343 results about "Imaging study" patented technology

The invention belongs to the technical field of medical data processing and artificial intelligence, and particularly relates to a cardiovascular disease risk prediction system based on multi-modal fusion, which comprises a multi-modal data acquisition and preprocessing module, a cross-modal association graph construction module, a dynamic fusion and prediction module based on a graph neural network and an interpretability analysis module. By constructing a heterogeneous graph fusing prior knowledge and data driving and utilizing a graph attention network to perform multi-level dynamic feature fusion, deep integration and interaction of multi-modal data such as genomes, iconography, clinical and intestinal flora metabolism are realized, so that the accuracy and interpretability of cardiovascular disease risk prediction are improved.

The invention relates to the technical field of hepatitis diagnosis and treatment management, and discloses a block chain-based hepatitis patient full-period management method. The method comprises the following steps: creating a patient identity chain in a block chain network, and storing patient registration data and initial diagnosis information; full-cycle diagnosis and treatment data of a patient are obtained and divided into clinical examination, iconography, medication records and life sign monitoring data. Performing multi-dimensional cleaning on the clinical examination data to generate a standardized data set; inputting the iconography data into a pre-trained hepatitis feature extraction model, and outputting a liver injury feature set; verifying the integrity of the medication record, and generating a non-tampering medication time sequence chain; comparing the life sign monitoring data with preset health reference parameters to generate a sign deviation coefficient set; according to the method, data safety and traceability are guaranteed through the block chain, diagnosis and treatment data quality is improved through multi-dimensional data processing, patient privacy is protected, and reliable support is provided for full-period management of hepatitis patients.

The invention relates to the technical field of medical equipment, and provides a multi-modal lymphedema assessment and surgical navigation system based on image processing, which comprises a 3D scanning modeling module used for acquiring three-dimensional point cloud data of the body surface of a patient through structured light 3D scanning equipment, an iconography examination analysis module used for performing U-Net image segmentation on CT / MRI image data, and an image processing module used for processing the CT / MRI image data. The ultrasonic result calculation module is used for carrying out Canny edge detection on an ultrasonic image to determine the boundary of a surgical site, and the data integration and navigation generation module is used for realizing multi-modal data fusion through mutual information maximization, generating a surgical path based on FMM and carrying out real-time navigation in combination with an AR technology. According to the method, objective evaluation of lymphedema and accurate navigation of the operation are realized through accurate integration of multi-modal data and an image processing technology, the diagnosis and treatment accuracy is effectively improved, the operation risk is reduced, complications are reduced, and a scientific basis is provided for personalized treatment.

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 discloses an image text report quality control method and system based on LLM and a structured report, relates to the technical field of medical text report quality control, and solves the problems that a current text report quality control method is difficult to accurately extract key fields and is difficult to judge complex diagnostic reasoning logic. According to the technical scheme, the method is characterized in that iconography feature description related to a diagnosis conclusion is extracted from a text report, cue words are constructed according to a matched structured report template to analyze the iconography feature description into structured data, and an interface is called to fill the structured data into the structured report template. And obtaining a template diagnosis conclusion according to the built-in judgment logic of the structured report template, and comparing the template diagnosis conclusion with the current diagnosis conclusion to realize quality control auditing of the text report.

This invention discloses a system for predicting the efficacy of inflammatory bowel disease based on multimodal data, relating to the field of medical information processing. The system includes: a multimodal data acquisition module for receiving multimodal data; a text feature extraction module for determining text feature vectors using a text feature extraction model; an endoscopic intestinal mucosal feature recognition module for determining intestinal mucosal features using an endoscopic intestinal mucosal feature recognition model; an image feature recognition module for determining imaging features using an image feature recognition model; a pathological feature extraction module for determining pathological features using a pathological recognition model; and an efficacy prediction module for predicting the efficacy of biologics based on multimodal data features using a Transformer model and fully connected layers. Compared to existing technologies, this invention achieves efficacy prediction of biologics based on the fusion of clinical, pathological, endoscopic, and imaging multimodal features, improving the accuracy and clinical rationality of diagnostic and treatment decisions.

