255 results about "Diseases/diagnoses" patented technology

The invention discloses a multi-agent large-model disease diagnosis knowledge reasoning system based on data dual drive, and relates to the technical field of artificial intelligence assisted medical diagnosis. The system collects patient symptom follow-up records, laboratory test results, observation diagnosis probabilities and expert diagnosis recommendation results in a multi-source manner; time sequence evolution characteristics are extracted, a time sequence diagnosis sensitivity coefficient is calculated, and early recognition of disease risks is achieved; in combination with anti-fact simulation and statistical reasoning, a causal consistency coefficient is obtained and is used for verifying causal reasonability of observation diagnosis and contrast results; based on agent group consensus analysis, calculating a game consistency coefficient for judging the credibility of a diagnosis conclusion; positioning and multi-level verification are carried out on abnormal reasoning steps and knowledge fragments, so that the reliability and safety of a result are guaranteed; continuous optimization of the diagnosis model is realized through a log analysis and knowledge backflow mechanism; according to the invention, the accuracy, interpretability and safety of disease diagnosis can be obviously improved.

The invention discloses an AI-driven disease diagnosis and curative effect monitoring analysis system, which comprises a data processing module used for collecting multi-source diagnosis and treatment data and constructing a multi-modal sample set; the diagnosis prediction module is used for generating a diagnosis label, a diagnosis confidence value and a prediction curative effect trend through a multi-task learning model; the curative effect comparison module is used for collecting actual curative effect data, constructing a curative effect observation sequence and generating a curative effect deviation sequence based on the curative effect observation sequence and the predicted curative effect trend; the credibility evaluation module is used for executing credibility review based on the curative effect deviation sequence and the diagnosis confidence value; and the path evolution module is used for recording continuous multi-round diagnosis correction results, generating a jump type diagnosis path and updating a diagnosis and treatment data chain. The dynamic closed-loop verification mechanism between the multi-mode diagnosis and treatment data and the predicted curative effect trend improves the accuracy of disease diagnosis, the reliability of curative effect prediction and the adaptability of the diagnosis and treatment process.

The invention relates to the cross technical field of artificial intelligence and medical image processing, in particular to a medical image segmentation system and method based on a wavelet bridge diffusion model and an efficient conditional random field. A WBDM-ECRF framework is constructed and comprises a discrete wavelet transform module, a BDM-T module, a BDM-S module and an ECRF module; decomposing the image through discrete wavelet transform, extracting a low-frequency sub-band, and enhancing the contrast ratio of a focus and normal tissues; the BDM-T takes U-Net as a backbone, integrates a FlashAttention mechanism, and optimizes a variance formula to realize efficient training; the BDM-S adopts a leapfrog sampling strategy, so that the reasoning time is greatly shortened; the ECRF introduces a multivariate potential function of a structural similarity index and smooth operation through edge expansion, and accurately optimizes edge segmentation. According to the method, the dependence of marked data is reduced, the training and reasoning efficiency is improved, the problem of fuzzy edge segmentation is solved, the Dice coefficient and intersection-union ratio performance on the ISIC data set is excellent, and reliable quantitative support is provided for disease diagnosis and treatment.

The present invention relates to the technical field of dental medical treatment, and specifically discloses a method for using a CBCT (Cone Beam Computer Tomography) for automatically positioning a tooth, which comprises the following steps: S01, archiving CT data; S02, calculating AI and generating AI results; S03, entering CT reading by a client; S04, downloading and loading the CT data; S05, opening a function of a tooth lens; S06, selecting a corresponding tooth position in a tooth position list; and S07, selecting a 3D tooth rendering mode. The function of the ‘tooth lens’ is added into traditional CT reading, which is suitable for clinical disease diagnoses such as ‘tooth extraction, root canal therapy, tooth repair’ and the like in oral treatment, and doctors can precisely and quickly position the single tooth through selection for the tooth position, which facilitates more comprehensive analysis.

