187 results about "Clinical information" patented technology

The invention is applicable to the technical field of clinical medicine, provides a clinical multi-modal cancer drug response prediction method based on feature reconstruction, constructs a clinical multi-modal model for drug response prediction of diffuse large B-cell lymphoma, and aims to predict the drug response of diffuse large B-cell lymphoma by integrating gene sequencing and clinical multi-modal data. And accurate drug reaction prediction is realized. The model adopts an end-to-end multi-stage processing flow: firstly, extracting gene features through TransP-Net, and processing multi-modal clinical data by using a clinical information encoder; then, pseudo-gene features are generated through a clinical-genome filling module to deal with the data missing problem; and finally, multi-modal deep fusion is realized through a clinical information decoder, and a prediction result is output. According to the method, data characteristics and working processes in a real clinical environment are fully considered, two conditions of complete gene data and missing gene data can be processed at the same time, and the method has a good clinical transformation prospect and application value.

The invention discloses a device for awareness level evaluation and a storage medium. The apparatus comprises: a processor; the device realizes the following operations: collecting clinical information of a person to be assessed and a task state electroencephalogram signal under a target stimulation normal form; extracting frequency domain characteristics of a specific frequency band and spatial-temporal characteristics of a target event related potential based on the task state electroencephalogram signal, and combining the frequency domain characteristics and the spatial-temporal characteristics into a corresponding electroencephalogram topographic map; inputting the corresponding electroencephalogram topographic map into a multi-modal large language model, and performing image feature extraction by using an image encoder to obtain electroencephalogram features; inputting clinical information into the multi-modal large language model, and performing text feature extraction by using a text encoder to obtain text features; and performing cross-modal attention calculation fusion on the electroencephalogram features and the text features by using a cross-modal fusion module to realize consciousness evaluation so as to output a consciousness evaluation result. By means of the scheme, the consciousness level of the patient can be automatically and accurately evaluated.

The invention discloses a clinical intelligent decision-making method based on proxy workflow and a storage medium. Comprising the following steps: acquiring and preprocessing clinical information of a patient, and constructing a candidate disease set; and constructing a proxy directed workflow. In the retrieval stage, the diagnosis criteria corresponding to the candidate diseases are retrieved and aggregated from the diagnosis criteria library rechecked by the experts to form working memory. The preliminary diagnosis stage model node generates a preliminary candidate diagnosis set in combination with work memory and patient medical history and physical examination. And the final diagnosis stage generates a final diagnosis result based on the preliminary candidate diagnosis and the complete clinical information. And when the global confidence is lower than a threshold value, the model node pointedly checks an information source and updates reasoning and confidence. A successful reasoning track forms a demonstration set after manual auditing, the demonstration set is used for supervising a fine tuning model to obtain an initial strategy, multiple structured outputs are generated through grouping sampling, relative strategy updating is carried out in combination with reward signals and reference strategy regularization constraints, and optimization and stable improvement of the model diagnosis capability are achieved.

The invention specifically relates to a clinical deterioration risk prediction and early warning system and method based on machine learning, and relates to the technical field of medical artificial intelligence and clinical informatics, and the method comprises the steps: obtaining multi-dimensional time series data in real time; constructing a dynamic feature engineering vector; machine learning risk prediction; judging a risk threshold value and triggering early warning; and interpretation and suggestion generation driven by the large language model. According to the method, multi-dimensional time sequence data is continuously acquired in real time, multi-sliding window statistical features and standardized clinical deterioration scores are extracted in combination with dynamic feature engineering, and accurate quantitative risk prediction of multiple disease deterioration types such as sepsis and respiratory failure is realized by means of machine learning models which are specifically trained by XGBoost, LSTM and the like. The problem that traditional early warning depends on manual judgment and is high in hysteresis is effectively solved; medical staff can be helped to quickly grasp the core inducement of disease deterioration, and a standardized and landing action scheme is provided.

