34 results about "Rare disease" patented technology

The invention provides a rare disease risk screening model training system and method, a screening system and a medium, and belongs to the technical field of medical information. According to the invention, an innovative three-stage architecture design is adopted, unified identification of all rare disease types is realized, and the blank in the prior art is filled. When only basic clinical information exists, a rare disease high-risk patient can be accurately identified, the diagnosis time is remarkably shortened, and a treatment precedent is won for the patient. The system adopts a small parameter quantity model integration strategy, reduces hardware requirements and maintenance cost, can be widely deployed in basic medical institutions, and promotes balanced distribution of medical resources. Through the technical measures of nested cross validation, algorithm comparison, threshold optimization and the like, the system has high accuracy and reliability, and the credibility of doctors to the system is enhanced.

The invention discloses a plateau newborn rare disease multi-mode combined screening system and a storage medium, and belongs to the technical field of medical software development. The plateau adaptability newborn rare disease multi-mode combined screening system is constructed, the plateau adaptability of the screening system is enhanced, the cost is remarkably reduced, and the screening system has a good application prospect.

The invention discloses a method and a system for assisting in inputting clinical information of rare diseases and analyzing gene mutation based on phenotypes. The method comprises the following steps: firstly, acquiring clinical information in voice, text and image forms of a patient through a multi-source data acquisition module, converting the clinical information into characters, and performing entity recognition and standardization processing to generate structured medical record data; secondly, extracting clinical phenotypes from the structured data; furthermore, a candidate gene list is obtained according to the gene-disease relationship, comprehensive scoring and sorting are carried out, and a concerned gene list is output. According to the method, efficient structured input and standardization of clinical information are realized, the accuracy and automation level of phenotype-gene matching are remarkably improved, the gene variation interpretation period is effectively shortened, and intelligent support is provided for precise diagnosis of genetic diseases.

The invention provides a multi-modal fusion Bayesian medical auxiliary diagnosis method and system, belongs to the technical field of medical auxiliary diagnosis, and is used for solving the problems of inquiry redundancy, insufficient adaptation of small sample / emergency treatment / rare disease scenes and large disease probability calculation deviation in related technologies. According to the method, a knowledge graph containing disease-symptom exclusion weight and a patient database are constructed, intelligent inquiry is combined with regional correction, core symptoms are screened through real-time evidence, the disease probability is calculated through multi-modal data processing and Bayesian reasoning, and tool collaborative verification and scene adaptation optimization output are carried out. The system comprises a core calculation unit, a multi-modal acquisition unit, a database unit, a detection interaction unit and a result display unit for supporting method execution. The method can improve the inquiry efficiency and diagnosis precision, is suitable for multiple clinical scenes, and provides reliable assistance for doctors.

The invention relates to an information interaction method and device for a rare disease, a storage medium and electronic equipment, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining the query information of a target user for the rare disease; according to the query information, determining an initial query intention of the target user for the rare disease through a preset general intention model; extracting a plurality of actual keywords from the query information, and adjusting the initial query intention according to each actual keyword to obtain a target query intention; and determining associated knowledge data corresponding to the target query intention according to a preset general rare disease knowledge base, generating reply information according to the query information and the associated knowledge data, and sending the reply information to a terminal of the target user. The method has the effect of improving the matching between the rare disease related information queried by the user and the query demand of the user.

The application discloses an information processing method and device for medical image classification, equipment and medium, relates to the technical field of image recognition, and comprises the following steps: dividing a plurality of preset disease categories into a preset head category and a preset tail category; clustering each preset disease category in the preset tail category into a plurality of abnormal disease categories, and determining third training data corresponding to each abnormal disease category based on second training data corresponding to each preset disease category in each abnormal disease category; determining a target data set based on the first training data, the preset disease category corresponding to the first training data, the third training data, and the abnormal disease category corresponding to the third training data; training an initial classification model based on the target data set to obtain a target classification model, and generating a target disease classification result of an initial medical image to be recognized by using the target classification model. The application can efficiently process long-tail distribution data of medical images, and improve the recognition accuracy and reliability of rare diseases.

The application provides a rare disease occurrence probability prediction method and device and an electrocardiogram analysis system. The method comprises the following steps: acquiring a plurality of to-be-evaluated electrocardiogram data and corresponding to-be-evaluated medical record text data; inputting the to-be-evaluated electrocardiogram data into an electrocardiogram pre-training model to obtain a target electrocardiogram feature vector; inputting the medical record text data corresponding to the to-be-evaluated electrocardiogram data into a medical text pre-training model to obtain a target text feature vector; inputting the to-be-evaluated electrocardiogram data into a preset risk detection model to obtain an electrocardiogram rare disease occurrence probability coefficient; presetting a rare disease feature dictionary; determining the occurrence probability of each rare disease according to the similarity between the target electrocardiogram feature vector and the reference text feature vector of each rare disease, the similarity between the target text feature vector and the reference text feature vector of each rare disease, and the electrocardiogram rare disease occurrence probability coefficient. The occurrence probability of the rare disease can be predicted without training a large number of electrocardiogram samples of the rare disease.