The invention discloses a similarity measurement-based few-sample electrocardiosignal classification method. The method comprises the following steps of: acquiring an electrocardiosignal sequence from a public library and dividing the electrocardiosignal sequence into a support and query set according to a few-sample format; performing normalization processing and zero filling operation on the obtained sequence; inputting the sequence data with the uniform length into a parameter-shared one-dimensional convolutional neural network to extract a time sequence embedded feature vector; after the vectors are spliced and multiplied, weighting the vectors into weighted features through an attention network; inputting the weighted features into a multi-layer perceptron to calculate a similarity score, and training and fixing a neural network by using positive and negative sample pairs; and calculating the maximum similarity between the query sample and the support set, and outputting a prediction category. According to the method, the electrocardiosignal labeling cost can be remarkably reduced, the accuracy and efficiency of abnormal heart rhythm detection can be improved, dependence on large-scale labeling data is reduced, the practicability and expandability of electrocardiosignal classification are improved, and the method is applied to the field of medical signal processing and has important significance in the aspects of abnormal heart rhythm detection, disease diagnosis, wearable equipment application and the like.

The invention discloses a method for testing and evaluating safety effectiveness of a digital therapy product for mental diseases, which relates to the technical field of mental diseases and comprises the following steps: S1, evaluating objective indexes of the digital therapy product; s2, testing a man-machine interaction feedback mode; s3, mapping measurement of product output content and patient feedback; and S4, performing real-time adjustment and optimization. According to the method for testing and evaluating the safety effectiveness of the mental disease digital therapy product, aiming at the mental disease digital therapy product, a special testing and evaluating system is constructed, objective indexes and output contents of the digital therapy product are tested, and in the interaction process of a patient and the product, the safety effectiveness of the product is tested. By detecting data such as micro-expression, electroencephalogram signals and eye movement signals of a user in real time, feedback of a patient is obtained in real time, corresponding relation analysis is conducted on the feedback and output content of a digital therapy product, therefore, the treatment effect of the product can be accurately evaluated, and the accuracy of mental disease diagnosis and treatment is improved.

The invention discloses a disease diagnosis method and system based on multi-modal space-frequency domain adaptive fusion, and relates to the field of artificial intelligence and biomedical engineering.The method comprises the steps that multi-modal data are standardized, the unified and standardized multi-modal data are coded, and multi-modal initial feature representation is obtained; after projection and gating alignment and cross-modal interactive attention alignment are carried out on the initial feature representation of each modal, enhanced representations of each modal are obtained, and then the enhanced representations of each modal are fused into a shared feature representation; performing deep feature extraction on the enhanced representation of each mode to obtain deep features of each mode, and performing adaptive multi-domain feature enhancement processing to obtain multi-domain enhanced features of each mode; performing semantic alignment on the multi-domain enhanced features of each mode, and then performing fusion through a hierarchical attention mechanism to obtain fusion features; and the fusion features are input into a diagnosis network for prediction, a disease diagnosis result is obtained, and the intelligent diagnosis precision and robustness of papillary thyroid carcinoma are improved.

The invention discloses a portable detection system for early diagnosis of orthopaedic joint diseases, and relates to the technical field of orthopaedic joint disease diagnosis, which comprises the following steps: acquiring acoustic and motion feature vectors of joint activities, calculating a quality score based on an acoustic signal-to-noise ratio and a motion goodness of fit, performing weighted splicing on features by taking the quality score as a weight after normalization, generating a fusion feature vector; thirdly, reconstructing features by using a reverse decoding model, comparing the same degree with the original features, weighting to obtain fusion same degree, and iteratively adjusting the weight to optimize the fusion effect; and performing clustering analysis on the candidate feature vectors, replacing the candidate feature vectors if the candidate feature vectors deviate from a clustering center, and finally inputting a health model to output a diagnosis report. Through dynamic weighting of quality scores, high-fidelity features are processed preferentially, the information density and discrimination capability are improved, and low-quality data interference is reduced; and self-checking and dynamic correction are realized through reconstruction and similarity comparison, so that the fidelity and consistency of the characteristics are ensured, and the robustness of the system is enhanced.