The invention relates to the field of medical information technology and medical data analysis, in particular to a VTE risk prediction method and system based on multi-modal data fusion and dynamic risk modeling, and the system comprises a first feature extraction module, a second feature extraction module, a two-stage risk assessment module and a final risk judgment module. According to the application, through integration of clinical diagnosis and treatment, iconography and genomics data, a deep learning algorithm, a dynamic weight adjustment mechanism and a graph neural network technology are utilized to realize accurate prediction of the VTE risk of the active-stage lung cancer patient. And meanwhile, the problem of cross-mechanism data distribution difference is solved by adopting transfer learning, and the generalization ability of the model is improved. The method can significantly improve the prediction accuracy, reduces the manual intervention demands, and optimizes the medical resource distribution efficiency.

The invention relates to the field of medical data analysis, in particular to a patient multi-dimensional feature clustering analysis method for a mental health database. The method comprises the following steps: acquiring multi-modal data of a patient from a mental health database, and performing missing value filling and standardization processing; constructing a biomarker pathway mapping knowledge domain, mapping the genetic information of the patient and biomarker data to nodes of the mapping knowledge domain, and extracting pathway activation features; fusing the dimensionality-reduced iconography and clinical features with the pathway features by adopting a multi-modal embedding method to obtain a comprehensive feature vector of the patient; grouping the patients by using a metric learning clustering algorithm and combining with pathway consistency constraints to obtain subtypes with biological significance; and further analyzing the characteristic contribution degree and the molecular mechanism difference of each subtype by adopting an interpretability method. According to the method, multi-modal data can be effectively integrated, the clustering precision and interpretability are improved, and a basis is provided for subtype recognition and personalized intervention of mental health diseases.

The invention belongs to the technical field of data processing, and particularly discloses a method and system for predicting the lifetime of a non-small cell lung cancer patient after radiotherapy, and the method comprises the steps: collecting multi-time-point imaging and hematology follow-up visit data of the patient after radiotherapy, building a longitudinal follow-up visit sequence, and mapping the longitudinal follow-up visit sequence to a unified time axis; determining a bimodal first-time significant improvement point through preset rule judgment, and constructing and standardizing an asynchronous index according to the bimodal first-time significant improvement point; then inputting a feature vector at a follow-up time point, encoding a bimodal sequence through a Transform time sequence encoder, and generating joint representation through cross-modal interaction; and finally, inputting the standardized asynchronous index, the joint representation and the clinical and radiotherapy characteristics into a DeepHit model, and outputting a patient survival distribution prediction result. According to the method, bimodal indexes can be accurately captured, time dislocation is improved, prediction accuracy and individualization degree are improved, and a reliable basis is provided for clinical prognosis evaluation.

The invention relates to an oxidase response type molecular probe as well as a preparation method and application thereof, and belongs to the technical field of medical imaging. Aiming at the problem that the inflammation activity is difficult to accurately evaluate in the prior art, the invention provides a molecular imaging probe capable of responding to the activity of in-vivo myeloperoxidase (MPO), and the molecular imaging probe can be used for real-time and non-invasive imaging monitoring of the activity of the in-vivo MPO, so that the activity of the inflammatory reaction of a focus is accurately evaluated. The invention provides a chelating agent with structural characteristics of a chelating unit and an enzyme response unit, and is characterized in that the enzyme response molecular probe is constructed by taking rigid framework cyclohexanediamine (CDTA) as a transition metal chelating unit and taking electron-rich functional groups (such as phenol, catechol or 5-hydroxytryptamine) as the enzyme response unit. The probe is chelated with paramagnetic metal (such as Mn < 2 + >) for magnetic resonance imaging (MRI) or radioactive metal (such as 68Ga) for positron emission tomography (PET), so that specific response to MPO activity is realized. The dynamic stability of the probe is remarkably improved; after the MPO is responded, the relaxation efficiency is improved by 2-3 times, and the sensitivity is optimized; the cytotoxicity is low; in an acute pancreatitis model, the MRI contrast noise ratio (delta CNR) reaches 77.39 + / -7.04 and is 4.2 times that of a control group, the PET uptake rate is remarkably improved, bimodal imaging is supported, and accurate positioning and quantitative evaluation of the inflammation activity are achieved.

