41 results about "Outcome prediction" patented technology

The invention relates to the technical field of computers, and discloses a legal case strategy analysis method and system, electronic equipment and a storage medium, and the method comprises the steps: analyzing a to-be-analyzed legal case, extracting key information, and constructing a dynamic affair graph based on the key information; matching with a historical case database by adopting a multi-dimensional class case retrieval method on the basis of the dynamic event atlas, and retrieving similar historical cases; based on the judgment result information of the similar historical cases, performing analysis in combination with key information or a dynamic affair map of the to-be-analyzed legal case, and generating judgment result prediction of the to-be-analyzed legal case; and fusing the dynamic affair atlas, similar historical cases and judgment result prediction, calling a large language model adjusted by a law field instruction through a retrieval enhancement generation architecture, and generating a litigation policy analysis report with a preset structure part. According to the method, the automation degree and depth of case information processing are improved, and closed-loop generation from analysis to strategy is realized.

One or more systems, devices, computer program products and / or computer-implemented methods of use provided herein relate to predicting an optimized result for a graph neural network (GNN). A system can comprise a memory configured to store computer executable components; and a processor configured to execute the computer executable components stored in the memory, wherein the computer executable components comprise: a fusion component that that models non-linear modality correlations within and across entities through Hirschfeld-Gebelein-Re'nyi maximal correlation (MaxCorr) embeddings that generates a multi-graph that preserves identities of modalities and entities; and a multi-graph neural network (MGNN) component for task-informed reasoning in multi-graphs, that learns parameters defining entity-modality graph connectivity and message passing in an end-to-end fashion.

The invention relates to a fusion prediction and interpretable report generation system for multi-modal medical data, and the system comprises a data access and preprocessing module which is used for processing at least two kinds of heterogeneous modal data from a target object; the modal feature encoding module comprises a plurality of encoders corresponding to modal types and is used for extracting feature representation of each modal data; the outcome prediction module is used for outputting an outcome prediction result of each mode; the missing modal fusion module is used for dynamically adjusting a fusion weight based on a modal existence vector when at least one modal data is missing so as to realize robust prediction; and the multi-agent interpretation module comprises a plurality of interpretation sub-modules corresponding to the modals and a language model convergence device, and is used for generating an interpretable natural language report based on the multi-modal evidence. According to the method, the problems that an existing prediction technology is unavailable under the condition of lack of modals, the fusion capability is insufficient and clinical interpretability is lacked are solved, and a stable and interpretable auxiliary decision-making tool is provided.

The invention discloses a Crohn disease postoperative bad outcome prediction method combining knowledge base construction and an experience-driven self-asking mechanism, and solves the problems of insufficient interpretability, insufficient clinical data utilization, weak generalization ability and the like of an existing prediction method. The method comprises the following steps: constructing a Crohn disease special medical knowledge base, extracting disease progress related factors from medical literatures and real medical records, and carrying out expert score weighting processing to form searchable knowledge entries; retrieving related knowledge injection context from a knowledge base based on a medical record input by a user, and generating preliminary prediction by using a large language model; structured reasoning is guided through a multi-layer prompt engine, diagnosis tasks are decomposed through the thinking chain technology, and self-check before reasoning is achieved by driving a self-asking mechanism through experience. And generating an introspection problem chain according to historical error cases, iteratively optimizing the reasoning process, and finally outputting a bad outcome prediction result and a detailed analysis report. According to the method, both traceable interpretability and prediction accuracy are considered, and reliable assistance is provided for postoperative clinical management of Crohn's disease.