A learning device includes a memory and processing circuitry configured to register at least inspection data and medical care data of a patient who has developed a rare disease from a plurality of medical institutions perform predetermined preprocessing on inspection data and medical care data of a patient estimate an onset probability of an estimation target patient for each of a plurality of rare diseases based on the inspection data and medical care data of the estimation target patient after the preprocessing by using an estimation model that estimates an onset probability for each of the plurality of rare diseases and use at least the inspection data and the medical care data of the patient who has developed a rare disease after the preprocessing as learning data, and cause the estimation model to learn a relationship between the inspection data and the medical care data and an onset probability.

This application discloses a retrieval model training method, a retrieval-enhanced generation-based Bert-Hogg-Dubbs syndrome (BHD) identification method and system, relating to the field of rare disease identification. The training method includes acquiring positive and negative sample pairs; inputting the positive and negative sample pairs into an initial retrieval model; calculating the loss function of the initial retrieval model; and dynamically adjusting the angle margin of the loss function in real time according to a metric variance adaptive mechanism. The initial retrieval model is then optimized based on the loss function calculation results. This application forcibly expands the angle interval between BHD and non-BHD by using the angle margin of the loss function, and uses a metric variance adaptive mechanism to dynamically adjust the angle margin based on the statistical variance of the cosine similarity among all positive sample pairs in the current training batch. This solves the problem of weak image differences and blurred category decision boundaries in DCLDs caused by the highly similar imaging features of various rare diseases, thus improving the recognition accuracy of large models for query information.

The present disclosure provides a vertebral rare disease vertebral body morphology analysis processing method, system and device, the method comprising: acquiring a plurality of sample vertebral body images under a target scene; performing segmentation processing on each sample vertebral body image to obtain a plurality of sample segmentation results of vertebral bodies; acquiring a plurality of dimension sample morphology representation parameters corresponding to each vertebral body; performing clustering processing based on different sample morphology representation parameters to obtain different clustering ranges to acquire target classification association information for distinguishing different vertebral body morphologies of vertebral rare diseases. In the present disclosure, a multi-segmentation model fusion segmentation processing is adopted for any one unlabeled sample vertebral body image to obtain the optimal segmentation area of any one vertebral body; clustering processing is performed based on the sample morphology representation parameters of each vertebral body to quickly and accurately obtain relevant information for distinguishing different vertebral body morphologies of vertebral rare diseases, thereby ensuring the accuracy and efficiency of the analysis and processing of vertebral body morphology.

The present application relates to a kind of medical entity identification and knowledge base alignment method for rare disease scene, belong to medical data processing technical field, solve the weak rare disease recognition ability in existing medical entity identification, serious mis-matching problem, single matching strategy, knowledge base alignment difficulty and lack of explainability problem.It includes obtaining medical text to be processed, the semantic analysis of the medical text, identify drug entity and disease entity, and based on the identified entity, obtain the candidate entity name list;The candidate entity name list is input into the pre-constructed medical standard terminology knowledge graph, and structured verification is carried out using a four-stage cascade matching strategy, including: for the current candidate entity, sequentially matching from the first strategy, if the current level strategy successfully matches the entity name in the knowledge base, then output the matching result and terminate subsequent matching, otherwise enter the next level strategy.Medical entity identification and knowledge base alignment are realized.

The invention discloses application of a DNAH9 composite heterozygous mutation site in identification of a rare disease primary ciliary dyskinesia 40 type companion inversion, and belongs to the technical field of biology. According to the invention, DNAH9 composite heterozygous mutation sites (including c.11845G and c.308del) are used as biomarkers for identifying the rare disease primary ciliary dyskinesia type 40 accompanying inversion, a detection reagent and a kit are developed, and rapid screening and diagnosis of primary ciliary dyskinesia type 40 accompanying inversion gene mutation can be assisted.