The invention belongs to the technical field of medical image intelligent diagnosis, and provides a breast cancer focus benign and malignant discrimination method based on a gated multi-expert mechanism. The method comprises the following steps: firstly, carrying out standardization and semantic preprocessing on a mammary gland X-ray image, a BI-RADS imaging report and structured clinical data, embedding age, mammary gland density and focus position information into a text template in a natural language form, and realizing unified expression of multi-modal input; secondly, extracting image features by utilizing a ResNet network and a simplified CLIP model, obtaining a text semantic vector by adopting a Bio-ClinicalBERT model, and establishing two sub-paths of a lump expert and a calcification expert in a Transform structure; further, an expert weight is dynamically generated through a gating routing mechanism, and soft routing fusion is executed; and finally, outputting benign and malignant results of the breast cancer focus by the binary classification module. According to the method, deep fusion and dynamic collaboration of the mammary gland X-ray image, the BI-RADS text and the clinical information are realized, and the accuracy and interpretability of breast cancer discrimination can be remarkably improved.

The invention relates to the technical field of data synchronization, in particular to a digestive system department clinical information acquisition and synchronization system, which comprises a multi-source heterogeneous acquisition module for generating a standardized diagnosis and treatment sequence, a characteristic spectrum construction module for mapping a text and an image into vector nodes and constructing a dynamic diagnosis and treatment characteristic spectrum, the state fingerprint verification module calculates a map fingerprint Hamming distance to position a difference feature node, and the incremental collaborative synchronization module constructs an incremental data packet and sends the incremental data packet to the central server. According to the method, the multi-dimensional characteristic spectrum based on the diagnosis and treatment time sequence is constructed, discrete images and texts are converted into topological association units, heterogeneous data semantic level alignment and integrity verification are achieved, version conflicts and information faults are eliminated, meanwhile, a dynamic hash fingerprint difference comparison strategy is adopted, accurate recognition is achieved, and only substantial change nodes are transmitted; and the network load is reduced, and high real-time consistency and zero-loss circulation of whole-flow information are ensured.

The present invention relates to an AI-based system and method for generating enhanced radiology reports. The system comprises a database for storing multimodal patient data, a natural language processing (NLP) module for extracting clinical information, and a machine learning module for correlating the clinical information with radiology images to identify diagnostic insights. An AI-based report generation module analyzes the images and clinical information to generate a preliminary report, which is refined based on radiologist input. The generated report is then integrated into the patient's electronic health record. The system employs techniques such as multimodal deep learning, active learning, explainable AI, and federated learning to enhance diagnostic accuracy, capture expert feedback, provide transparency, and enable multi-institutional collaboration. The invention aims to improve the accuracy, efficiency, and value of radiology reporting in patient care.

The invention discloses a medical document processing method and system based on a double-pipeline architecture, and relates to the technical field of document processing. According to the medical document processing method based on the double-assembly-line architecture, through a closed-loop process of document classification, preprocessing, double-assembly-line directional parallel processing and hierarchical vectorization storage, precise adaptation and efficient processing of multi-format and multi-type medical documents are achieved, the information loss rate and the key information truncation rate are greatly reduced, and the medical document processing efficiency is improved. According to the method, the document processing efficiency and the data standardization degree are improved, the warehousing success rate and the data traceability of the vector library are ensured, high-quality and structured data source support is provided for subsequent medical intelligent retrieval, clinical question and answer and retrieval enhancement generation system application, the knowledge base construction and maintenance cost is remarkably reduced, and the method is suitable for popularization and application. The problems that in existing medical document processing, medical semantics are not taken into consideration, so that key clinical information is easy to cut off, and a single processing flow cannot adapt to a structured guide and an unstructured case are solved.

The invention belongs to the technical field of tumor molecular diagnosis and biological information analysis, and relates to a high-throughput sequencing method and system for monitoring acute lymphocytic leukemia (MRD). Through targeted sequencing with a unique molecular identifier and / or a double-chain tag, error modeling based on a background noise spectrum and statistics / machine learning pseudo variation filtering, ultra-deep accurate detection of IG / TCR cloning and related gene low-frequency variation is realized. And an artificial intelligence recurrence risk prediction model is established by combining a time sequence MRD index, cloning diversity and clinical information, and a structured clinical report is output and docked with LIS / HIS. According to the method, the sensitivity and the specificity of ALL minimal residual disease detection can be remarkably improved, dynamic evaluation on leukemia cloning evolution and recurrence risks is realized, and a reliable basis is provided for individualized treatment decision and long-term follow-up visit.