The invention relates to the technical field of medical data analysis, in particular to a hepatic fibrosis outcome prediction and intervention recommendation method based on dynamic data, which comprises the following steps: collecting biological samples and clinical index data from a target individual at a plurality of continuous time points, and constructing a multi-modal time sequence data set; processing the data set to extract dynamic features reflecting a change process, wherein the dynamic features comprise trend features and event features; inputting the baseline features and the dynamic features of the target individual into a pre-trained hepatic fibrosis outcome prediction model, and outputting hepatic fibrosis index prediction values at a plurality of time points in the future to form a prediction outcome trajectory; when the trajectory meets an early warning condition, matching candidate donors from a pre-constructed donor feature library and generating an intervention recommendation list based on the current state and trajectory features of the target individual; and when new time point data is obtained, updating the data set and repeating the process. According to the method, dynamic and quantitative prediction of the hepatic fibrosis outcome trajectory is realized, and data-driven personalized intervention recommendation is provided.

A system and method for generation and use of radiation outcome prediction (response & side effects) score for patients undergoing radiotherapy for various medical conditions wherein the score is a personalized score, provided by analyzing multiple parameters including the tumor specific, patient specific, gene specific and treatment planning specific parameter(s), during and post therapy.

The present application relates to the medical technical field, specifically relates to a kind of early postoperative adverse outcome prediction method for craniopharyngioma patient, comprising: extracting preoperative data and label data of patient, store in database;The data in database is preprocessed, is randomly divided into training set and verification set according to preset proportion, and the training set is handled using the minority class oversampling technique, obtain the training set after processing;From the training set after processing, the most predictive value feature subset is identified and screened;Predictive model is constructed;The predictive model is evaluated and compared, and the best model is selected as the final deployment model;Receive new preoperative data of patient, input final deployment model, output prediction result.The present application fills the blank of existing prediction craniopharyngioma postoperative early overall adverse outcome model, overcomes the single data processing method in existing prediction technology, feature selection is not accurate enough, prediction accuracy is not high and lacks clinical usability and other problems.

The invention relates to the technical field of biology, in particular to 5-hydroxymethylcytosine and application of a prognosis model of 5-hydroxymethylcytosine in nasopharynx cancer prognosis evaluation. According to the method, the prognosis scoring model capable of layering the outcome of the patient is constructed, and the model shows strong distinguishing performance in a training set and a verification set. When it is integrated into a prognostic model together with tumor staging and EBV status, the 5hmC score results in higher calibration accuracy and net clinical benefit in decision curve analysis. According to the findings, a 5hmC prognosis model is determined as a noninvasive marker rich in biological information, can be used for prognosis risk layering of NPC patients, and has potential significance for individualized management and improvement of outcome prediction.

The invention relates to the technical field of artificial intelligence and psychological health assessment, and particularly discloses a psychological health longitudinal tracking and effect prediction method based on a time sequence, which reflects the change process of the psychological state of an individual by utilizing longitudinal time sequence data in data acquisition and processing, and predicts the psychological state of the individual. In result prediction, future psychological health trend prediction is realized through a time sequence model, accuracy of diagnosis and intervention evaluation is improved by fusing subjectivity such as scale, inquiry and objectivity such as cognitive test, physiological signal data and past examination reports, and the method is suitable for various application scenes such as clinical follow-up visit, school psychological health management and online psychological counseling platforms.

The application provides a kind of legal judgment result prediction intelligent feedback error correction method and electronic equipment, including the secretary intelligent agent obtained by the training of large language model, the prosecution intelligent agent, the defense intelligent agent and the judge intelligent agent of sentencing, method includes: through secretary intelligent agent, event-level disassembly and extraction are carried out to case fact text, and fact atom set is obtained;Through the prosecution intelligent agent, the charge set is generated based on fact atom set, and through the defense intelligent agent, the refutation set is generated based on fact atom set;Through the judge intelligent agent of sentencing, the charge set and the refutation set are carried out to confrontation type collaborative reasoning, and initial sentencing result is obtained;The initial sentencing result is checked, if there is doubt charge, then the prosecution intelligent agent and the defense intelligent agent are used to generate correction content for doubt charge;Through the judge intelligent agent of sentencing, the final sentencing result is generated according to the correction content.Can assist judicial organ to realize more stable, interpretable and efficient legal judgment prediction.

