509 results about "Diagnostic accuracy" patented technology

The invention provides an artificial-intelligence auxiliary interrogation diagnosis system which comprises a corpus training module, a question classification module, a chat processing module, an automatic response module for medical commonsense questions and an interrogation diagnosis interaction module, wherein the corpus training module automatically extracts a patient's illness sign information, treatment and medicine use information and diagnosis suggestion information in historical conversations between a doctor and the patient, semantic representation is conducted according to the various types of extracted information, the information is converted into numerical vectors, training learning is conducted on the numerical vectors, and a diagnosis model which contains relations between the illness sign information existing in historical doctor-patient interactions and the doctor's diagnosis suggestions is obtained; and the question classification module judges and classifies information consulted by the patient into an online interrogation diagnosis question or a question outside the interrogation diagnosis service scope or a medical commonsense question, and the questions are answered by the interrogation diagnosis interaction module, the chat processing module and the automatic response module of the medical commonsense question respectively. The system is characterized in that by various models obtained through training, an interrogation diagnosis mechanism can be learned independently and perfected; and system universality and diagnosis accuracy are increased.

The invention discloses a method for performing fault diagnosis by using model conversion. According to the method, processed related information of a fault mode effect analysis model is converted into a corresponding Bayesian network model by using a self-defined data structure while ensuring complete data, an elementary event, a logic gate and an intermediate event of a fault tree in a fault tree analysis model are converted into nodes in a Bayesian network respectively, and a corresponding conditional probability table in the Bayesian network is set. The fault diagnosis is performed through the converted Bayesian network model. The method of the invention expands the use of the Bayesian network model in the fault diagnosis, improves the diagnosis accuracy of a fault diagnosis model in practical application, ensures the universality of model conversion, and can realize cross-tool conversion among different fault mode effect analysis, fault tree analysis result and the generated Bayesian network.

The invention discloses an automatic screening system and an automatic screening method using a thin-prep cytology test. The system comprises an image acquisition module, an image segmentation module, a target recognition module and a comprehensive analysis module. The method provided by the invention respectively adopts a three-point evolutionary uniform sampling method to guarantee the reliability of automatic focusing, the coarse-to-fine segmentation algorithm to improve the segmentation accuracy of pathological cell nucleuses, a series of filters to rapidly filter various impurities, and a cascade classifier training method to greatly reduce the false positive rate, and uses relative features to conduct comprehensive analysis. The method disclosed by the invention has the advantages of high sensitivity over pathological cells, high specificity over normal cells, high recognition speed and high automation degree, so that the diagnostic accuracy can be improved and the workload of a cytopathologist is reduced at the same time.

The present invention provides a dermoscopy image automatic segmentation method based on a full convolutional neural network. The method comprises the following four steps: 1: obtaining of dermoscopy images and true value graphs; 2: performing structure design of a full convolutional neural network; 3: performing design of feature fusion and a per-pixel segmentation method; and 4: performing network training and segmentation. According to the steps mentioned above, an end-to-end depth convolutional neural network is obtained through training to perform accurate segmentation of the dermoscopy images and allow a small area skin lesion area to be effective so as to solve actual problems that the skin lesion area segmentation is not good to influence the subsequent diagnosis accuracy in the dermatology computer auxiliary diagnosis system.

The invention provides an abnormal engine sound fault on-line diagnostic system and diagnostic method. The diagnostic system comprises a sound signal sensor, a sound signal collecting and storing module and an analysis and diagnosis module. The analysis and diagnosis module comprises a spectral analysis and diagnosis module, a wavelet analysis and diagnosis module and a psychological acoustics analysis and diagnosis module, wherein the spectral analysis and diagnosis module is used for obtaining frequency spectrum parameters through the spectral analysis, the wavelet analysis and diagnosis module is used for obtaining energy parameters through the wavelet analysis, and the psychological acoustics analysis and diagnosis module is used for obtaining psychological acoustics parameters through the psychological acoustics analysis. The spectral analysis and diagnosis module, the wavelet analysis and diagnosis module and the psychological acoustics analysis and diagnosis module are used for extracting the frequency spectrum parameters, the energy parameters and the psychological acoustics parameters respectively to show abnormal sound fault characteristics. The abnormal engine sound fault on-line diagnostic system replaces manual work to carry out on-line diagnostic working on the abnormal engine sound fault, the diagnostic accuracy and consistency are high, the abnormal sound fault diagnosis requirements for real-time performance, fastness and accuracy of an engine production line can be met, and the working efficiency of an engine detection line is improved.