The invention provides a doctor influence comprehensive evaluation method based on multi-modal data fusion, and the method comprises the steps: collecting media reports, clinical information, social platforms and other multi-source heterogeneous data, introducing timestamp marks, and achieving the time sequence semantic vector representation of a doctor entity through a medical field pre-training model and sine function embedding; in combination with BiLSTM-CRF and an attention mechanism, doctor attributes and key events are identified, and an LSTM and a dynamic clustering algorithm are adopted to extract and divide doctor attribute evolution trajectories in stages; an influence score is calculated through an exponential decay function, stage knowledge graph nodes are generated, and incremental updating and node merging and splitting are supported; a graph convolutional network and a graph attention mechanism are adopted, a doctor influence evolution chain is constructed, information dynamic association and trend prediction are achieved, and the time sequence precision and data comprehensiveness of doctor influence evaluation and the dynamic evolution ability of a knowledge graph are improved.

The invention discloses a method and a system for constructing an intelligent typing diagnosis model of psoriasis. The method comprises the following steps: S1, collecting pairing multi-modal data of a psoriasis patient whose fingernails are not tired, wherein the pairing multi-modal data comprises clinical information, scanning electron microscope images of fingernail surface morphology and infrared spectrum data of fingernail protein; s2, constructing a clinical feature encoder; s3, constructing a spectral feature encoder; s4, constructing an image feature encoder; s5, constructing a hierarchical fusion module; and S6, three loss function components are constructed, multi-objective optimization of the model is realized, and a total loss function is adopted to carry out model training. According to the method, the clinical information, the scanning electron microscope image of the nail surface morphology and the infrared spectrum data of the nail protein are integrated, and the multi-mode deep learning technology is combined, so that early recognition and prediction of psoriatic arthritis in a psoriasis patient are realized.

Approaches are described for facilitating remote ophthalmic examinations using three-dimensional (3D) imaging and mixed reality technology. A system obtains real-time or stored 3D data of a patient's eye, capturing detailed anatomical structures. The system analyzes the 3D data to identify specific regions of the eye, such as the cornea or retina, and retrieves corresponding diagnostic data and patient-specific clinical information. The 3D data, diagnostic metrics, and clinical records are integrated to generate an interactive visualization, which is presented through a mixed reality interface. The system allows healthcare professionals to manipulate diagnostic overlays, investigate flagged abnormalities, and adjust the visualization using gesture-based inputs. Machine learning models may be applied to detect potential abnormalities in the eye, while the system also determines stages of the examination based on changes in the anatomical structure of the eye.

The application discloses a kind of Zhang Xizheng pure Chinese medicine theory prescription medicine digital inheritance and intelligent analysis system.The system includes disease and syndrome differentiation rule engine subsystem, prescription three-way search subsystem, drug knowledge enhancement subsystem, medical record intelligent search subsystem, Zhang Xizheng school review subsystem and large language model interaction subsystem.The method performs the following steps: LLM clinical information extraction and disease prediction (extract symptoms, tongue, pulse, and 8 big disease classification prediction), disease and syndrome differentiation KB rule matching (8 big disease classification x multiple syndrome type weighted scoring matching), multi-source search (prescription three-way search+drug knowledge double-way vector recall and prescription-drug dynamic association+medical record multidimensional mixed scoring search), LLM fusion generation (inject all search results Zhang Xizheng school special prompt word generation theory prescription medicine analysis report) and Zhang Xizheng school four-dimensional review (western medicine perspective+air theory+drug use principles+theory prescription medicine consistency).The present application first realizes the computer formalization expression of Zhang Xizheng "disease differentiation before syndrome differentiation" mode and "air monism" theory, based on "Medical Western Medicine Record" 172 prescriptions+213 drugs / medical theory / medical conversation+137 cases to build a complete knowledge search and intelligent analysis system, with the ability of western medicine perspective+air theory+drug use principles+theory prescription medicine consistency.

The invention provides an innovative model based on the correlation between the PLEKHA4 gene expression level and the low-grade glioma radiotherapy sensitivity and a prediction method. According to the model, a multi-factor Logistic regression analysis framework is constructed, and the PLEKHA4 gene expression level and key clinical pathological parameters are organically combined, so that a radiotherapy sensitivity prediction model is established. The method comprises a series of steps of sample collection, gene expression detection, clinical information collection, model calculation, result interpretation and the like, and can realize accurate prediction of radiotherapy response of low-grade glioma patients. Compared with the prior art, the method has the remarkable advantages of simplicity and convenience in operation, high prediction accuracy, high clinical transformability and the like. The method has great potential in the aspect of guiding individualized radiotherapy scheme formulation, can significantly improve the treatment effect, and has important value for medical application. Besides, the model can be optimized through further clinical verification, so that clinical practice can be better served, and the life quality and prognosis effect of patients are improved.

