure element-based background superposition method, and simultaneously acquiring another background suppressed mammogram image by adopting a Top-hat transformation method in morphology; segmenting the two images by adopting double thresholds to acquire a preliminary microcalcification image, and performing postprocessing to form a coarse microcalcification detection image; and extracting characteristics from coarsely detected microcalcification destination area, classifying by using a support vector machine, removing false microcalcification destination areas, marking the rest microcalcification into the mammogram image for the reading of doctors. The method can free the doctors from trivial mammogram reading and classifying work, and assists the doctors in better understanding and judging the images so as to reduce error diagnosis and missed diagnosis and fulfill the aim of improving diagnostic accuracy.

The invention relates to a generalized-morphology-based automatic filling system fault diagnosis method. The method includes that aiming at high-speed motion of each mechanism when an automatic filling system works, measuring points are arranged at each angle motion component position, a driving motor and a power source portion to measure vibration acceleration, angle motion parameters and load current response signal for data analysis and fault classification recognition; experiment testing, signal processing, feature extracting and fault diagnosis are integrated, and automatic diagnosis, alarming and predicting can be realized. Aiming at different fault types of the automatic filling system, a generalized-morphology-based early fault diagnosis method is developed, convenience and quickness in fault diagnosis and prediction of the automatic filling system are realized, the problem that a medium-large-caliber artillery automatic filling system is backward in maintenance means and needs to be demounted greatly for inspection is solved, and the fault diagnosis method is high in intelligence level, low in maintenance cost, short in period, less prone to misdiagnosis and missed diagnosis and adaptable to needs of equipment and weapon development.

The invention discloses a gastroscope image auxiliary processing system and method based on ensemble learning. The system includes an image acquisition module, a data preprocessing module, a neural network training module and an ensemble learning module, optimization of processes of screening, data classification and expansion of image data is realized through the data preprocessing module, the neural network training module realizes expansion of an adopted convolutional neural network model, and at the same time, a method for integrating different classifiers that are generated is provided through the ensemble learning module, aiming at obtaining a final decision classifier to improve overall performance of the classifier, thereby meeting requirements of clinical auxiliary diagnosis and treatment in four indexes of sensitivity, specificity, rate of missed diagnosis and misdiagnosis rate, and the method effectively improves recognition efficiency and accuracy, truly playing a role of assisting diagnosis and treatment.

The invention relates to the field of medical image processing and pattern recognition, and provides a computer-assisted lump detecting method based on a mammary gland magnetic resonance image. The computer-assisted lump detecting method based on the mammary gland magnetic resonance image aims at solving the problems that in the prior art, the lump partition effect is not good, and the accuracy, the sensitivity and the specificity in a classification test are not high. The computer-assisted lump detecting method includes the following steps: S1, extracting an interest area from the mammary gland magnetic resonance image; S2, extracting an initial lump area from the interest area in a separated mode, and determining the contour line of the initial lump area; S3, calculating the weight distribution of characteristic parameters of the initial lump area; S4, selecting the characteristic parameters, with the weight coefficients larger than a standard weight coefficient, of the initial lump area, and carrying out training classifying to obtain optimized characteristic parameters; S5, inputting the optimized characteristic parameters into a classifier, analyzing the optimized characteristic parameters with a support vector machine classification method, determining a final lump area, and displaying the final lump area for a user. The detecting method has the good lump partition effect, the accuracy, the sensitivity and the specificity in the classification test are effectively improved, the detecting result serves as a second opinion to be provided for a radiologist, and the misdiagnosis rate and the missed diagnosis rate of the radiologist can be effectively reduced.

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 construction method and application of a thoracic disease detection model. The construction method comprises the following steps: A, acquiring chest X-ray images of a specified number of thoracic disease patients; B, performing image preprocessing on the X-ray images to obtain pre-processed image data; and C, inputting the pre-processed image data to a convolution neural network model for training to obtain the trained model for thoracic disease detection. The thoracic disease detection model can accurately predict a type and a location of a thoracic disease, the diagnosis time of the doctor is greatly shortened, and the rate of missed diagnosis and misdiagnosis is lowered.