The invention relates to the field of medical image analysis and neuropsychiatric disease diagnosis, and particularly discloses a brain network typing diagnosis system and method for attention deficit hyperactivity disorder. Comprising a data acquisition and preprocessing module, a topological feature extraction module, a supervised manifold learning module, a network reconstruction and diagnosis module, a parameter optimization module and a result verification module, continuous homology analysis is performed on a brain region function connection matrix based on an algebraic topology theory, and a brain region topological feature matrix is generated; utilizing supervised manifold learning to map the brain region topological feature matrix to a Riemannian manifold, and calculating a brain region importance weight map based on ADHD phenotypic features as supervised signals; according to the method, the brain network is subjected to subtype specific reconstruction through the curvature flow theory of Riemannian geometry, the topological difference between different subtypes is calculated, accurate typing diagnosis of ADHD is achieved, and the diagnosis accuracy is improved by 15%-20%.

The invention discloses a fusion method for balancing medical text positive and negative sample training, belongs to the field of medical text classification crossing natural language processing and medical informatics, and solves the problems of long text information loss, labeling noise interference and positive and negative sample imbalance in the prior art. The method comprises the steps of preprocessing a medical text, defining five types of medical entities such as disease diagnosis and the like, and labeling and establishing entity association; segmenting a long text by using a sliding window, extracting text features based on BioBERT, adding a gradient inversion layer, and setting a classifier and a discriminator; in positive and negative training, a weighted cross entropy loss function is used for learning real label mapping of a text in positive training, and a noise suppression loss function is used for reducing the influence of noise on model learning in negative training. According to the method, the delirium-symptom-containing medical record can be accurately identified, and the accuracy and robustness of medical text classification are improved.

The invention discloses a cell fate distinguishing and differentiation prediction method based on a three-dimensional form, and belongs to the technical field of cell fate distinguishing and differentiation prediction. In order to solve the problem that cell fate differentiation is difficult to distinguish and predict under a small sample size, the method comprises the following steps: acquiring a living biological sample, and imaging by using a three-dimensional delay microscope to obtain cell three-dimensional image data; segmenting the cell membrane boundary and extracting geometric and high-order three-dimensional morphological characteristic parameters; labeling cell fate and differentiation states based on means such as cell lineage tracking or gene fluorescence labeling; and finally, determining an optimal morphological index and a minimum sample size through single-side double-sample t test, and applying the optimal morphological index and the minimum sample size to two new cell populations and outputting a cell fate distinguishing and differentiation prediction result. The method can realize cross-species, low-loss, low-cost and high-precision cell fate differentiation and differentiation prediction, and is suitable for development research, stem cell culture and disease diagnosis.

The invention discloses a case misdiagnosis identification method and device and a storage medium, relates to the technical field of medical technology and machine learning, and discloses the steps that multi-source heterogeneous medical detection data of a to-be-identified case and a disease diagnosis result and a treatment strategy of a doctor on the to-be-identified case are acquired; based on the multi-source heterogeneous medical detection data, the disease diagnosis result and the treatment strategy, diagnosis and treatment path matching is conducted on a pre-constructed typical disease knowledge base, the corresponding diagnosis and treatment path deviation degree is obtained, and the typical disease knowledge base comprises multiple disease entries and reference diagnosis and treatment paths corresponding to the disease entries; and inputting the multi-source heterogeneous medical detection data, the disease diagnosis result, the treatment strategy and the diagnosis and treatment path deviation degree into a misdiagnosis recognition model, and performing misdiagnosis recognition on the to-be-recognized case based on the misdiagnosis recognition model to obtain a corresponding misdiagnosis recognition result, thereby solving the technical problem of low misdiagnosis recognition accuracy of the actual diagnosis and treatment path of the case in the prior art, and improving the accuracy of the misdiagnosis recognition. And the misdiagnosis identification accuracy is improved.