The invention discloses a hip joint surgical robot grinding and rubbing detection system and a detection tool thereof, and relates to the technical field of medical instruments. The device comprises a bottom plate and a grinding and rubbing assembly, a hip model assembly is fixedly connected to one side of the top of the bottom plate, an acetabular fossa model assembly is fixedly connected to the other side of the top of the bottom plate, and a tracer support is fixedly connected to the position, located at the rear end of the acetabular fossa model assembly, of the top of the bottom plate. The grinding and rubbing precision can be monitored in real time in the operation process, immediate feedback is provided, an operation team can find and correct errors in the grinding and rubbing process in time, the success rate of the operation is increased, tiny grinding and rubbing errors can be detected by accurately measuring the ball center position of the grinding and rubbing contour, it is ensured that the surgical robot achieves high-precision grinding and rubbing operation, and the working efficiency is improved. Multiple times of X-ray or CT scanning needed by traditional postoperative imaging evaluation are avoided, the radiation risk of a patient is reduced, the grinding and rubbing precision detection process is simplified, the operation time is shortened, and the operation complexity is reduced.

The invention discloses a searcher model training method, and a Burt-Hooge-Durob syndrome recognition method and system based on retrieval enhancement generation, and relates to the field of rare disease recognition. The training method comprises the steps of obtaining a positive sample pair and a negative sample pair; the positive sample pair and the negative sample pair are input into an initial searcher model, a loss function of the initial searcher model is calculated, and the loss function adjusts the size of an angle margin in real time according to a measurement variance adaptive mechanism; and optimizing the initial retrieval model according to a loss function calculation result. According to the method, the angle interval between the BHD and the non-BHD is forcibly expanded through the angle margin of the loss function, and the angle margin is dynamically adjusted according to the statistical variance of the cosine similarity between all positive sample pairs in the current training batch by using a measurement variance adaptive mechanism; the problems of weak DCLDs image difference and fuzzy category decision boundary caused by very similar iconography features of various types of rare diseases of DCLDs are solved, and the recognition precision of a large model on query information is improved.

The invention discloses a medical image report automatic generation method and device based on artificial intelligence, relates to the technical field of artificial intelligence and medical image crossing, and aims to shorten the report generation period and improve the report generation efficiency while improving the accuracy and consistency of report diagnosis. The method comprises the following steps: carrying out cross-modal image space alignment processing on a PET image and a CT image; a pre-trained PET focus segmentation model is adopted, a high-metabolism focus part is identified and segmented in the processed PET image, and focus iconography parameters are acquired; segmenting an organ image in the processed CT image by adopting a pre-trained CT image organ segmentation model, and determining focus position information; sorting the lesion iconography parameters and the lesion position information to generate an initial report, and processing the initial report by adopting the fine-tuned large language model to obtain a medical image report and outputting the medical image report.

The invention discloses an intracranial SEEG electroencephalogram data analysis method and system based on artificial intelligence, and relates to the technical field of data analys.The method comprises the steps that original electroencephalogram signals are obtained, and electrode space coordinates and imaging information are synchronized; extracting robust features in a time-frequency domain through synchronous extrusion in combination with learnable chirp and multi-resolution attention; according to the electrode coordinates and the anatomical topology, topology sensing alignment is executed, and an anatomical function diagram is constructed; performing continuous time evolution modeling on a node state through a graph neural continuous dynamic structure, and introducing an energy conservation and event jump mechanism to realize focus transmission chain inference; estimating orientation information in different physiological contexts, and screening a stable direction as a prior constraint; and outputting a continuous risk curve according to propagation embedding, and generating a coverage-controllable confidence interval through online conformal calibration. Energy consistency, causal stability and space-time continuous modeling of SEEG signals can be realized, and the accuracy and interpretability of electroencephalogram focus recognition and clinical risk assessment are improved.

Systems and methods are described that implement displaying grouped medical images for review. Current medical image management systems generally use thumbnail image grouping of images that have the same position and side. These groups of thumbnail images can be large, making it difficult for a clinician (e.g., a radiologist) to quickly find a particular view they are interested in reviewing. The disclosed systems and methods provide for user-selectable grouping and ungrouping of thumbnail images in a graphical user interface, which makes it easier for a clinician to quickly find a particular view that they are interested in and enables a more-efficient review of an imaging study.