The embodiment of the invention provides a medical image processing method, a fundus image processing method, a model generation method, equipment, a storage medium and a program product, which are applied to the field of image processing, and comprise the following steps: obtaining a first medical image and clinical information obtained by shooting a first organ object of a target user; inputting the first medical image and the clinical information into an image processing model, extracting image features in the first medical image by using a first image encoder in the image processing model, and extracting text features of the clinical information by using a text encoder; using a feature reconstruction module to obtain reconstruction features associated with the second organ object based on the image features, and using the text features to correct the reconstruction features; generating, using a third image decoder, a second medical image based on the corrected reconstruction feature; the second medical image comprises a second organ object and is used for analyzing the second organ object. According to the scheme of the embodiment of the invention, the obtaining cost of the second medical image is reduced.

The invention relates to the technical field of data analysis, and particularly discloses a data analysis method and system for acute kidney injury, and the method comprises the steps: integrating the multi-source acute kidney injury data of a patient, including clinical information and historical medical records, constructing the time series data of serum creatinine, urine volume and the like, extracting the risk features in an unstructured text by using an NLP technology, and analyzing the risk features in the unstructured text. Physiological index dynamic and semantic risk factors are deeply fused through space-time alignment, a basic feature vector is formed, a difference risk score of historical and current states is calculated in combination with sliding window features of historical medical records, when the score exceeds a preset threshold value, it is judged that a dynamic evolution risk exists, an advanced intervention instruction is automatically generated, and early intervention is achieved. Therefore, the problems of single data dimension, risk identification lagging and untimely intervention of a traditional method are solved.

This invention proposes an aneurysm image segmentation method, a rupture risk prediction method, and a system. It constructs a 3D deep learning framework combining multi-view convolution and Laplacian feature pyramids. The multi-view convolution module models the 3D aneurysm structural features from multiple orthogonal directions, and a gated Laplacian frequency fusion mechanism is introduced to compensate for the loss of high-frequency detail information during downsampling, thereby improving the ability to characterize small vessels and complex boundary regions. Furthermore, the morphological features obtained from segmentation, the deep visual features of CTA images, and patient clinical information are jointly modeled. A cross-attention mechanism is used to achieve effective interaction between morphological semantics and image features, thereby completing a binary classification prediction of aneurysm rupture risk. This invention helps to fully explore the complementarity between imaging information and clinical features, more closely reflecting real clinical diagnosis and treatment processes.

The invention discloses a device for acute abdominal disease recognition and a storage medium. The apparatus comprises: a processor; the device realizes the following operations: constructing a knowledge graph for acute abdominal disease identification; performing multi-dimensional sub-graph division on the knowledge graph, forming a corresponding coherent text, and obtaining a multi-dimensional sub-graph set; calculating the correlation of each sub-graph, and extracting a target sub-graph set with the maximum correlation from the multi-dimensional sub-graph set; inputting the clinical information and the target sub-atlas into a language encoder for text encoding to obtain text features; based on the image data, performing image encoding by using an image encoder, and extracting image features; and based on the text features and the image features, a multi-mode decoder is used for decoding, and an identification result of acute abdominal disease identification is obtained. By using the scheme of the invention, accurate and credible identification results and decision support can be provided for clinic.

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 invention discloses a fundus lesion evaluation method and device for chronic nephropathy, and the method comprises the steps: obtaining an ultra-wide-angle fundus image of a to-be-detected target object with chronic nephropathy, and extracting fundus features from the ultra-wide-angle fundus image; based on the extracted fundus features, utilizing a multivariable regression model to predict a fundus lesion state; performing multi-modal fusion processing on the fundus lesion state, the patient information of the to-be-detected target object and the clinical information; and performing lesion probability prediction processing of a dynamic confidence interval on the multi-modal fusion features to generate a chronic kidney disease fundus lesion risk level. Therefore, the fundus lesion risk grade of the chronic kidney disease can be accurately predicted, and the reliability of evaluation of the fundus lesion risk grade of the chronic kidney disease is improved; in addition, a set of evaluation standards for the risk level of fundus lesions complicated by early chronic nephropathy are also constructed, so that the purposes of early discovery and early treatment of patients are achieved, and irreversible fundus lesions caused by CKD are reduced.