A plurality of electrodes for measuring electrocardiogarphic waveforms are attached on body surface positions that constitute a subset of the standard 12-lead system. The measured electrocardiogarphic waveforms of said subset of said standard 12-lead system are used to calculate the electrocardiogarphic waveforms of remaining leads in the said standard 12-lead system. The measured and calculated electrocardiographic waveforms are synthesized to form a standard 12-lead electrocardiogram. The invention is capable of monitoring the 12-lead electrocardiogram with reduced number of electrodes, wherein a portion of waveforms are directly measured and used as primary information for diagnosing heart disease, and the other portion of waveforms are derived from the measured leads and are used as a secondary information for improving the accuracy of diagnosis. The invention is especially useful in monitoring ischemic heart disease and acute myocardial infarction in cases where mounting and maintaining ten electrodes to obtain 12-lead electrocardiogram is difficult, such as in ambulatory monitoring, long-term monitoring and home monitoring.

A system for administering a health care system, comprising up to five processing means and a database. The system accepts inputs from enrollees, insureds, providers, insurers and vendors, but requires no input from an administrator. The apparatus optionally includes devices for: measuring patient treatment outcomes, a “just-in-time” appointment rescheduler, an editable archive of prototypical treatment plans, measuring the diagnostic accuracy of diagnosing physicians and measuring the complication rate of treatment providers.

The invention discloses a method and equipment for identification and classification of an electrocardiogram, which identifies a QRS wave configuration pattern to obtain the configuration characteristics of P waves, T waves and QRS waves by calculation of time-domain characteristics, configuration extraction pretreatment and point calculation of QRS wave configuration characteristics. The technical scheme of the invention fully uses the identification process of ECG configuration characteristics from thinking and experience of doctors to realize the effect using small amount of work to achieve a high diagnostic accuracy of cardiac arrhythmia and provides a good base for the application of an ECG diagnostic apparatus.

A method and system for detecting the presence of restriction to expired airflow in humans or animals by analyzing the expired capnogram and oxygram, as well as the geometric analysis of the real-time plot of the waveform that depicts the instantaneous ratio of CO₂ to O₂ (the carboxygram ratio). Airway obstructions causes an increase in the Q-angle between the slope of phase 11 and slope of phase III in the expired carboxygram. The diagnostic accuracy of the detection of airways obstruction is further enhanced by measuring the ratio of time spent in exhalation (Te) versus inhalation (Ti). The system uses the combination of an increased carboxygram Q-angle, and a prolonged Te/Ti to detect presence of airways obstruction.

The invention provides methods for determining the probability that a given region of a tissue sample contains tissue of a given category, such as CIN 1 (cervical intraepithelial neoplasia, grade 1), CIN 2/3 (cervical intraepithelial neoplasia grades 2 and/or 3), normal squamous, normal columnar, and metaplasia, for example. The invention provides increased diagnostic accuracy by combining a plurality of statistical classification techniques. Furthermore, in one embodiment, the invention comprises combining one or more statistical techniques with one or more non-statistical classification techniques.

The invention discloses an image-based early diagnosis system for senile dementia. The early diagnosis system comprises a brain image collector (1) and a memorizer (2) and is characterized by comprising image preprocessing equipment (3), an image asymmetry characteristic extraction device (4), a consolidator (5), a characteristic vector selector (6), an optimization classifier (7) and an output device. The early diagnosis system has the remarkable effect of being capable of obviously improving the accuracy of early diagnosis of the senile dementia when being applied to the early diagnosis of the senile dementia. The early diagnosis system not only can be used for realizing the clinical application of sequential asymmetry characteristics of anatomical structures in the brain MR (Magnetic Resonance) images in the early diagnosis of the senile dementia, but also is high in diagnosis accuracy and real-time performance.