The invention relates to a double-attention-enhanced drug prediction model MTDADP under a multi-task learning framework, and belongs to the technical field of medical artificial intelligence. According to the method, disease diagnosis detail information can be fully utilized, association between the disease diagnosis detail information and medical history can be deeply mined, and safer and more accurate personalized drug prescription recommendation can be realized. The method comprises the following steps: firstly, performing embedded representation on historical treatment records, current symptoms and test details of a patient by the MTDADP; the state representation of the patient is enhanced through double mechanisms of access level attention and feature dimension attention; then constructing a patient memory bank and carrying out two-channel retrieval, and comprehensively utilizing drug distribution information of similar and dissimilar patients; then, the safety of drug representation is enhanced through a graph convolutional network by using an EHR graph and a DDI graph; and finally, integrating multi-source information to carry out joint prediction, introducing disease diagnosis prediction as an auxiliary task, and carrying out optimization through a multi-objective loss function. Experimental results on a disclosed PIC data set show that the MTDADP is remarkably improved in Jaccard, F1 and PR-AUC indexes compared with an existing advanced baseline model, and the effectiveness of the method in the aspect of improving the drug prediction accuracy is verified.

The invention discloses an intelligent inquiry method and device and electronic equipment, and relates to the technical field of intelligent inquiry and the technical field of data processing.The intelligent inquiry method comprises the steps that symptom information provided by a patient in the current round of inquiry is obtained, and the symptom information serves as latest symptom information; generating a current candidate disease set based on the latest symptom information; for each candidate disease in the current candidate disease set, calculating the posterior probability of the candidate disease after the current round of inquiry based on the latest symptom information by adopting a Bayesian algorithm; based on the posterior probability of each candidate disease in the current candidate disease set, the uncertainty entropy of the current candidate disease set is calculated, and the uncertainty entropy represents the uncertainty of the disease diagnosis result of the current candidate disease set; and determining whether to end the intelligent inquiry or not based on a size relationship between the uncertainty entropy and a preset entropy threshold value. By adopting the scheme, the inquiry efficiency and the inquiry integrity in intelligent inquiry are effectively balanced.

The invention discloses a laying hen genetic disease knowledge graph construction and intelligent decision support system. The system comprises a genetic disease weak supervision graph extraction module, a disease multi-factor diagnosis module, a group health risk clustering module, a multi-omics graph analysis module, a genetic disease knowledge graph construction module and an intelligent decision support module. According to the system, laying hen genes, physiological indexes, breeding environments and multi-omics data are collected through gene sequencing and the like, a gene disease association graph is generated through weak supervision graph extraction, a disease diagnosis result is generated through multi-factor diagnosis, health risk grades are divided through risk clustering, and an association graph is generated through multi-omics analysis; and a genetic disease knowledge graph is constructed, and an intelligent decision scheme is generated in combination with real-time data and an algorithm. The system improves the accuracy of laying hen genetic disease prevention and control and the intelligent level of breeding management, guarantees the economic benefits of breeding, and is suitable for large-scale breeding scenes of laying hens.

The invention relates to the technical field of health monitoring, in particular to an electrocardiograph monitoring intelligent diagnosis system based on deep learning and multi-modal fusion technology, which is characterized in that a multi-modal synchronous sensing unit synchronously acquires ECG, PCG and PPG data streams with aligned timestamps, and an edge intelligent diagnosis unit operates a lightweight model with the parameter smaller than 1MB, so that the ECG monitoring intelligent diagnosis system is obtained. The cloud fusion analysis unit is used for dynamically fusing multi-source data by adopting a cross-modal attention mechanism to generate a ventricular fibrillation probability PVF, and the dynamic collaborative decision-making unit is used when R is equal to III or Cilt, and the cloud fusion analysis unit is used for dynamically fusing the multi-source data by adopting a cross-modal attention mechanism to generate a ventricular fibrillation probability PVF; deep diagnosis, PVFgt, is triggered within 5 seconds at 0.9; and when 0.95, a third-level audible and visual alarm is triggered, and the closed-loop optimization unit continuously optimizes the model by distilling and compressing the misdiagnosis sample. According to the intelligent diagnosis system for electrocardiograph monitoring, rapid edge diagnosis is achieved through a lightweight model, critical cases can respond rapidly through cloud cooperation, and the accuracy and timeliness of heart disease diagnosis are improved.