System and methods for determining high-impact, non-informative features in datasets can include obtaining a first dataset including a first set of features and first set of tags, determining a second dataset including a second set of features and the first set of tags, the second set of features transformed from the first set of features, and each of the second set of features including a binary value, predicting, by a model, a third dataset including a third set of features and second set of tags based on the second dataset, comparing the third dataset to the second dataset to determine a correlation therebetween, based on the comparison, filtering at least one feature having missing feature values from the first dataset, and determining a training dataset with the remaining features in the first dataset, wherein the model being trained using the training dataset improves the outcome prediction accuracy of the model.

The invention relates to a law decision result prediction method and system, and the method comprises the steps: constructing an intellectual property case keyword dictionary, and obtaining a law text of an intellectual property related case from a first public database through the intellectual property case keyword dictionary; obtaining enterprise economic data corresponding to each legal text from a second public database; constructing a structured topic extraction model, and training the structured topic extraction model by using the legal text and the enterprise economic data; constructing a legal judgment result prediction model, and training the legal judgment result prediction model by using a topic distribution result output by the trained structured topic extraction model; and obtaining a legal text and enterprise economic data of a to-be-predicted object, inputting the legal text and the enterprise economic data into the trained structured topic extraction model to obtain a topic distribution result, and inputting the topic distribution result into the trained legal judgment result prediction model to perform legal judgment result prediction. According to the method, the prediction accuracy and the interpretability of the legal judgment result of the legal text of the intellectual property case can be remarkably improved.

The invention relates to the technical field of machine learning and disease prediction, in particular to a KAN-based necrotizing enterocolitis outcome prediction model construction method. According to the method, firstly, the key variables related to the necrotizing enterocolitis outcome are screened, the necrotizing enterocolitis outcome can be predicted through the key variables, and the modeling complexity is reduced; secondly, a KAN is used for constructing an end prediction model, complex relations in data can be captured, and good performance and beforehand interpretability are achieved;

The invention relates to the technical field of medical treatment, and discloses a transfer learning method for bad outcome prediction of small-sample early-stage cancer patients. According to the method, the neural network risk function is introduced and the transfer learning framework is combined, so that the nonlinear interaction of high-dimensional features can be better fitted, and the prediction accuracy of the small-probability bad outcome of the early-stage tumor is improved. Although the existing RSF can overcome linear hypothesis to a certain extent, the existing RSF is susceptible to noise interference in a small sample queue, and the prediction stability is insufficient. According to the migration survival forest method provided by the invention, by training and migrating structure priori on a large-scale database, a stable tree structure can be kept under a small sample condition, and a split threshold is re-estimated in combination with target data, so that global structure knowledge and local adaptability are taken into consideration, and the reliability of the system is improved. Therefore, the robustness and generalization ability of the model are remarkably improved.

The invention discloses a blood fat-clinical outcome correlation analysis system and decision-making method oriented to an assisted reproduction and pregnancy period, and belongs to the field of medical information technology and intelligent diagnosis and treatment, and the method comprises the following steps: obtaining blood fat data and clinical baseline characteristics of a patient on an ovulation promotion starting day and a trigger day; adjusting a preset standard metabolic response curve by taking the clinical baseline characteristics as adjusting parameters based on blood fat data, and performing constraint fitting on the adjusted curve through parameter inversion to generate a personalized metabolic response curve and extract dynamic response characteristics; inputting the dynamic response features and the clinical baseline features into a risk-outcome prediction model to obtain a prediction result; and generating a clinical decision suggestion according to the prediction result and the decision rule. According to the method, discrete blood fat data are constructed into a continuous personalized metabolic response curve by adopting a method based on parameter inversion and constraint fitting, and dynamic response characteristics are extracted from the continuous personalized metabolic response curve for outcome prediction, so that the clinical outcome can be predicted more accurately, and a reliable basis is provided for personalized clinical decision making.