The invention relates to the technical field of control of intelligent implanted medical instruments, and discloses an aneurysm neck sealing system and a control method. Digital subtraction angiography is adopted to measure the neck diameter in multiple directions, an average value is calculated, the target neck sealing radius is set by combining the size of an embolism device, and it is ensured that instrument type selection is scientific and reasonable. The triggering temperature and recovery kinetics of the shape memory polymer are determined through a constant-temperature experiment, an actual expansion radius model is established, and quantitative control over the heating process is ensured. A heat balance model is established by using the heat capacity and cooling curve of the embolism device, and temperature rise and heating power expressions are deduced, so that fine regulation and control of temperature and power are realized. And natural cooling is performed after heating to a set terminal point, the plugging effect is evaluated in real time through iconography measurement, and compensation heating or model replacement is performed according to actual differences to ensure the plugging effect.

The invention discloses a specific disease iconography feature extraction method and system based on deep learning, and relates to the technical field of image analysis, and the method comprises the steps: collecting X-ray image data, stem cell markers, magnetic resonance imaging image data and dynamic ultrasonic image data of an osteoarthritis patient; performing standardization processing on the X-ray image data, the stem cell marker, the magnetic resonance imaging image data and the dynamic ultrasonic image data to obtain a standardized X-ray image data set, a standardized stem cell marker, a standardized magnetic resonance imaging image data set and a standardized dynamic ultrasonic image data set; a standardized X-ray image data set, a standardized magnetic resonance imaging image data set and a standardized dynamic ultrasound image data set are acquired. According to the invention, through multi-modal image data fusion and biomechanics-stem cell characteristic dynamic coupling, early accurate diagnosis and repair potential evaluation of osteoarthritis are realized.

The invention relates to the field of medical imaging and clinical data machine learning, and discloses an ICAS stroke risk prediction method and system, and the method comprises the steps: carrying out the preprocessing of image data; clinical data such as past medical history, clinical symptoms and family history of the patient are collected; analyzing related biomarkers of the blood sample to obtain biomarker data, and selecting variables closely related to the stroke risk; then carrying out feature splicing to form a multi-modal fusion feature vector; and finally, training, verifying and testing by adopting a machine learning algorithm, and constructing to obtain an ICAS stroke risk prediction model. And based on the ICAS stroke risk prediction model, generating a stroke risk score for each patient, and generating a clinical decision support strategy. The performance of the stroke risk prediction method and system can be improved continuously, accurate risk prediction can be provided in different clinical environments, and accurate stroke risk assessment and targeted treatment schemes can be provided for ICAS patients.

The invention provides a pulmonary nodule patient treatment prediction method based on artificial intelligence, and the method comprises the steps: obtaining a historical tongue picture feature and a historical pulmonary nodule CT image, extracting a tongue coating color value RGB component, a spatial frequency of a tongue surface texture feature, a gradient value of pulmonary nodule edge sharpness, and a CT value standard deviation of pulmonary nodule density distribution, and obtaining an initial feature set; extracting the change trend of the tongue coating thickness and the pulmonary nodule edge sharpness in the treatment cycle from the weighted feature set, analyzing the dynamic evolution rule of the correlation between the change rate of the tongue coating thickness and the gradient value of the pulmonary nodule edge sharpness, and obtaining a time sequence feature vector in combination with the time rhythm characteristics of the traditional Chinese medicine tongue diagnosis; and carrying out standardization and normalization processing on the fusion feature set, predicting by combining a clinical curative effect evaluation standard and a pulmonary nodule iconography improvement index to obtain the correlation strength of the change rate of the thickness of the tongue coating and the improvement trend of the edge sharpness of the pulmonary nodule in a treatment cycle, and obtaining a predicted curative effect result.

The invention relates to a lung elastography and AI-assisted pulmonary nodule benign and malignant identification system and a medium, and solves the problem of limited identification accuracy of a traditional iconography examination method. The system comprises at least one processor which is configured to realize the lung elastography and AI-assisted pulmonary nodule benign and malignant identification method when executing a computer program, and the specific steps are as follows: inputting obtained comprehensive consistency indexes, individual characteristics of patients and key information in multi-source data into a pre-trained correlation model; outputting the prediction probability that the case pulmonary nodule is malignant; based on the prediction probability and the initial confidence coefficient, adopting a preset confidence coefficient adjustment algorithm to calculate the adjusted confidence coefficient; and displaying the pulmonary nodule benign and malignant judgment result, the confidence coefficient, the multi-physical field simulation result and the comprehensive consistency index on a display terminal in a multi-modal mode. The method has the advantages that the accuracy of identifying benign and malignant pulmonary nodules is improved, and the limitation of a traditional method is made up.