Comprehensive and high-quality disease dictionaries are invaluable resources for tasks such as building ontologies, automated relation extraction, text summarization, question answering etc. Such curated resources are useful to clinicians, researchers, and various Biomedical Natural Language Processing tasks. However, these are manually curated and are labor and time intensive, and additionally suffer from lower recall and coverage is also less. Present disclosure provides systems and methods for augmenting rare disease dictionaries, wherein the system retrieves (new) rare diseases terms from medical literature that are related to the given dictionary terms (seed terms) and recommends new terms (or NPs) in a ranked order. This method is useful for rare diseases dictionary augmentation as a significant fraction of the top recommendations are new synonym candidates for dictionary augmentation. The method uses syntactic and semantic similarity measures in combination with efficient nearest neighbor search for efficient retrieval.

The invention discloses a method for evaluating and recommending a treatment scheme of a rare patient. The method comprises the following steps: acquiring data including structured data, semi-structured data and unstructured data; carrying out unified indexing on the integrated multi-modal data; setting a unified time grid to divide the multi-modal data, and mapping each observation data to the latest time according to timestamp information; carrying out feature extraction on the multi-modal data, unifying the modal features, splicing, and carrying out dimensionality reduction on the unified modal features through a linear projection layer to obtain a multi-modal fusion sequence; taking the multi-modal fusion sequence as input, and performing risk prediction on an output vector corresponding to a time step in the time sequence state matrix of the complete sequence through a prediction function; performing forward propagation on the same input by using a Monte Carlo method to obtain a plurality of groups of prediction results; the completeness of patient state features and the risk prediction precision are improved, and the problems that traditional reinforcement learning is prone to over-fitting and unstable in training are solved; and the purposes of transparency and verifiability of the result are achieved.

The invention relates to a rare disease identification method and system based on multi-scale vision-language prompt multi-instance learning, and the method faces a full-view digital pathological image (WSI), and introduces the prior knowledge of the pathological field under the weak supervision condition of only using a slice-level label, so as to identify the rare disease. Constructing a multi-scale vision-language collaborative prompt mechanism: generating descriptive text prompts corresponding to image blocks of different scales by utilizing a large language model, designing a prototype-guided image block decoder to aggregate massive image block features, and meanwhile, realizing cross-modal alignment in combination with a context-guided text decoder; the method simulates the diagnosis process of a pathologist, achieves high-precision classification under small samples, remarkably improves the recognition accuracy of rare diseases, has good cross-center generalization ability, supports interpretable evidence output, does not need large-scale data labeling or de novo pre-training, and facilitates clinical rapid deployment.

The invention discloses an incremental tuning and knowledge injection method and system for few samples / rare diseases, and relates to the field of continuous learning of a large language model (LLM) and medical artificial intelligence. In order to solve the technical problem that catastrophic forgetting is easily generated when an existing model is subjected to incremental tuning in small-sample and high-value fields (such as rare diseases), the invention provides an optimization strategy based on PEFT (Parameter Efficient Fine Transfer) and knowledge integration loss. The method comprises the following steps: initializing a PEFT module and adding the PEFT module to a pre-training LLM; taking few-sample data labeled by experts as new task data; constructing a dual loss function containing new task loss and knowledge distillation loss (measuring the difference between the output of the new model and the output of the basic model); only by optimizing the parameters of the PEFT module, the basic universality is protected from being damaged to the maximum extent while it is guaranteed that the model quickly masters rare disease knowledge. According to the invention, the efficiency and the stability of the medical LLM in a continuous updating scene are remarkably improved.

The invention discloses a disease auxiliary diagnosis system based on an interactive symptom union set. The disease auxiliary diagnosis system comprises a user interaction layer, a core processing layer and a data storage layer, the user interaction layer is used for realizing interaction between a patient and the system, receiving complaint information input by the patient and displaying symptom collection guide information, a preliminary diagnosis result and matched doctor information to the patient; meanwhile, personalized demand information submitted by the patient is received. According to the method, the pre-trained BERT-BiLSTM-CRF medical entity recognition model is adopted, so that the unstructured text of the complaint of the patient can be accurately analyzed, and the medical entity and the modification relationship thereof are extracted; meanwhile, symptom information is supplemented through interactive guidance, a complete symptom union set is constructed, the problem that in the prior art, the generalization ability of fuzzy or rare disease description is poor is effectively solved, and accurate data support is provided for follow-up disease diagnosis.

This invention relates to the field of medical artificial intelligence technology, and in particular to an AI-assisted diagnostic decision-making system for rare bladder diseases. The system includes modules for data acquisition, feature extraction and quantification, an AI rule-based reasoning engine, error warning, and treatment decision output. Its core lies in using natural language processing technology to transform unstructured clinical text into feature tags. The AI ​​engine then executes a multi-level identification process of "excluding infection – rapid screening of medical history – precise matching," outputting a list of candidate diseases sorted by probability. Simultaneously, the system can automatically warn of clinical errors such as antibiotic abuse and overtreatment, and generate a comprehensive report including primary care treatment plans and referral criteria. This invention combines a rule engine with intelligent reasoning, providing primary care physicians with a standardized and easy-to-use auxiliary diagnostic tool, which helps improve the efficiency and standardization of rare disease diagnosis and has high clinical translational value.