The present disclosure proposes an artificial intelligence (AI) based System (100) for preoperative assessment and surgical outcome prediction. The AI-based system (100) comprises one or more client modules collect data from electronic medical records (EMRs), Data repositories and Databases via a data collection module (122), then standardize and pre-process it with a standardization & pre-processing module (124) for integrity and compatibility. The pre-processed data is transmitted to a server (104) through a network(102) using an API module (126) and thereafter forwarded to an input module (112). The input module (112) authenticates and validates data of at least one patient. The data is subsequently sent to a processing module (114), and a prediction module (116) for predicting surgical duration and blood loss, and the extreme Gradient Boost model for postoperative patient placement predictions.

An intelligent system for optimizing radiotherapy for personalized radiotherapy, consisting of: • a genome data acquisition module configured to acquire and process a patient's genomic, molecular, and biomarker information; • a multimodal medical imaging module configured to acquire medical image data from one or more imaging modalities; • an image processing and automatic segmentation module configured to preprocess the acquired medical images, register multimodal images, segment tumors and organs at risk, and extract quantitative image features; • an artificial intelligence and data integration module configured to integrate genome data, imaging data, radiomic features, and clinical information to generate a patient-specific predictive model;• A treatment planning and dose optimization module configured to automatically generate and optimize a personalized treatment plan based on the integrated patient-specific model; • A digital twin and adaptive therapy module configured to simulate the patient-specific treatment response and continuously adjust the treatment plan during therapy; • A treatment monitoring and outcome prediction module configured to predict treatment response, disease progression, and radiation-induced toxicity based on longitudinal patient data; • A clinical decision support module configured to generate personalized treatment recommendations and support clinical decision-making;• A communication and data management module configured for the secure exchange, synchronization, and storage of clinical, imaging, genomic, and treatment-related information; and • An autonomous learning and systems management module configured to continuously improve predictive models and treatment optimization algorithms based on collected treatment outcomes, with the modules working together to generate, optimize, monitor, and continuously adapt personalized radiotherapy based on integrated genomic and multimodal medical imaging information.

The application provides a gastric low-grade intraepithelial neoplasia outcome evaluation method based on deep learning imageomics features, and relates to the field of medical information technology. The method comprises the following steps: acquiring a gastric medical image, extracting a lesion region of interest and expanding outward to construct an extended background region containing a surrounding mucosa microenvironment; extracting a high-dimensional imageomics feature set from the lesion region of interest; inputting the above two regions into a multi-branch convolutional neural network to extract deep semantic feature maps, and calculating the spatial feature difference to obtain an edge weight prior matrix; then constructing a spatial heterogeneity graph, and obtaining deep imageomics fusion features through collaborative updating of a graph attention network; finally, inputting the deep imageomics fusion features into an outcome prediction classifier to obtain a final outcome evaluation result. The application realizes deep fusion modeling of macro-microenvironment differences and micro-tissue heterogeneity, and significantly improves the accuracy of outcome prediction.

A Computer Aided Diagnosis, CADx, system and method for predicting an outcome score for a patient is described. The system comprising: an input circuit configured to receive input data comprising at least one input medical image for a patient; an outcome prediction circuit operably coupled to the input circuit configured to receive input data from the input circuit for outcome prediction analysis; wherein the outcome prediction circuit is further configured to receive details of a suggested future intervention for the patient; and the input data to the outcome prediction circuit is analysed to generate the outcome score for the patient accounting for the details of the future intervention.

The invention belongs to the field of legal artificial intelligence application, and particularly relates to a dispute result prediction system based on legal specification and fact matching, which is suitable for intelligent pre-judgment of judgment results and mediation schemes of litigation cases. A system architecture and a working process comprise small-premise structured input, large-premise intelligent matching, conclusion logic deduction, class case comparative analysis and document intelligent generation, and the system strictly follows a legal applicable logic structure. By introducing a similarity algorithm (such as a BERT legal semantic model), the fact-law-conclusion accurate mapping is realized, and it is ensured that different fact inputs inevitably correspond to differentiated prediction conclusions.