The invention relates to the technical field of biological detection, and discloses a reagent composition for detecting urothelial carcinoma, a kit and application. The composition provided by the invention is used for detecting the methylation state of a specific target sequence in a characteristic gene of the urinary tract epithelial cancer, and the detection sensitivity and specificity of the urinary tract epithelial cancer are at a relatively high level. The composition provided by the invention can be used for noninvasive urinary tract epithelium cancer screening, and when the composition is used for urinary tract epithelium cancer screening, compared with conventional urine cytology examination and iconography examination, the detection sensitivity is greatly improved, the early screening accuracy of urinary tract epithelium cancer is improved, and the misdiagnosis rate and the missed diagnosis rate are greatly reduced.

The invention relates to a kidney lump benign and malignant analysis method and device based on a laparoscopic ultrasound image, and belongs to the technical field of medical image processing.The method comprises the steps that radiomics characteristics of the laparoscopic ultrasound image are obtained, and radiomics scores of lump malignant risks are calculated according to the radiomics characteristics; determining an independent risk factor corresponding to the patient clinical variable, and constructing a clinical prediction model according to a mapping relationship between the patient clinical variable and the independent risk factor; and according to the radiomics score and the clinical prediction model, constructing a kidney lump benign and malignant analysis model. The technical problem that in the prior art, massive quantitative image features cannot be efficiently and accurately mined from medical images, and the most valuable iconography features cannot be screened out for analyzing clinical information is solved.

The invention relates to the technical field of medicine, in particular to an early evaluation method for endocrine remission of pituitary growth hormone adenoma based on low Knosp grading, which comprises the following steps: acquiring preoperative and postoperative image data and preoperative and postoperative clinical data of a patient with low Knosp grading pituitary growth hormone adenoma; extracting imaging features used for early evaluation of postoperative endocrine remission from the preoperative and postoperative image data; extracting clinical features for early evaluation of postoperative endocrine remission from the preoperative and postoperative clinical data; and performing statistical analysis on the iconography characteristics and the clinical characteristics to obtain an early evaluation result of postoperative endocrine remission. According to the method, risk factors related to whether low-Knosp graded pituitary growth hormone adenoma postoperative endocrine remission can be relieved or not are screened out, then a prediction model is established through a statistical method, and then early prediction of low-Knosp graded pituitary growth hormone adenoma postoperative endocrine remission is achieved.

The invention discloses a cerebral artery blood flow velocity prediction method, device and equipment and a medium, and relates to the field of blood flow velocity prediction.The method comprises the steps that cerebral artery nuclear magnetic imaging data are obtained and preprocessed; performing image segmentation to obtain a cerebral artery three-dimensional model; extracting geometric topology information of the cerebral artery; constructing a 0-1-dimensional geometric multi-scale model of the crowd averaged cerebral artery; determining entrance boundary conditions of the 0-1-dimensional model to obtain a 0-1-dimensional patient personalized hemodynamic model; according to a clinical preset contrast agent concentration-time curve and based on the law of conservation of mass, a 0-1-dimensional patient personalized hemodynamic model is used for simulating a contrast agent injection process, and a contrast agent concentration-time curve of each position in the cerebral artery topology network is obtained; based on the concentration data of the contrast agent at the near end and the far end of the target blood vessel, the flow velocity of the contrast agent at different blood vessel segments of the cerebral artery is determined, and the method can be suitable for blood flow velocity estimation of an artery system with a complex topological structure.

According to the subclinical non-obstructive coronary atherosclerosis metaphase plaque progress risk prediction method based on interpretable machine learning provided by the invention, multi-dimensional images and clinical features are integrated, and individualized risk prediction of subclinical non-obstructive coronary atherosclerosis is realized. Quantitative imaging biomarkers are managed for a long time, a key tool is provided for early screening of high-risk individuals, clinical doctors can be helped to judge whether comprehensive strategies such as enhanced lipid lowering, anti-inflammation and lifestyle intervention need to be adopted in time or not, plaque stabilization or fading is maximized, and the clinical effect is improved. And further, the future cardiovascular event risk is reduced. According to the method, a Shapley additive SAHP interpretation technology is introduced, the relative contribution of each prediction feature is quantified, and the interpretability of a machine learning model is improved. According to the method, a webpage tool for calculating the plaque progress probability on line is deployed, and the contribution degree of each feature to the plaque progress prediction result of a specific patient can be displayed more visually.