The application provides a rule cultivation quality full life cycle monitoring method and system, which comprises the following steps: in a training period, collecting operation logs of rule cultivation personnel in multiple clinical information systems, and reconstructing the operation logs into behavior trajectory sequences reflecting clinical core capabilities; evaluating the quality deviation degree of the rule cultivation personnel in the clinical case processing process based on the behavior trajectory sequences; obtaining the skill mastering vector of the rule cultivation personnel in each clinical core capability based on the quality deviation degrees of multiple clinical operations in a historical period and a preset capability dimension mapping rule; generating the capability evolution characteristics of the rule cultivation personnel in the training period according to the change law of the skill mastering vector over time; and generating a multi-dimensional evaluation result including skill maturity, risk deviation degree, diagnosis and treatment logic consistency and clinical efficiency factors based on the capability evolution characteristics and the quality deviation degree. The application improves the timeliness and accuracy of rule cultivation quality evaluation.

The present invention provides a computer implemented method for the prediction of clinical outcomes in patients with cancer or pre-neoplastic conditions through the integration of genomic evolutionary, genomic and clinical. More specifically, the invention provides systems and algorithms that generate prognostic and predictive models based on the combined analysis of molecular features, inferred evolutionary routes, and clinical parameters, enabling patient risk stratification and individualized outcome estimation.

The invention discloses a dynamic trajectory construction and visual mapping method for association of mental disorder brain network damage and systemic system diseases, and belongs to the field of artificial intelligence medical application. According to the method, high-resolution MRI images, biomarkers and clinical information of major mental disorder patients are collected, and the influence of factors such as age, gender, medication and diagnosis on the braingut axis and the cardio-cerebral axis is evaluated through multi-modal data fusion. By constructing a disease dynamic trajectory model, brain structures and function change modes corresponding to different mental disorders are identified. Large-scale samples are analyzed through machine learning, potential risks and protection factors are extracted, and a visual tool is developed to visually display changes of the brain under different disease systems. A closed-loop feedback mechanism is established through follow-up visit, the disease progress and the intervention effect are dynamically tracked, and key evaluation indexes are identified. According to the invention, theoretical basis and practical guidance are provided for early screening, precise intervention and personalized treatment of mental disorders, and the diagnosis and treatment accuracy and efficiency are remarkably improved.

The invention provides an image-clinical characterization-combined multi-endpoint prognosis evaluation method for transjugular vein intrahepatic portal vein shunt, which comprises the following steps of: constructing a few-label portal vein segmentation module to obtain a preoperative CT portal vein label of a full-dose patient, and extracting deep learning features and radiomics features of the region; establishing a multi-modal interactive representation learning module for implementation, and performing cross-modal fusion with clinical features to form unified representation; and designing a multi-endpoint prognosis prediction module, inputting the data to a plurality of prognosis task decoders for postoperative survival, portal vein pressure gradient, hepatic encephalopathy prediction and the like, and adopting a multi-task learning optimization model to obtain a postoperative multi-endpoint prognosis evaluation result. According to the method, efficient fusion and multi-endpoint prognosis prediction of images and clinical information can be realized under limited labeling, clinical doctors can be assisted in preoperative patient screening and treatment scheme making, and the method has good clinical application value.

The present application relates to the field of liver cancer immunotherapy efficacy evaluation method combining imageomics and deep learning, and specifically discloses a liver cancer immunotherapy efficacy evaluation method combining imageomics and deep learning. The method comprises the following steps: acquiring multi-modal medical images and clinical information of a liver cancer patient; performing standardization preprocessing and automatic lesion segmentation on the images; extracting features through a multi-scale deep network and realizing semantic alignment by using a cross-modal attention mechanism; fusing the images and the clinical data to construct a multi-source heterogeneous feature matrix; adopting a hierarchical model structure, modeling the spatial distribution of tumor immune microenvironment by using a graph neural network at the bottom layer, dynamically tracking the evolution of efficacy by using a gated recurrent unit at the upper layer, and finally outputting an immune response probability, a tumor load trend and a treatment response grade. The present application can objectively and quantitatively evaluate the efficacy of liver cancer immunotherapy, and improve the precision and automation level of individualized diagnosis and treatment decision-making.

