24 results about "Retrospective analysis" patented technology

The invention provides a personalized anesthesia management method and system. The method comprises the following steps: collecting real-time physiological data flow and previous medical record information of a patient to preliminarily evaluate operation requirements, and generating a preliminary anesthesia scheme; performing simulation prediction processing on the individualized reaction mode through a state-action value function updating mechanism by applying a deep Q network algorithm, and optimizing anesthetic selection and dosage by adopting a dynamic adaptive dosage adjustment technology to generate an optimized anesthetic scheme; the method comprises the following steps: extracting continuous physiological state parameters of a patient by applying an adaptive feedback control algorithm, presetting a safety threshold value, performing real-time comparison and analysis through a real-time parameter adjustment mechanism, performing automatic early warning processing on a trend deviating from a normal range by adopting an abnormal trend early warning technology, and generating a real-time adjustment instruction; and based on the real-time adjustment instruction, obtaining a patient recovery trajectory and subjective comfort evaluation, performing retrospective analysis, and generating a personalized anesthesia management scheme. According to the technical scheme provided by the invention, the safety of personalized anesthesia is remarkably improved.

The invention discloses a multi-agent cooperation molecular optimization method based on large language model driving. The method comprises the following steps: S1, task initialization and input analysis; s2, constructing an intelligent agent cluster; s3, task decomposition and scheduling execution; s4, performing expert optimization and tool calling; s5, performing evaluation and feedback screening; and S6, multi-round optimization and result output: the scheduling agent adjusts an optimization strategy and redistributes tasks according to a feedback result of the step S5, drives the expert agent to execute a next round of optimization operation, circulates the steps until the evaluation agent judges that an optimization target is met, and outputs a final optimization molecule and related attribute information thereof. And outputting an optimized path and an intermediate result for tracing analysis. According to the method, multi-round optimization and evaluation feedback iteration of a molecular structure are realized by fusing the knowledge reasoning ability of a large language model and an efficient interaction mechanism between intelligent agents.

The invention discloses a lightweight customer service operation backtracking analysis method and system based on an event stream snapshot, and relates to the technical field of backtracking recording, and the method comprises the steps: recording a DOM state of an initial page, traversing a DOM Tree of the whole page, converting the DOM Tree into JSON structure data through nodeType mapping, and generating a first screen DOM snapshot; updating incremental data after first screen snapshot, hijacking incremental change data by monitoring DOM incremental change, and storing the data into JSON structure data; and on the basis of the timestamp, assembling and restoring an event queue of the snapshots after analyzing by querying a screen recording index JSON file, reconstructing the snapshots into a corresponding DOM tree, and reproducing customer service operations covered in the timestamp. According to the invention, the technical design of event stream traceable recording is realized, and a new thought is provided for the application of the traceable recording technology.

Automated systems and methods are presented for retrospectively analyzing clinical trial data. A plurality of image derived from biological samples of patients in a cohort population are accessed. Image features are computed based on the plurality of images. A diagnostic feature metric is derived based on the computed image features. A cut point value is determined by applying a statistical minimization method using the derived diagnostic feature metric and patient outcome data from the cohort population, in which the cut point value identifies a patient in the cohort population as positive or negative for a diagnostic test.

Facilitating analytics such as retrospective analysis includes ingesting session data. It further includes summarizing the session data and generating leaf-level summarizations of leaf groups present in the session data. It further includes storing the leaf-level summarizations. It further includes deriving parent-level summarizations for a parent group from leaf-level summarizations of leaf groups that are children of the parent group.

The present invention provides an intelligent analysis system for porcelain insulators that combines mechanical and electrical properties, and relates to the field of monitoring data processing technology. The present invention can accurately track the path, speed, and direction of pollution flashover propagation from the starting porcelain disk to other porcelain disks by fusing multimodal images and time series analysis, thereby realizing dynamic modeling of the pollution flashover evolution process, effectively breaking through the limitations of the traditional static assessment model, and significantly improving the real-time and accuracy of risk assessment. The present invention marks the earliest pollution abnormal porcelain disk as the starting point of pollution flashover propagation through retrospective analysis of the integrated fusion abnormal points, and constructs a propagation chain and time series, which helps to identify key node porcelain disks and establish fault prediction logic based on evolutionary relationships. Through the dual-channel triggering mechanism, the early warning capability is significantly improved, ensuring that intervention can be made in the early stages of the risk, effectively preventing flashover from spreading to the entire string of insulators and causing system accidents.