The invention provides a temporal-mandibular joint disease diagnosis method and system based on a large language model, and the method comprises the steps: collecting text data of a temporal-mandibular joint disorder syndrome, carrying out the preprocessing of the text data, obtaining a temporal-mandibular joint disease professional data set, and employing a pre-trained large language model to extract metadata; segmenting the extracted metadata into text blocks by adopting a self-adaptive semantic partitioning algorithm, and obtaining a set of the text blocks, namely a text block library; constructing a temporal-mandibular joint special disease knowledge map; according to the type of a user query statement and a knowledge graph, integrating direct semantic search and knowledge graph guide search by adopting a mixed search strategy, and performing knowledge retrieval and reasoning to obtain retrieval information; after retrieval information is input into the fine-tuned large language model, a diagnosis result is generated; according to the method, higher diagnosis accuracy can be achieved, intelligent auxiliary diagnosis of the temporomandibular joint disease can be achieved, and the accuracy and reliability of diagnosis of the temporomandibular joint disease can be improved.

The invention discloses a swallowing type cow rumen sensor and a disease monitoring and early warning system. Comprising a swallowing type dairy cow rumen sensor, a data receiving host and a cloud platform upper computer, the sensor is of a sealed columnar capsule structure, a pH detection assembly, a TDLAS technology-based methane gas sensing system, a temperature sensor, a three-axis attitude sensor and a vibration power generation unit are integrated in the sensor, and the sensor resides in a rumen through a counterweight; the cow behavior recognition module is used for processing the collected posture data in real time based on an edge calculation algorithm built in the main control module so as to realize cow behavior recognition; the data receiving host is configured with a LoRa wireless receiving module and a 4G communication module and is used for receiving and synchronously uploading data; and the cloud platform upper computer analyzes the uploaded multi-dimensional physiological feature data and behavior recognition results based on a deep learning algorithm so as to realize disease diagnosis and accurate feeding decision. The device solves the problem that the deep body sign of the dairy cow is difficult to obtain, and has the characteristics of small size, corrosion resistance, discharge prevention, self power supply and the like.

The invention provides a multi-agent collaborative semantic transformation framework for a rare disease recognition problem, and belongs to the technical field of artificial intelligence and medical image analysis. According to the method, the problems of scarcity of marked data and difficulty in cross-domain knowledge migration in rare disease recognition are solved. The method comprises the steps that S1, a multi-agent parallel network architecture is constructed, each agent is provided with a special semantic focusing module, and diversified features are extracted from different attribute perspectives; s2, a cooperative gating mechanism with dynamic temperature parameters is realized, and the cooperative and competitive relationship between intelligent agents is balanced in a self-adaptive manner; s3, applying a double-constrained cross-domain semantic alignment strategy to ensure that the converted semantic features are consistent with the original features and keep diversity at the same time; s4, adopting a progressive training strategy and a semantic consistency loss function to reduce an overfitting phenomenon in cross-domain knowledge migration; and S5, classifying and identifying rare diseases through multi-agent cooperation, and applying source domain knowledge to a target domain. The method is prominent in medical image application, the accuracy rate of identifying rare skin diseases by using only common skin disease data reaches 52.13%, and the method is remarkably superior to an existing method. The framework has wide applicability in cross-domain zero sample learning tasks, is particularly suitable for application in the field with definite definition of semantic attributes, and provides a new solution for diagnosis of rare diseases in medical images.

The present invention relates to a diagnosis assistance device and method based on artificial intelligence processing of a radiographic image, wherein a disease diagnosis assistance device according to an embodiment of the present invention may comprise an information providing unit that constructs a diagnosis assistance model by learning a training radiographic image and additional information through a first feature processing unit, a second feature processing unit, and a feature fusion processing unit, determines a bone mineral density value and disease classification information with respect to a radiographic image to be diagnosed on the basis of the diagnosis assistance model, and provides disease diagnosis assistance information including the bone mineral density value, the disease classification information, and diagnosis basis information.