This invention discloses a method and system for predicting severe ischemic events in the nervous system caused by large vessel arteritis based on multimodal deep learning. Belonging to the field of intelligent medical image analysis and clinical risk prediction technology, the method includes: acquiring the patient's magnetic resonance angiography data and clinical information; preprocessing the image data; extracting three-dimensional morphological feature vectors from the point cloud of the blood vessel surface based on a point cloud neural network; extracting blood vessel wall feature vectors from the images based on a three-dimensional convolutional neural network; encoding clinical information to generate clinical feature vectors based on a pre-trained basic model; fusing the above three feature vectors to generate a multimodal fusion feature representation; and outputting an individualized risk score based on this fusion feature through a discrete-time survival analysis model. This invention, through multimodal deep collaborative modeling, achieves automated deep fusion analysis of three-dimensional vascular morphology, blood vessel wall structure, and clinical information, significantly improving prediction accuracy, objectivity, and clinical applicability.

The application discloses a method, device and equipment for predicting pCR based on DCE-MRI, the method comprises the following steps: acquiring a DCE-MRI set and auxiliary information corresponding to the DCE-MRI set; inputting the DCE-MRI set and the auxiliary information into a trained pCR prediction model; determining image features corresponding to the DCE-MRI set through the pCR prediction model; determining target features based on the image features and the auxiliary information; and predicting a pCR category based on the target features. The application takes SER information as supplementary information of clinical information, combines the auxiliary information including the clinical information and the SER information with the DCE-MRI set, so that the pCR prediction model can learn the image features carried by the DCE-MRI set and the clinical features carried by the auxiliary information, and combines the image features and the clinical features to predict the pCR, thereby improving the accuracy of the pCR prediction, and further improving the accuracy of the NAC reaction evaluation.

The application relates to a chronic kidney disease prediction method based on clinical information graph representation, belongs to the technical field of computer-aided diagnosis of chronic kidney disease (CKD), and aims to solve the problem of missed diagnosis caused by the fact that the kidney function index of early CKD is not obvious. The incidence of CKD is high, the early symptoms are hidden, and the disease is easy to be missed, thus developing into end-stage renal disease. Therefore, the application provides a chronic kidney disease prediction method based on clinical information graph representation. The method first extracts fundus image and clinical index features; the clinical index is fused into a clinical index joint feature through text embedding and numerical feature fusion, and the joint feature is fused with the fundus image feature through cross-modal attention; the similarity between subjects is calculated based on the fused feature, and a subject relationship graph is constructed by using an adaptive dynamic threshold mechanism; finally, a hybrid graph neural network is used for graph representation learning, pathological similarity between subjects is mined, and accurate prediction of chronic kidney disease is realized. The application can improve the detection rate of early CKD and is applied to non-invasive early screening and risk early warning.

The present application relates to the field of otology disease prediction, and particularly relates to an otology disease prediction method based on multi-modal data fusion and confidence evaluation. The technical scheme comprises: collecting an ear endoscope digital image, wideband tympanometry measurement data, and structured or unstructured clinical information; performing integrity check and quality check on the collected data, checking whether the ear endoscope digital image resolution meets the minimum pixel requirement and whether the information is complete; then performing otology data preprocessing and multi-modal feature extraction, extracting an ear endoscope digital image feature vector, a feature vector representing wideband tympanometry data, and a clinical information feature vector; after feature extraction, fusion is performed, and after fusion, confidence evaluation and recurrence risk prediction are performed. The present application significantly improves the accuracy of otology disease prediction by deeply fusing image, physiological signal and clinical text three modal data through an attention mechanism. The present application is suitable for otology disease prediction.

The application relates to a method and device for predicting the risk of depression of a post-stroke patient, a medium and an electronic device. The method comprises: acquiring a brain image of the post-stroke patient and a standard graph of a depression functional network, the standard graph of the depression functional network being used to represent a brain functional network related to post-stroke depression; calculating a network damage score of the post-stroke patient according to the brain image and the standard graph of the depression functional network; acquiring a standard graph of depression structural disconnection, the standard graph of the depression structural disconnection being used to represent a brain disconnection distribution related to post-stroke depression; calculating a structural disconnection score of the post-stroke patient according to the brain image and the standard graph of the depression structural disconnection; inputting the network damage score, the structural disconnection score and clinical information data of the post-stroke patient into a prediction model to determine the risk of depression of the post-stroke patient after several months of onset through the prediction model. The application can predict the risk of depression according to the direct impact of a stroke lesion on a brain network responsible for emotion regulation in the brain.