The invention discloses an intelligent detection method for business logic vulnerabilities, and belongs to the technical field of vulnerability detection. The method is used for solving the technical problems that in an existing scheme, the multi-dimensional influence traceability analysis effect of existing vulnerabilities is poor, and the self-evaluation optimization effect of a detection effect is poor. The method comprises the steps of performing supervision, digital processing and dynamic marking on newly added business records based on business logs to obtain normal business records or abnormal business records, and performing tracing analysis on abnormal logic rules according to the abnormal business records obtained through analysis. Determining the logic vulnerability influence corresponding to the abnormal logic rule to which the abnormal business record belongs and performing digital representation, and performing business logic vulnerability coverage analysis in different aspects on logic vulnerability influence data obtained by processing all newly added business records correspondingly; and according to an analysis result, carrying out dynamic optimization management of the service logic rule on subsequent implementation of the existing service logic vulnerability detection scheme.

The present invention provides a method, system, medium and electronic device for apex calibration based on magnetocardiogram, the method comprising the following steps: acquiring target data, the target data comprising magnetocardiogram data and angiography image data; performing gradient calculation based on the magnetocardiogram data to obtain a first apex point; reconstructing a heart model based on the angiography image, thereby obtaining a second apex point based on the heart model; performing coordinate conversion based on the first apex point and the second apex point to complete the calibration of the apex. The method, system, medium and electronic device for apex calibration based on magnetocardiogram of the present invention can solve the problems in the prior art of CT and MCG data calibration deviation caused by manual positioning, inability to use past CT and MCG data for retrospective analysis, and difficulty in batch processing of data calibration; at the same time, large quantities of data can also be automatically processed in batches, which is convenient and efficient, and can assist in the diagnosis of heart diseases and preoperative guidance.

Devices and processing systems are configured to enable physiological event prediction based on blepharometric data analysis. For example, some embodiments provide methods and associated technology that enable retrospective analysis of blepharometric data driving subsequent hardware / software configuration, thereby to provide for personalized and / or generalized biomarker identification.

This invention discloses a trial auxiliary data management system and method based on judicial case quality evaluation, belonging to the field of data processing technology. By decomposing and reorganizing core and auxiliary cases, a cache of related cases is constructed. This allows for further mining of related cases from both core and auxiliary case perspectives, outputting the most similar related cases to the submitted cases to be evaluated. Simultaneously, using the optimal solution as a reference, cases requiring random checks can be quickly identified, significantly reducing errors and inaccuracies caused by manual operation, improving trial quality and efficiency. Furthermore, the optimal solution found can provide positive or negative preventative references for trial retrospective analysis.

The invention discloses a digital twinning modeling and evaluation method for typhoon resistance toughness of a culture platform, and relates to the technical field of ocean engineering digital twinning and risk evaluation.According to the scheme, a high-fidelity digital twinning body integrating multiple physical field models is constructed, and real-time data bidirectional interaction with a physical platform is established; on the basis, anti-typhoon toughness quantitative indexes including robustness, redundancy and restorability are systematically defined; by using the digital twinborn body, dynamic deduction, real-time evaluation and backtracking analysis of the whole process of the typhoon can be carried out, and the toughness state of the platform can be accurately observed; and finally, based on an evaluation result, simulating, optimizing and recommending an optimal toughness enhancement strategy in a digital space. According to the method, the anti-typhoon capability of the culture platform is predicted from static evaluation to dynamic prediction and from a single safety coefficient to comprehensive toughness management, and a core technical support is provided for safe operation of deep and far sea culture.

Systems, methods and computer-readable media for providing explainable selective attention in multi-source or multi-speaker environments. A selective attention module receives multimodal sensor data including audio, video, gaze, text, and physiological signals from a plurality of sources. An attention inference engine generates attention distributions over the sources and fuses them into a probabilistic belief state. An explainability module produces interpretable outputs corresponding to the fused belief, including attention matrices, confidence scores, reliability measures, margin-based differentiators, and natural language rationales. The explainability outputs are rendered through visual, auditory, or augmented / virtual reality interfaces to indicate the attended source, suppressed sources, and reasoning for the selection. The system enables user interaction by providing justifications in real time, logging explanations for retrospective analysis, and supporting adaptation of thresholds and model weights based on feedback. The disclosed technology improves transparency, interpretability, and trust in selective attention systems, while maintaining real-time performance in dynamic multi-speaker environments.

The present invention provides a retrospective analysis of subject and / or implant positioning. Provided is a method and device for assisting positioning of a subject having an implant in X-ray imaging. The method comprises the steps of receiving (S1) X-ray image data of the subject including the implant, receiving (S2) exam request data indicating at least an exam type to be performed for the specific implant, and determining at least one set of image features being assigned to the requested exam type and indicative for positioning of the subject and / or the implant relative to an X-ray imaging device. Further, the method comprises analyzing (S3) an image content of the received X-ray image data for the determined set of image features, and determining (S4), from the analysis of the image content, a feedback being indicative for positioning the subject and the X-ray imaging device to each other with respect to a target positioning, to be provided for assisting the positioning. Advantageously, the set of image features defines at least a contour of the implant, wherein the contour is determined to meet a corresponding target contour defined by the set of image features in case of meeting the target positioning.