The present application relates to the technical field of disease prediction, and discloses a phenotype analysis method for disease prediction, which comprises the following steps: constructing a database, determining the phenotype similarity of rare diseases and common diseases, referring to the difference of gene data of rare diseases and common diseases, constructing a "rare disease-common disease" common database, processing patient data to find the best combination of phenotype characteristics, calculating the disease matching score, calculating the cross-entropy loss of the phenotype characteristic matching model of disease prediction, taking the weighted sum of the classification loss function in the "rare disease-common disease" common database and the cross-entropy loss of the phenotype characteristic matching model as the total loss function of the disease prediction model, outputting the prediction result, extracting effective phenotype characteristics based on a graph convolution network for data comparison of rare diseases and common diseases, and extracting disease matching difference from the commonality of rare diseases and common diseases to predict diseases, which can solve the problem of easy confusion between diseases and to a certain extent, avoid misdiagnosis and missed diagnosis of rare diseases.

This application discloses a device and storage medium for calculating incidence rates based on medical record data. The device includes: a data acquisition module for acquiring medical record data from multiple patients; a data filtering module for filtering the medical record data from multiple patients to obtain medical record data from a target patient suffering from a rare disease; a data prediction module for analyzing the number of second inpatient cases provided by a second category of medical institutions based on the number of first outpatient cases and first inpatient cases provided by a first category of medical institutions to obtain a prediction result of the number of second outpatient cases suffering from rare diseases in the second category of medical institutions; and a first data determination module for determining the incidence rate of the rare disease based on the medical record data of the target patient and the prediction result. Thus, even in the case of missing medical data, a relatively reliable incidence rate estimation scheme can still be achieved, thereby improving the accuracy of the final obtained incidence rate of rare diseases.

The application relates to the field of intelligent medical treatment, in particular to a method, equipment and program product for identifying rare intestinal diseases. The method comprises the following steps: acquiring pathological data of a to-be-detected person; the pathological data comprises goblet cells, Paneth cells, neutrophilic granulocytes and plasma cells; the type of the rare intestinal disease of the to-be-detected person is determined based on the pathological data, the type of the rare intestinal disease comprises autoimmune enteropathy and common variant immunodeficiency; when any two or more of the following conditions are met, namely, the number of goblet cells is less than a preset threshold, the number of Paneth cells is less than a preset threshold, neutrophilic granulocyte infiltration occurs, and the number of plasma cells remains unchanged, the to-be-detected person is determined to have autoimmune enteropathy; and when the number of plasma cells is less than a preset threshold, the to-be-detected person is determined to have common variant immunodeficiency. The application can effectively distinguish the types of rare intestinal diseases and has good clinical value.

An object of the system according to the embodiment is to support identification of a disease from a medical interview result or a medical examination result.SOLUTION: A system includes an inquiry result analysis part, a medical examination result analysis part, an intractable disease specification support part, and a nursing research support part. An inquiry result analysis part analyzes the inquiry result and lists possible diseases. The medical examination result analysis part analyzes the medical examination result and supports the specification of a disease. The incurable disease specification support unit supports specification of an incurable disease. The nursing research support unit supports nursing research and case research.SELECTED DRAWING: Figure 1

The invention discloses an enhanced generation method and system for a virtual case of a rare disease, and the method comprises the following steps: S1, extracting core features of the rare disease from an authoritative database through an RAG technology, carrying out the standardization, and screening a matched basic case template for subsequent transformation; s2, non-specific features of the basic case are replaced with rare disease typical expressions based on a rule engine, a disease course timeline is reconstructed, and a structured case framework is generated through medical logic verification; and S3, generating a complete case text by utilizing the medical large model iteration, performing error analysis and knowledge base optimization on substandard cases in combination with independent medical model verification, and outputting high-quality teaching cases. According to the method, the accuracy and timeliness of basic medical knowledge are ensured, and a self-optimization closed loop is formed through a large model iteration generation strategy and independent medical model verification based on error analysis and knowledge base optimization.