This invention discloses a method and system for following up on magnetic resonance imaging (MRI) images of amyloid-related radiological abnormalities, belonging to the field of medical image processing. The method includes: acquiring multimodal MRI image data; performing bias field correction and intensity normalization based on a bias field fitting model to generate a first image set; performing global alignment and edge detection on the first image set to extract anatomical physical boundaries, generating brain region mask data; performing branch deformation field registration on the first image set within the brain parenchyma defined by the brain region mask data to generate a spatially aligned second image set; calculating the average intensity of normal brain parenchyma regions in the second image set to obtain a scaling factor, acquiring a three-dimensional residual image; performing anisotropic diffusion filtering denoising on the three-dimensional residual image, and performing differential denoising and morphological screening processing, outputting the detection results. This invention eliminates systematic artifacts caused by insufficient bias field correction.

This invention discloses an AR-assisted puncture navigation spatial registration method and system based on environmental reference features, relating to the field of medical image navigation. The system includes an augmented reality navigation device, a permanent ceiling marker, and a temporary marker board. The augmented reality navigation device includes a binocular augmented reality camera group, a global tracking camera, a global spatial registration module, and a display module. The binocular augmented reality camera group constructs a binocular overlapping field of view, acquiring the pose information of the binocular augmented reality camera group relative to the temporary marker board during the calibration phase, and acquiring near-field images of the surgical area in real time during the surgical phase and displaying them on the display module. The global tracking camera acquires the pose information of the global tracking camera relative to the global coordinate system of the ceiling. The global spatial registration module solves the extrinsic parameter matrix between cameras and the environmental relationship matrix based on the pose information. During the surgical phase, the path planned by the imaging equipment is converted into a real-time navigation path and superimposed on the display module. This solves the complex calibration problem under non-overlapping fields of view.

The invention relates to the technical field of biology, and particularly discloses a tumor imaging biomarker quantitative analysis system based on multi-mode MRI voxel space coupling and application. According to the application, the spatial coupling relationship between blood brain barrier destruction and perfusion in a tumor can be quantitatively reflected through the voxel overlap region proportion of T1CE and CBF signals, the 1p / 19q state can be efficiently predicted, and accurate prediction of IDH mutation and WHO grading (2-3 grade vs.4 grade) is realized in cooperation with clinical factors, so that the application has a wide application prospect. And an efficient and potential noninvasive imaging biomarker is provided for ADG preoperative molecular typing and grading prediction.

The invention belongs to the technical field of biological medicine, and particularly relates to application of actinomycin D in preparation of a ctDNA detection sensitizer. The invention provides a ctDNA release method based on actinomycin D as an enhancer, which can significantly improve the release level of ctDNA in early small-volume tumors and tiny residual lesions, thereby improving the sensitivity and positive detection rate of liquid biopsy in MRD detection. Under an extremely low dosage, the ActD can generally promote a plurality of human tumor cells to release ctDNA under the condition of not causing obvious cell death, and has good biological safety and adaptability. The method is especially suitable for tiny, biological inert or anatomical position hidden focuses which are difficult to identify by iconography, namely blind area patient groups in disease monitoring, and has a wide clinical application prospect.

The invention relates to an obesity risk prediction system based on multi-modal iconography, and relates to the field of electrical digital data processing, and the system comprises a layer-by-layer construction device which is used for obtaining an obesity risk prediction model; and the risk prediction device is used for intelligently predicting the fatty liver grade and the easily fat food name of the to-be-analyzed person according to the liver QCT picture information of the to-be-analyzed person and the fMRI imaging data of the obesity-related brain region by adopting an obesity risk prediction model. The invention also relates to an obesity risk prediction method based on multi-modal iconography. According to the invention, the technical problems of insufficient obesity risk parameter analysis precision and single parameter type of the to-be-analyzed person are solved; synchronous intelligent prediction of two obesity risk parameters, namely the fatty liver grade and the easy-to-fat food name, of a person to be analyzed can be realized based on multiple items of basic data screened in a targeted manner by adopting an obesity risk prediction model based on multi-modal iconography of a customized structure, so that the technical problem is solved.