The invention discloses a bearing fault mode diagnosis method and system based on deep learning. The method comprises: a learning rate is adjusted automatically, noises are introduced, and a linear correction unit is introduced as an activation function, and thus a deep belief network is improved based on the activation function; a data set and long- and short-term memory networks are migrated toobtain a synthetic data set, a trained data set is extended based on the synthetic data set, and the improved deep belief network is trained based on training data set training to obtain a bearing fault diagnosis model; and a vibration signal of the bearing is collected and a bearing failure mode is diagnosed based on the bearing vibration signal and the bearing fault diagnosis model. On the basisof combination of a semi-supervised learning and a migration learning algorithm, the diagnostic accuracy is improved while no insufficient data are provided.

The invention discloses a rolling bearing fault diagnosis method based on improved variational model decomposition and an extreme learning machine. The method comprises: vibration signals of a rollingbearing under different types of faults are collected, the vibration signals are filtered by means of maximum correlation kurtosis deconvolution, parameter optimization is carried out on the maximumcorrelation kurtosis deconvolution method by using a particle swarm algorithm, and an enveloped energy entropy after signal deconvolution is used as a fitness function; the mode number of variationalmodel decomposition is improved by an energy threshold and improved variational model decomposition of the filtered vibration signals is realized to obtain mode matrixes of the corresponding vibrationsignals; singular value decomposition is carried out on the mode matrixes to obtain a singular value vector and a rolling bearing fault feature set is constructed; and the fault feature set is trained by using an extreme learning machine and a rolling bearing fault diagnosis model is established. Therefore, stable feature extraction of the complex vibration signal of the rolling bearing is realized, so that the diagnostic accuracy is improved.

The invention relates to a stacked SAE (Sparse Autoencoder) deep neural network-based bearing fault diagnosis method. The first layer of a network is applied to the qualitative judgment of a bearing fault, that is, the first layer of the network is applied to the fault type judgment of the bearing fault; and the second layer of the network is applied to the quantitative judgment of the bearing fault, that is, the second layer of the network is applied to the severity judgment of the bearing fault. According to the method of the invention, empirical mode decomposition (EMD) and an autoregressive (AR) model are combined together to perform pre-processing on original bearing signals, the parameters of the AR model are extracted and are adopted as the input of the network, and therefore, the input dimensions of the network can be greatly reduced, the simplification of calculation can be facilitated, and the training and testing of the network can be accelerated; a deep neural network on which the method of the invention is based can further automatically extract features of the input and qualitatively and quantitatively determine the bearing fault automatically, and therefore, the diagnostic accuracy of the method of the present invention can be ensured, and at the same time, dependence on signal processing expertise can be decreased, manual judgment is not required, the consumption of manpower can be decreased; and thus, the method has a higher practical value in the era of big data.

A method for eliminating motion artifacts in digital subtraction angiography comprises the following steps: step 1, reading; step 2, selecting points; step 3, constructing a DSA space body; step 4, carrying out space slicing; step 5, connecting tracks; step 6, analyzing the space motion characteristics of DSA pixel; step 7, carrying out triangulation; step 8, carrying out affine transformation; step 9, carrying out space winding; step 10, optimizing; step 11, registering; step 12, carrying out grey correction; step 13, carrying out logarithmic subtraction angiography. The invention belongs to the technology of image processing. The method of space analysis is used so that the registering of DSA image is more accurate, thereby efficiently eliminating motion artifacts and obtaining clear angiography images. Thus, the diagnostic accuracy of a doctor is improved, and the working efficiency is increased.

The invention provides a power grid fault diagnostic model and a diagnostic method thereof and belongs to the technical field of power grid fault diagnosis. A mathematical expression of the traditional fault diagnostic analytic model can be abstracted as the formula; analysis on the action state of protecting a switch is converted into description on the probability of protecting the switch; the description is information transmission uncertainty description based on an information theory; an objective function is established as an optimal solving function; mutual information between an information sink and an information source under every failure mode is calculated when multiple optimal solutions, namely multiple failure modes exist; corresponding failure modes are most likely to occur if the quantity of condition self-information is smallest; and a principle that the mutual information of the information sink and the information source is maximum is utilized to determine a fault sorting result when the plurality of condition self-information is similar. The power grid fault diagnostic model and the diagnostic method thereof have the advantages of enabling the uncertainty of fault diagnosis to be integrated in an analytic model, enabling the fault tolerance of the model to be improved and the dimension of the model to be greatly reduced, being high in diagnostic speed and diagnostic accuracy, being capable of being well applied to a scheduling terminal and playing a positive and important role in the field of the power grid fault diagnosis.