The invention relates to the technical field of deep learning, and discloses an acute ischemic stroke function outcome prediction method and device based on deep learning. The method comprises the following steps: segmenting a preprocessed stroke lesion image sequence and a preprocessed deep marrow vein image sequence by using a deep learning image segmentation model; generating a first image sequence according to the preprocessed stroke lesion image sequence and the lesion segmentation mask sequence; generating a second image sequence according to the preprocessed deep marrow vein image sequence and the deep marrow vein segmentation mask sequence; processing the first image sequence and the second image sequence by using a deep learning classification model to obtain a first functional result prediction result and a second functional result prediction result; and generating a stroke functional result prediction result based on the first functional result prediction result and the second functional result prediction result. According to the invention, the stroke function outcome can be predicted more quickly and accurately.

This invention proposes a deep learning-based method for predicting student learning outcomes. First, it collects student learning behavior data, evaluation data, interaction data, and attribute data through a compliant authorization mechanism, generating standardized data after anonymization and preprocessing. Second, it extracts behavioral feature vectors, evaluation feature vectors, interaction feature vectors, and attribute feature vectors using bidirectional LSTM, fully connected networks, a pre-trained language model (BERT), and an embedding network, respectively. Finally, it generates prediction results through an outcome prediction model, which includes a feature fusion layer, a main feature extraction layer, a memory enhancement layer, an attention decision layer, and an output layer. This invention achieves accurate prediction of student learning outcomes through multi-source data fusion and external memory enhancement mechanisms, while ensuring compliance and privacy security in the data processing process.

This invention provides a method, apparatus, and device for establishing a customer service outcome prediction model, relating to the field of artificial intelligence technology. The method includes: obtaining a training text dataset based on historical customer service voice data; performing quantum measurement processing on the target statement text training data in the training text dataset to obtain the semantic features of the target statement; using a bidirectional long short-term memory (BiLSTM) model and a self-attention mechanism to obtain the semantic features of the target dialogue based on the semantic features of the target statement; and obtaining a customer service outcome prediction model based on the semantic features of the target dialogue and the corresponding historical customer service outcomes. This invention's solution, by obtaining a customer service outcome prediction model based on the semantic features between the contexts of the target dialogue, fully models the semantic features of the target statement and the semantic features between the contexts of the target dialogue, increasing the accuracy of the customer service outcome prediction model in predicting customer service outcomes.

The invention discloses a judgment result prediction method, device and equipment based on legal relevance and a medium, and the method comprises the steps: inputting a case statement into a plurality of prediction models, and obtaining a plurality of law article prediction results, a plurality of command and control prediction results and a plurality of penalty prediction results; respectively combining the plurality of law article prediction results, the plurality of command and control prediction results and the plurality of penalty prediction results to obtain a plurality of law decision prediction results; determining the relevance degree of each legal decision prediction result; according to the case statement, determining similar legal cases, and obtaining a plurality of commands of the similar cases; and determining a final judgment prediction result corresponding to the case statement according to the plurality of commanding and control of the similar cases, the plurality of legal judgment prediction results, a preset commanding and control criminal name penalty principle and the relevance degree of the plurality of legal judgment prediction results. The invention belongs to the field of legal penalty prediction. The prediction accuracy can be improved.