The invention provides a working method of a medical diagnosis auxiliary system, and the method comprises the following steps: inputting symptom information into a server, and sampling patient examination data; enabling the server to identify keywords and related sentences in the symptom information and extract the corresponding keywords; according to sample training and statistics obtained from historical medical record data, obtaining detailed features of the disease, finding out potential patients, and depicting a disease model; and matching the depicted disease model with a diagnosis modelin a database to obtain a diagnosis model result closest to the symptom information of a patient. The accuracy of data analysis is further improved through continuous data accumulation; compared withInternet inquiry in the prior art, the method improves the accuracy by 70%; the new analysis vocabularies and logic algorithms are continuously expanded according to logic operation of data, the diagnosis capacity of doctors is improved through auxiliary diagnosis, and the misdiagnosis rate and the missed diagnosis rate are reduced. Doctors apply medicine according to symptoms and actual conditions and query results of the method, and the accuracy rate can reach 100%.

A thickness balancing method and device for a breast image. The thickness balancing method for the breast image includes the steps of: obtaining a breast image; filtering the breast image to obtain a low frequency breast image and a high frequency breast image; performing gray scale transformation on a preset area in the low frequency breast image to obtain a first image, a gray value of a preset area in the first image tending to be consistent with a gray value of a neighborhood, and the preset area referring to an area at a predetermined distance from an edge of the low frequency breast image; and reconstructing the first image and the high frequency breast image to generate a thickness-balanced breast image. The breast image after balancing which is obtained through the technical scheme of the invention has uniform gray scale while details of the breast image are not lost, satisfies actual clinical needs, the breast image after balancing is adopted to be diagnosed on a window level of a certain window width, and loss of edge information of the breast image does not occur, thereby lowering the rate of missed diagnosis, and improving the accuracy rate of diagnosis.

The invention discloses a HPV typing and integrated high-throughput sequencing detection method. The method selects the genes of the current HPV subtypes and combines the second-generation high-throughput sequencing technology to more comprehensively detect the type of HPV infection in patients. It overcomes the difficulties of low accuracy rate, high false positive rate, poor repeatability and high missed diagnosis rate of traditional detection methods. In the field of molecular diagnosis, the most direct and clear technology is gene sequencing. Compared with the current classic Sanger sequencing method, the second-generation high-throughput sequencing technology has higher detection throughput, faster sequencing speed, and higher accuracy. Advantages such as lower cost and richer information. With the help of the second-generation high-throughput sequencing technology, this method can accurately type high-risk HPV and low-risk HPV and detect whether it has been integrated with the human genome, and accurately and individually evaluate the tester to prevent The risk of lesions, thereby preventing the occurrence of tumors.

The invention discloses a medicinal composition containing dimeticone/simethicone. The composition is mainly prepared from the following raw material medicaments in portion by weight: 1 to 1,000 portions of acetylcysteine or pharmaceutically acceptable salt thereof, and 1 to 500 portions of the dimeticone/simethicone. The composition can be prepared into various pharmaceutically acceptable preparations; and the composition can be applied to gastrointestinal endoscopy and treatment or an adjuvant drug administered before imaging examination, and compared with the acetylcysteine or the dimeticone or simethicone which is in single use at the same dosage, the composition has stronger defrothing effect, and stronger grume removal effect, so that the composition can effectively improve the visual definition of examination and treatment, reduce misdiagnosis and missed diagnosis, and improve effectiveness of diagnosis and treatment, contributes to early diagnosis discovery of digestive tract diseases, and has wide application prospect.

The invention provides an intelligent endoscope image analysis method and system. The method comprises steps as follows: acquiring an image and auxiliary information associated with the image; analyzing the image and performing classified marking on the image with a focus zone; extracting the focus zone of the image, comparing the focus zone with a focus feature database corresponding to classified marking of the image, and outputting focus feature information of the focus zone; receiving and displaying focus feature information of the focus zone. Disease feature information contained in the image can be judged by acquiring the image and processing the focus zone of the image and a focus zone selected in an interaction mode, auxiliary diagnosis information is provided for medical diagnosis, and missing diagnosis and misdiagnosis caused by insufficient professional experience of grass-roots doctors can be avoided.