The invention provides a power transmission line fault backtracking analysis method and system and a storage medium, and the method comprises the following steps: carrying out the hierarchical processing of lightning events, traveling wave waveforms, meteorological parameters and video image data of a power transmission line, and obtaining corresponding fault characterization features; and carrying out logic combination on the multi-source fault characterization characteristics according to preset judgment conditions of all fault types, obtaining the fault types and determining fault positions and time. The precision, efficiency and reliability of power transmission line fault diagnosis are effectively improved, and the fault type, the fault occurrence position and the fault reason are accurately determined.

The invention relates to the technical field of clinical diagnosis, and particularly discloses a construction method and application of a primary liver cancer diagnosis model GAAAD. According to the method, clinical information of patients with primary hepatocellular carcinoma, patients with other liver diseases and healthy people is subjected to retrospective analysis, a multi-factor Logistic regression method is utilized, and a GAAAD combined diagnosis model composed of five variables including gender, age, AFP, AKR1B10 and PVIKA-II is constructed. Clinical data verifies that the model is excellent in prediction distinction degree and accuracy, can accurately evaluate the risk of an individual suffering from primary hepatocellular carcinoma, has important clinical value in early prevention, diagnosis and treatment and prognosis evaluation of diseases, and can assist in formulating a scientific diagnosis and treatment scheme.

The present invention relates to the technical field of multi-modal perception monitoring, specifically including a method and system for monitoring the operation status of park facilities based on multi-modal perception, which includes: collecting multi-modal perception data, generating a fusion matrix, extracting operation features, constructing a facility status evolution map, configuring a monitoring period, identifying potential failure modes, matching with historical data to optimize monitoring rules, and outputting a monitoring report. It solves the technical problem that the potential failure features of cross-modal associations cannot be effectively extracted, affecting the accuracy and comprehensiveness of failure identification, and realizes the construction of a facility status evolution map. By combining the time prediction of the health index and the spatial annotation of the failure risk probability, latent failures are identified in advance. At the same time, through multi-modal data retrospective analysis and state transition feedback optimization, the sensitivity of failure feature extraction is improved, and incremental learning and model iteration are used to continuously optimize the adaptive association model, achieving the technical effect of improving the accuracy and comprehensiveness of failure identification.

The invention provides a puerpera and fetus two-dimensional synergistic painless delivery anesthesia dynamic regulation and control monitoring method, and belongs to the technical field of painless delivery anesthesia. The method comprises the following steps: constructing a puerpera-fetus two-dimensional monitoring system, and synchronously collecting multidimensional signals such as puerpera side pain perception, circulatory respiration and drug metabolism and core signals such as fetus side heart rate and blood oxygen saturation; performing adaptive filtering preprocessing and attention mechanism driven multi-source data fusion analysis to obtain a puerpera anesthesia state, a fetus safety state and a collaborative correlation coefficient of the two states; constructing a regulation and control model by taking a parturient painless threshold-a fetus safety threshold as double constraints, outputting an initial anesthesia scheme, and performing real-time dynamic adjustment in combination with a delivery process and drug interaction; meanwhile, full-process data recording and three-level early warning are achieved, and the scheme is optimized through backtracking analysis after delivery. The problems that in the prior art, puerpera pain relieving is emphasized, and fetus safety cooperative monitoring and regulation lack dynamics are neglected are solved, and anesthesia precision and mother and infant safety are improved.

Methods of reliably estimating genomic fraction (e.g., fetal fraction) from polymorphisms such as small base variations or insertions-deletions are disclosed. Sequenced data from a multigenomic source is used to determine allele counts for one or more of the polymorphisms. For one or more of the polymorphisms, zygosity is assigned, and genomic fraction is determined from the zygosity and allele counts. Certain embodiments employ SNPs as the relevant polymorphism. The disclosed methods can be applied as part of an intentional, pre-designed re-sequencing study targeted against known polymorphisms or can be used in a retrospective analysis of variations found by coincidence in overlapping sequences generated from maternal plasma (or any other setting where a mixture of DNA from several people are present).

A robotic surgical system includes a robotic instrument coupled to a controller and configured to interact with biological tissue. A physiologic sensing subsystem has one or more sensors configured to detect real-time physiological parameters. The sensors include at least one of: tissue oxygenation sensors, perfusion sensors, bioimpedance sensors, or neural conductivity sensors. A boundary mapping engine includes an artificial intelligence (AI) module trained to: generate dynamic boundary zones within a surgical field based on sensed physiological thresholds and contextual procedural data; classify zones as high-risk based on learned thresholds derived from patient-specific models, historical surgical outcomes, or intraoperative signal patterns; and generate an instrument constraint map corresponding to the boundary zones and continuously refine it based on ongoing sensor inputs and AI-inferred trends; A control module configured to: receive the instrument constraint map; restrict, dampen, or modify robotic instrument movement preventing intrusion into AI-classified high-risk boundary zones; and log override attempts, constraint violations, and AI confidence scores for real-time or retrospective analysis. A visualization interface presents real-time physiological boundary overlays, constraint zones, and AI-generated risk indicators to the surgeon.