The application discloses an assisted reproductive clinical outcome inference method based on a deep neural network, relates to the technical field of medical intelligent data analysis, and comprises the following steps: collecting a multi-modal reproductive health data set of a target patient group, which comprises time-series hormone detection records, follicle ultrasound measurement sequences and embryo morphology score vectors; performing feature alignment and missing value filling processing on the data set to construct a standardized feature tensor; sending the standardized feature tensor into a deep residual contraction network to complete multi-level abstract feature extraction, outputting an embryo implantation potential encoding matrix, and then performing time-series attention weighted fusion on the encoding matrix to generate a clinical pregnancy outcome prediction probability distribution, according to which, a target patient assisted reproductive outcome inference label is determined. The method realizes unified regulation and deep feature mining of multi-class reproductive health heterogeneous data, combines time-series feature fusion logic to complete outcome quantization determination, and is suitable for an assisted reproductive clinical intelligent analysis application scene.

The invention discloses a multi-target reinforcement learning recommendation method and system oriented to long-term user participation, and relates to the technical field of data processing, and the method comprises the steps: extracting a user interaction record from an offline log, constructing a state representation characteristic vector set, and generating a long-term participation label set, so as to train a long-term result prediction model, generating a long-term reward component; and constructing a multi-target Q network and a dynamic weight network, and carrying out alignment constraint joint training on the dynamic weight multi-target network by introducing a long-term reward component to obtain a multi-target reinforcement learning recommendation model. According to the method, the technical problems of low model learning efficiency and poor adaptability caused by long-term feedback delay and stiffness of a multi-target weighing mechanism when the long-term user participation degree is optimized in the prior art are solved, multi-target dynamic weighing is realized by introducing progressive feedback and a dynamic weight mechanism, and the user participation degree is optimized. And the learning efficiency and adaptability of the model to the long-term user participation degree are effectively improved.

The invention provides an intelligent feedback error correction method for law judgment result prediction and electronic equipment, and the method comprises a bidder agent, a controller agent, a discriminator agent and a crime-measuring judge agent which are obtained through large language model training, and the method comprises the steps: carrying out the event-level disassembly and extraction of a case fact text through the bidder agent; obtaining a fact atom set; generating a crime name set based on the fact atom set through the control agent, and generating a reverse set based on the fact atom set through the dialect agent; carrying out adversarial collaborative reasoning on the crime name set and the reverse set through a crime quantity judge agent to obtain an initial crime determination result; the initial crime determination result is verified, and if a doubtful crime name exists, correction content is generated for the doubtful crime name through the control party agent and the party distinguishing agent; and generating a final crime determination result according to the correction content through the crime determination judge agent. And a judicial institution can be assisted to realize more stable, explainable and efficient law decision prediction.

The invention discloses a cerebral apoplexy early-stage clinical outcome prediction method, system, device, medium and product, and relates to the field of clinical outcome prediction.The method comprises the steps that an electronic medical record of an old ischemic cerebral apoplexy patient is collected, and medical features and medical feature values are extracted; according to the discharge diagnosis information in the electronic medical record, the medical history and disease diagnosis information of the senile cerebral arterial thrombosis patient are extracted, the number of common diseases is calculated, and different common disease modes are constructed; evaluating according to the physical examination condition of the old ischemic stroke patient during discharge, determining an mRS score, and determining a post-stroke clinical outcome dichotomy result according to the mRS score; constructing a machine learning model according to the medical characteristic value, the common disease mode and the post-stroke clinical outcome dichotomy result; the early clinical outcome result of the cerebral apoplexy of the patient is predicted according to the machine learning model, a personalized early treatment strategy is formulated based on each medical feature, and the clinical outcome result can be accurately predicted in an early stage.

The purpose of the present disclosure is to provide a program, a device, and a system for assisting in formulation of a treatment plan for a patient. More specifically, the present disclosure provides a program for assisting in formulation of a treatment plan for a patient, the program causing one or more computers to execute: a step for acquiring patient information; a step for generating a treatment draft plan on the basis of the patient information; and a step for outputting the treatment draft plan. In addition, a device and a system having the same functions are also provided. Furthermore, a program, a device, and a system that assist in outcome prediction for assisting in formulation of a treatment plan for a patient are also provided.