The invention provides a single-lead electrocardio abnormal signal identification method based on generative adversarial networks and convolution recurrent neural networks, and mainly solves the problem that data concentration samples are unbalanced. Categories being low in data concentration data volume are subjected to data enhancement, and then identification and classification of electrocardioabnormal signals are performed, so that reference is provided for assistance of doctors, the wrong diagnosis rate and missed diagnosis rate are reduced, and the workload of the doctor can be alleviated. The generative adversarial networks are used for enabling the data concentration samples to achieve the relative balance, so that training the convolution recurrent neural networks can be realized, and good classification effects can be achieved.

The invention discloses a control method of an intelligent fault diagnosis system of a pneumatic regulating valve, comprising the steps of establishing an off-line fault diagnosis model and an on-linefault diagnosis; By establishing off-line fault diagnosis model and on-line fault diagnosis steps, not only the fault self-diagnosis of pneumatic regulating valve is realized, but also the accuracy of fault diagnosis is improved, especially the missed diagnosis rate is reduced, and the fault operation of valve is avoided. But also has good generality, and can complete fault diagnosis without complicated expert experience and knowledge reserve; Common operators can grasp, improve the pneumatic valve fault diagnosis automation and intelligence.

The invention discloses a high-speed milling chatter on-line identification method based on an AR model. The method comprises the steps that (1) the state information of the milling process is acquired; (2) forced vibration frequency filtering is carried out on a signal; (3) chatter sensitive frequency band filtering is carried out on the signal; and (4) a model characteristic root index R(k) is constructed based on the difference of the AR model of the signal in a stable milling state and a chatter milling state, time varying AR(1) modeling is carried out on the signal in the stable milling process, and a recursive least square method with a forgetting factor is used for identification to obtain the variation of the model characteristic root R(k) of the model in the whole cutting process to identify chatter. Compared with a traditional chatter detection method, according to the method, characteristic information reflecting chatter and characteristic information irrelevant to the chatter are separated, and substantive characteristic parameters of a milling system are obtained; and the physical property of the milling chatter is represented substantially, sensitivity, precision and reliability of chatter detection are effectively improved, and the misdiagnosis rate and the missed diagnosis rate are decreased.

The invention discloses a C0 complexity and correlation coefficient-based milling chatter detection method. The C0 complexity and correlation coefficient-based milling chatter detection method includes the following steps that: state information of a milling process is obtained through a vibration acceleration sensor; obtained signals are preprocessed through using a comb filter, so that periodic components can be filtered out; the complexity of residual signals is calculated through utilizing C0 complexity indexes, so that the degree of nonlinearity of chatter can be reflected; and the correlation coefficient of original signals and filtered signals is calculated, and therefore, the proportion of chatter components in the signals can be reflected, and chatter degree in the machining process can be described. Compared with a traditional chatter detection method, and according to the C0 complexity and correlation coefficient-based milling chatter detection method of the invention, characteristic information reflecting chatter and characteristic information irrelevant to chatter are separated out from each other, and various kinds of indexes are fused, and physical characteristics of milling chatter can be essentially characterized. With the C0 complexity and correlation coefficient-based milling chatter detection method adopted, the sensitivity, accuracy and reliability of chatter detection can be effectively improved, and misdiagnosis rate and missed diagnosis rate can be decreased.

A fully automatic ultrasonic scanner and a scan detection method is provided. The fully automatic ultrasonic scanner comprises: an ultrasonic probe (4); a driving system (5) for driving the ultrasonic probe (4) to move; and a flexible structure on which the ultrasonic probe (4) is mounted, wherein the flexible structure enables the ultrasonic probe (4) to be always along a curve of a skin surface and keep perpendicular to the skin surface during scanning. The flexible structure of the fully automatic ultrasonic scanner has self-adaptive effect, can adjust the scan trace and the probe angle in real time according to different curves of the human body, and ensure that the ultrasonic probe (4) scans against the skin surface and keeps perpendicular to the skin surface, which improves the quality of the scanned images, so as to enhance the detection rate and accuracy rate of early screening, and reduces the probability of missed diagnosis.