The application discloses a similar case recommendation method and system, belongs to the technical field of clinical decision support system, and aims to solve the technical problem that the implementation difficulty of similar case recommendation is extremely great and the feasibility is not high by using a deep learning algorithm and a natural language processing technology.The technical scheme is as follows:the method is specifically as follows: data preprocessing: extracting the symptom and diagnosis information of the case admission record electronic medical record; and processing the case into a standard symptom and standard diagnosis list, and extracting the treatment information of the case for storage in a dictionary to assist in subsequent acquisition of the similarity degree of the case; wherein, the treatment information comprises age and department; obtaining the symptom weight: based on the knowledge graph of the symptom and disease related knowledge, extracting the weight of the symptom in the disease diagnosis; obtaining the similarity degree of the case: according to the symptom and diagnosis list of the target case, and fusing the weight to obtain the similarity degree of other cases and the target case; similar case recommendation.

The invention discloses a rapid magnetic resonance fingerprint reconstruction method and system based on K uniform hypergraph regularization, and relates to the technical field of magnetic resonance fingerprint imaging. The invention aims to solve the problems that the imaging quality is reduced due to data undersampling artifacts in magnetic resonance fingerprint imaging at present, and a non-local irregular anatomical structure is difficult to model and the calculation complexity is high in the existing method. The method comprises the following steps: (1) acquiring undersampled magnetic resonance fingerprint data; (2) constructing a K uniform hypergraph; (3) establishing a magnetic resonance fingerprint reconstruction model based on K uniform hypergraph regularization; (4) constructing an iterative solution algorithm based on an incremental sub-gradient approximation method; (5) setting an iterative solution algorithm termination condition; and (6) reconstructing the magnetic resonance fingerprint data and the quantitative parameter image by using an iterative solution algorithm. Experiments show that under the condition of high-power undersampling, a high-quality magnetic resonance fingerprint image can be reconstructed by using less sampling data, the reconstruction time is shortened, and a more reliable quantitative image basis is provided for clinical disease diagnosis and treatment monitoring.

The invention provides a multiple detection technology MMDEC for high-sensitivity detection of tumor methylated DNA from samples such as blood, belongs to the field of disease diagnosis markers, and more specifically relates to a method for detecting methylation of target DNA through the following steps: constructing oligonucleotide; the kit comprises a target specific multiple Rooptag primer capable of being complementarily combined with a plurality of tumor specific high-methylated CG island target DNAs (Deoxyribose Nucleic Acid) and a universal tag primer combined with the specific multiple Rooptag primer, wherein the target specific multiple Rooptag primer can be complementarily combined with a plurality of tumor specific high-methylated CG island target DNAs (Deoxyribose Nucleic Acid); the method comprises the following steps: performing exponential amplification on multiple methylation target DNA by using a specific multiple Rooptag primer as a primary primer to obtain a multiple amplification product; performing homogeneity amplification on the obtained multiplex amplification primer by using oligonucleotide (which can be complementarily combined with the linearly amplified target DNA) using a universal tag primer as a secondary primer; and detecting whether an amplification product exists or not through a probe. According to the method disclosed by the invention, MMDEC can be used for early screening and diagnosis of plasma / urine sample tumors, accords with clinical practice on the basis of research results, and has a good clinical application prospect.

The invention discloses a multi-modal magnetic resonance image and pathological tissue slice fusion accurate diagnosis system and method, and relates to the technical field of medical image processing, and the system comprises a data collection module which collects multi-modal magnetic resonance image data and pathological tissue slice data of a target object; the data preprocessing module is used for performing denoising, registration and normalization processing on the acquired multi-modal magnetic resonance image data and performing image segmentation, digitization and feature labeling processing on the acquired pathological tissue slice data; according to the method, the multi-time-point and multi-mode magnetic resonance image data are collected and subjected to self-adaptive space-time fusion with pathological tissue slice data to construct the disease dynamic evolution model, the states of diseases at different time nodes can be reflected, more comprehensive dynamic information support is provided for disease diagnosis, and the disease diagnosis accuracy is improved. The problem that diagnosis key features are covered due to lack of time sequence data integration capability is effectively avoided, so that the technical requirements of accurate diagnosis are met, and the accuracy and reliability of disease diagnosis are improved.