This invention belongs to the field of medical data analysis and provides a retrospective analysis method and system for calculating physiological data thresholds based on mathematical inflection points. The method includes: acquiring the patient's historical physiological data and the patient's outcome data; using the patient's historical physiological data as the independent variable and the patient's outcome data as the dependent variable, constructing a regression analysis model using a machine learning classification algorithm; performing second-order differentiation on the regression model to obtain the point where the second derivative is zero, which is the threshold; and predicting the occurrence of postoperative abnormalities based on the threshold. This invention calculates the corresponding threshold using mathematical means, providing reliable mathematical support and theoretical basis for the origin of the threshold. Furthermore, it allows for automated calculation of the corresponding threshold by designing computer programs, ensuring the uniqueness of the results for each calculation and eliminating interference from human factors.

The invention discloses a burn assessment method and system based on image recognition, relates to the technical field of burn assessment, and realizes automatic calculation and grading judgment of pain grades by constructing a fusion index and a joint pain response index psi and mapping the fusion index and the joint pain response index psi into a standardized scoring index Ps. The scoring mechanism not only has standard uniformity and cross-patient comparison ability, but also has self-learning and error suppression ability through a subsequent correction mechanism, so that the system is more stable and credible. According to the method, a dynamic backtracking analysis mechanism of the psychological state is specially set, abnormal scores caused by psychological fluctuations such as anxiety and nervous tension can be recognized, and the score result is corrected through an emotion regulation function. The mechanism solves the problem that an existing AI evaluation system cannot distinguish'real pain 'and'emotional response', effectively prevents excessive response or misjudgment intervention, and improves the clinical applicability of an evaluation result.

The invention discloses a biomarker combination for grading diagnosis of esophageal squamous intraepithelial neoplasia and application of the biomarker combination, and relates to the technical field of biological medicines. The biomarker combination comprises p53, Ki-67 and Statmin-1, and is characterized in that the biomarker combination comprises p53, Ki-67 and Tissue specimens of esophageal squamous intraepithelial neoplasia patients and esophageal squamous cell carcinoma patients are subjected to retrospective analysis, the expression conditions of p53, Ki-67 and Statmin-1 are detected by using an immunohistochemical technology, and the application value of Statmin-1, p53 and Ki-67 protein joint detection in grading diagnosis of esophageal squamous intraepithelial neoplasia is discussed. The result shows that the Statmin-1, p53 and Ki-67 protein combined detection has certain diagnosis and prediction values on esophageal squamous intraepithelial neoplasia grading, and the diagnosis accuracy of esophageal squamous intraepithelial neoplasia can be improved.

The present invention relates to a method for trust management of group intelligence perception based on active and retrospective analysis, which belongs to the field of big data. The method comprises the following steps: obtaining task submission data of all participants and performing standardization processing on each dimension of data; sending a drone to an area with uncertain participant reliability to collect data and record it as real data; calculating the sum of the difference values ​​of each dimension of data to judge the participant's credibility; calculating the ratio of each participant's difference value to the real value and defining a cutoff point to judge the trust value trend of the participant's submitted data; setting a trust value update formula; comparing the traced historical data with the real data, performing retrospective comparison on the participant's previous behavior, analyzing and judging the participant's credibility; calculating the increase or decrease in the credibility value within each timestamp; calculating the increase or decrease in the total credibility value of the participant, and judging the participant's trust trend. The method of the present invention can be used to identify malicious participants in the system and improve the quality of group intelligence perception data.

The invention provides a porcelain insulator intelligent analysis system combining mechanical and electrical performance, relates to the technical field of monitoring data processing, and can accurately track the transmission path, speed and direction of pollution flashover from a starting porcelain plate to other porcelain plates by fusing multi-modal images and time sequence analysis, thereby improving the analysis accuracy of the pollution flashover. Dynamic modeling of the pollution flashover evolution process is achieved, the limitation of a traditional static evaluation mode is effectively broken through, and the real-time performance and accuracy of risk evaluation are remarkably improved. According to the method, through backtracking analysis of comprehensive fusion abnormal points, the earliest pollution abnormal porcelain plate is marked as a starting point of pollution flashover propagation, and a propagation chain and a time sequence are constructed, so that key node porcelain plates can be identified, and fault prediction logic based on an evolution relation is established. And through a double-channel triggering mechanism, the early warning capability is remarkably improved, intervention can be performed in the early risk stage, and flashover is effectively prevented from spreading to the whole string of insulators to cause system accidents.