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 present invention provides a hand-held device of the type useful to perform an in vitro immunodiagnostic assay for an analyte of interest. In embodiments of the invention, the device incorporates a carrier for conducting the assay which comprises an array of pads including a sample pad, a narrow detection pad for detecting analyte, and a bridging pad coupled in flow communication therebetween. The device is provided with a housing adapted to deliver sample onto the carrier as a linear band having a width greater than the detection pad. In combination, these features yield a device that retains the sensitivity required for diagnostic accuracy, yet is improved in terms of cost and ease of manufacture.

The present invention discloses a method for automatically recognizing white blood cells in leucorrhea based on the convolution neural network, and belongs to the field of automatically recognizing medical microscopic images by using a machine vision scheme. The method comprises: firstly artificially identifying a plurality of white blood cell images and non-white blood cell images; establishing a 9-layer neural network, and training the 9-layer neural network according to the artificially identified images; and carrying out corresponding adjustment in real time on training parameters and a learning rate of the neural network during the training process; and after completion of training the neural network, carrying out detection on to-be-detected target images. According to the method disclosed by the present invention, effects of reducing labor intensity of image reading by doctors and improving the diagnostic accuracy are achieved.

The invention discloses a numerical control machine tool operation state health diagnosis system and diagnosis method. The system comprises an edge data acquisition module, a cloud platform and a client, the cloud platform is connected with the edge data acquisition module and the client. The cloud platform comprises a predictive diagnosis module, an information management and storage module and an application server; and the predictive diagnosis module comprises a real-time diagnosis module, a post-processing diagnosis module and a comprehensive diagnosis module. According to the system, a cloud platform with a diagnosis function is used for replacing an upper computer, the expandability is greatly improved, the diagnosis and prediction tasks can be completed only by deploying corresponding sensors on the machine tool and connecting the sensors to the cloud platform, the real-time state data of the machine tool are managed in a centralized mode through the cloud platform, and the health condition of the machine tool is evaluated and predicted through the diagnosis technology based on the neural network. The diagnosis method is simpler, and the health diagnosis accuracy is high.

The invention discloses an analog circuit test node selecting method based on dynamic feedback neural network modeling. The method comprises the steps of selecting the frequency of a test signal, inputting the test signal into a circuit to be tested, simulating various typical fault conditions, collecting the voltage values of a normal sample and a fault sample of the circuit on a test node to be selected of the circuit to be tested so as to construct a fault dictionary table; according to a fault fuzzy voltage interval, analyzing a fuzzy fault set and obtaining a fault integer encoding table; constructing an initial training sample set, training an initial dynamic feedback neural network, and utilizing the dynamic feedback neural network for fitting the nonlinear mapping relation between the test node and the fault; and according to the target function calculated by the genetic algorithm output by the network, obtaining the optimal test node set by utilizing the genetic optimization algorithm. In the method, a fault dictionary is analyzed by the intelligent algorithm, so that the global optimum test node set can be found, and the subsequent diagnostic accuracy can be further improved.

The invention discloses an asynchronous motor fault monitoring and diagnosing method based on deep learning. The asynchronous motor fault monitoring and diagnosing method comprises the following stepsthat electric power load time series of an asynchronous motor in known working condition types are acquired, the time span is Num1 electric power load cycles, and electric power load data at each sample time includes data of three dimensions of voltage, current and power; the voltage data, the current data and the power data are separately used as the gray value of pixel points of three layers inRGB images, and the time series of the electric power load cycles are transformed into the RGB image in a segmented mode, and each electric power load time series correspondingly obtains a set of feature image time series; and a deep neural network is trained by the feature image time series of the asynchronous motor and the corresponding working condition types, and a fault diagnosis model is obtained and then used for classifying the working conditions of the asynchronous motor to be tested. The fault diagnostic accuracy of the asynchronous motor fault monitoring and diagnosing method is high, and the threshold of employees is reduced while saving the time of system development.