The embodiment of the invention provides a sample analyzer and a sample analysis method. The sample analyzer comprises a detection device, a display device and a control device, the detection device is used for obtaining a detection result of a sample in one or more items, the display device is used for displaying a first display interface, and the control device is used for responding to an operation of selecting a target sample on the first display interface by a user and controlling the first display interface to display an item detection result of the target sample and character abnormity prompt information of the target sample. On one hand, the character abnormity prompt information is directly presented on the first display interface, jumping to a second-level page is avoided, and the information viewing mode is simpler and more visual; and on the other hand, the item detection result is displayed in combination with the sample character abnormality prompt information, so that an observer can know the sample character abnormality while viewing the item detection result of the target sample, thereby providing more comprehensive clinical information for disease diagnosis and assisting medical personnel in disease diagnosis.

The application discloses an adaptive clustering federated learning modeling method for precision medicine, and relates to the technical field of precision medicine, and comprises the following steps: data collection and preprocessing and model construction and training; the application accurately determines the optimal number of the global model through adaptive clustering, and divides clustering groups according to the similarity of the client model parameters, realizes independent training of the groups, effectively improves the adaptability of the model to different client data characteristics, samples and label distribution differences, avoids the problem that a single model has poor performance in some clients, and significantly enhances the ability of the model to capture complex medical patterns; the model generalization is optimized through grouped training, so that the model can better cope with new data distribution, under the premise of ensuring the privacy and security of medical data, the accuracy and reliability of the model in precision medical scenes such as disease diagnosis and prognosis prediction are greatly improved.

The present application relates to the field of biological medicine, and particularly relates to biomarkers for migraine diagnosis, kits and application thereof. The biomarkers related to migraine disease according to the present application include COL4A2, and also include any one or a combination of multiple of MMP-14, LCAT, ADAMTS13 and CPXM2. The present application is based on proteomic data of clinical healthy controls and migraine patients, and uses bioinformatics analysis and machine learning methods to deeply mine protein combinations with the most early warning value for migraine disease diagnosis, and to verify in a clinical cohort with expanded samples, to provide a good prospect for clinical transformation of newly discovered biomarkers, and also to lay a foundation for subsequent mechanism research.

The application discloses a color fundus image classification system and method based on virtual multi-modal technology, and a classification model, and comprises a multi-modal generation module, a modal specificity feature calibration module and a modal dependency feature calibration module.The core of the application is integration of generation and diagnosis and self-adaptive calibration of features, and two tasks of image synthesis and disease diagnosis which are traditionally separated are integrated under a unified optimization target.Specifically, parallel deep neural networks are independently trained, and then learned representations are transferred to a multi-modal generation module in the form of corresponding virtual modal feature layers, and after the virtual modal feature layers generated by the multi-modal generation module are spliced and fused, self-adaptive calibration is performed in the modal feature space and between the modal feature spaces, so as to reduce redundancy and support flexible and information-rich feature integration.

The invention discloses a segmentation modeling method for improving bronchial branch identification precision, and belongs to the technical field of medical image processing and respiratory tract three-dimensional reconstruction. According to the method, high-resolution acquisition is performed on a respiratory tract CT image, and preliminary segmentation of a tracheal tree is completed in combination with gray threshold setting and a region growing algorithm; and a reference line is introduced at the bifurcation, and dynamic region growth based on density gradient is adopted, so that refined identification of the sixth-level bronchial branch is realized. And then carrying out iteration correction on the generated model for multiple times, manually comparing anatomical features to delete pseudo branches and complement missing branches, and completing smoothing and NURBS curved surface construction in Geomagic software. And finally, splicing and fusing the bronchial branch model and an independently generated upper respiratory tract model to obtain a complete three-dimensional respiratory tract combination model, and ensuring that the error between the three-dimensional respiratory tract combination model and an original CT image is controlled within 0.8 mm through deviation analysis. According to the method, the recognition precision of the fine bronchial branches can be remarkably improved, the defect of misrecognition or missing division of traditional threshold segmentation is overcome, and the obtained three-dimensional model can be used for aerosol drug deposition simulation, respiratory system disease diagnosis, surgical planning and design and evaluation of personalized inhalation devices.