The application discloses a kind of based on multimodal fusion's rare disease diagnosis and treatment resource collaborative scheduling method and system, the method includes: extracting multidimensional feature from multimodal data, based on multidimensional feature is adapted score calculation, based on comprehensive adaptation score from candidate diagnosis and treatment scheme Screening out recommended scheme, and matching required medical resources, through to patient state and resource state monitoring is adjusted and resource collaborative scheduling is carried out diagnosis and treatment path;The present application fully excavates the complex interaction between patients, diagnosis and treatment scheme and medical resources by multimodal fusion technology, improves the accuracy of comprehensive adaptation evaluation with the help of special model, combined with cross-agency collaborative scheduling and dynamic adjustment mechanism, effectively solves the problem of insufficient utilization of multimodal data, insufficient scheme adaptability, inefficient resource scheduling and lack of flexible adjustment of diagnosis and treatment path, significantly improves the individualization level of rare disease diagnosis and treatment, resource utilization efficiency and timeliness and rationality of diagnosis and treatment decision.

The invention belongs to the technical field of medical intelligent analysis and data mining, and particularly relates to a label clustering and classifier chain-based unbalanced multi-disease risk collaborative prediction method, which comprises the following steps of: dividing high-frequency co-occurrence diseases into the same label cluster by adopting a label clustering strategy based on disease correlation; for each label cluster, constructing a classifier chain structure by using mutual information; calculating the imbalance rate of each disease label in the classifier chain, and dynamically selecting an optimal imbalance processing strategy according to the imbalance rate; and integrating the classifier chain corresponding to each label cluster to realize multi-disease risk collaborative prediction of a new sample. According to the method, potential correlation between diseases can be fully mined and utilized, and meanwhile, a dynamic unbalance adaptive method is provided in a multi-disease prediction model training stage, so that the prediction performance of the model on rare diseases is enhanced, and intelligent decision support is provided for doctors.

This application provides a data analysis method and apparatus based on multidimensional feature similarity. In this application, by calculating the weighted similarity of multidimensional features, the most similar historical cases to the current patient are found. Then, based on a first prior probability, a first likelihood probability, a second prior probability, and a second likelihood probability, the posterior probability of the patient's condition and each target historical case is predicted and determined, i.e., the degree of similarity between the patient's condition and each target historical case. This determines the matching probability between the patient's condition and each target historical case. Through this method, historical cases can be effectively utilized, i.e., historical cases can be used as diagnostic references, improving interpretability. Simultaneously, it is also beneficial for predicting rare diseases and reducing the probability of misdiagnosis.

A processing system and method for rare-disease medical data are provided, where a server, based on a medical data analysis request initiated by a target client, generates global model parameters and an initial analysis model, and sends the global model parameters and the initial analysis model to a plurality of clients; each client trains the initial analysis model based on client-side medical data of the client, and sends local model parameters and loss function values of the local analysis model to the server; the server inputs a plurality of the local model parameters and the loss function values received in a current round into a particle swarm optimization model, so as to obtain optimized model parameters output by the particle swarm optimization model; the server determines whether the optimized model parameters meet a preset condition, wherein if no, the optimized model parameters are sent to each client.

The application provides a rare disease data sharing system and method, and relates to the technical field of medical treatment. The sharing system comprises a server and a plurality of clients respectively deployed in different medical institutions and different departments. The server and the clients are connected to form a medical data platform based on a federated learning mechanism. The server searches based on case characteristics uploaded by each client, determines case characteristics extracted from a plurality of medical data of a patient to be analyzed, and then analyzes through a rare disease analysis model to assist doctors in judging the disease. The privacy of the patient's medical data can be ensured, and the patient does not need to manually organize or provide previous case materials, which provides convenience for doctors and patients.

The invention discloses a rare disease diagnosis and treatment resource collaborative scheduling method and system based on multi-modal fusion, and the method comprises the steps: extracting multi-dimensional features from multi-modal data, carrying out the adaptive score calculation based on the multi-dimensional features, screening out a recommendation scheme from candidate diagnosis and treatment schemes based on a comprehensive adaptive score, and carrying out the matching of needed medical resources, diagnosis and treatment path adjustment and resource cooperative scheduling are carried out by monitoring the patient state and the resource state; according to the method, complex interaction association among patients, diagnosis and treatment plans and medical resources is fully mined through a multi-modal fusion technology, the accuracy of comprehensive adaptation evaluation is improved by means of a special model, and cross-mechanism collaborative scheduling and a dynamic adjustment mechanism are combined; the problems of insufficient utilization of multi-modal data, insufficient scheme adaptability, low efficiency of resource scheduling and lack of flexible adjustment of diagnosis and treatment paths are effectively solved, and the individuation level of rare disease diagnosis and treatment, the resource utilization efficiency and the timeliness and rationality of diagnosis and treatment decision are remarkably improved.