387 results about "Microvessel" patented technology

The invention discloses an ultrasonic positioning microscopic imaging method and system based on joint resolution perception and vision Mangbar, and the method specifically comprises the steps: obtaining and preprocessing a plurality of modal angiography data, generating a microbubble motion sequence in combination with blood flow velocity field simulation data, and constructing a data set; inputting the data set into a mixed hierarchical feature pyramid fusion network, carrying out probability density estimation on a microbubble position based on a real-time multi-scale density field estimation algorithm, and introducing a resolution perception algorithm to enhance microbubble features to obtain a plurality of short-track images; performing dynamic feature decoupling and time-frequency analysis on the short-track image, extracting and encoding velocity component features and spatial-temporal context information of a vascular structure, and constructing a double-branch processing network based on visual Mangbar to obtain images respectively reflecting forward motion contribution and backward motion contribution, and finally generating a high-resolution blood vessel image through image fusion. The method has an important value for improving the efficiency and precision in microvascular imaging.

The invention relates to the technical field of medical robot systems, and discloses a nerve plexus intelligent positioning anesthesia puncture robot system based on multi-modal fusion, which performs time-space synchronization on ultrasonic image flow and contact force data. The frequency domain analysis module extracts a main peak frequency by using optical flow field frequency domain transformation, and generates a probability mask for marking pulsatile tissues. The viscoelasticity verification module calculates phase lag parameters of visual deformation and contact force, identifies tissue rheological properties and calculates a dynamic damping coefficient. The deformation compensation module calculates a mechanical residual error based on a reference mechanical model, and inversely calculates the mechanical residual error into space deformation displacement of the tissue. The feedback control module combines the probability mask and the deformation displacement to generate a motion instruction. According to the invention, through multi-modal data fusion, precise identification of tiny blood vessels, dynamic identification of tissue attributes and real-time compensation of soft tissue conceding deformation are realized.

The invention relates to an adjusting system applied to continuous kidney replacement treatment of acute kidney injury patients. Comprising the following steps: capturing the blood flow velocity and vascular endothelial permeability of glomerular capillaries in real time by using a micro sensor or a nano-scale optical imaging technology; the method comprises the following steps: pre-judging a risk time point of insufficient tissue perfusion through dynamic modeling of a capillary structure and a functional relationship; combining the microcirculation data with conventional biochemical indexes of the patient to create an early intervention model; when it is monitored that microcirculation perfusion is greatly reduced or kidney oxygen supply is insufficient, continuous renal replacement therapy CRRT parameter fine adjustment is triggered in advance; the method comprises the following steps: collecting plasma toxins of different patients at clinical and molecular levels, and constructing a common toxin map of the acute kidney injury AKI; whether the specific toxin concentration exceeds the standard or not is judged by combining a real-time sensor, and the removal priority is output in real time; a membrane material with an adjustable aperture or a special coating is adopted, so that inflammatory mediators and bacterial endotoxin molecules are adsorbed or intercepted, and the permeability of micromolecular electrolyte is kept.

The invention discloses an automatic retinal vessel segmentation method fusing large-kernel multi-scale cavity convolution and an attention mechanism, which is suitable for accurate extraction of a fine-grained vessel structure in a fundus image. According to the method, an improved deep neural network (LKD-UNet) is constructed, a collaborative receptive field module is embedded in each layer of an encoder, and large kernel convolution and multi-scale cavity convolution are combined to enlarge the receptive field and enhance the texture analysis capability; an efficient Transform module is introduced into a bottleneck layer, a cross-regional semantic relationship is modeled through a multi-head self-attention mechanism, and the topological consistency of microvessels is improved; the decoder adopts a shallow symmetric structure and a kernel scale scheduling strategy, and realizes high-resolution reduction in combination with jump connection. In addition, a four-category pixel-level error visualization mechanism is introduced and is used for assisting segmentation error diagnosis and result interpretation. Experimental results show that the method is superior to an existing mainstream model in multiple public data sets, has remarkable advantages in the aspects of segmentation precision, structural coherence and deployment efficiency, and is suitable for remote fundus screening and blood vessel extraction tasks in embedded medical equipment.

The invention relates to ultra-small prussian blue nano-particles (T-USPB-C) carrying tirofian, a preparation method of the nano-particles and application of the nano-particles in preparation of drugs for reducing MVO and MIRI. According to the preparation method, the T-USPB-C is prepared by three steps. The T-USPB-C provided by the invention can carry tirofiban, and has catalase and superoxide dismutase simulated nano-enzyme activity. Moreover, the size of the nano-scale drug is required to reach a nano-scale size, and the drug can be efficiently cleared through kidney excretion, so that the potential long-term toxicity problem is minimized. The ultra-small prussian blue nanoparticle carrying the tirofian has an excellent active oxygen scavenging capability. Microvascular perfusion can be rapidly improved through the antithrombotic effect, then the protection effect is continuously achieved through the anti-oxidation and anti-inflammatory mechanism, and microvascular injury and MIRI are effectively prevented.

The invention belongs to the technical field of biomedical tissue engineering materials, and particularly relates to injectable composite plastic gel and a preparation method thereof. The injectable plastic gel provided by the invention is formed by suspending and compounding the cross-linked sodium hyaluronate gel particles and the biological tissue extracellular matrix particles, the two material particles in the injectable composite plastic gel are uniform in size and controllable in size, other toxic catalysts do not need to be added in the preparation process, and the injectable plastic gel is easy to prepare. The collagen / elastin composite material is rich in collagen and elastin, and shows good biomechanical properties. Meanwhile, the composite plastic gel contains a plurality of bioactive factors, and after the composite plastic gel is injected and filled into a body, the composite plastic gel is slowly degraded and can promote the generation and ingrowth of autologous fat cells and microvessels, so that the balance of degradation and remodeling is achieved. The preparation method of the injectable plastic gel provided by the invention can be widely applied to the fields of plastic surgery, medical beauty and the like.

The invention discloses a robot-assisted real-time three-dimensional Doppler OCT (optical coherence tomography) angiography method and system, relates to the technical field of medical imaging, and can be used for carrying out real-time and large-view-field three-dimensional Doppler optical coherence imaging evaluation during a microvascular anastomosis operation and improving the anastomosis operation evaluation accuracy and efficiency. According to the system, a spectral domain optical coherence tomography system SD-OCT is used for performing spectral domain optical coherence tomography on a sample to obtain a sample cross-sectional image, and the sample cross-sectional image is sent to a display to be displayed. The OCT imaging probe and the binocular depth camera are integrally installed at the tail end of the robot arm, and the robot arm operates the OCT probe and the binocular depth camera to move to different positions and angles to image a sample. The OCT imaging probe collects optical signals reflected by a sample, and the signal processing host and the displayer receive spectral signals transmitted by the OCT imaging probe and display the three-dimensional structural characteristics of the microvessel and the blood flow velocity condition in the blood vessel.

The invention discloses a diagnosis model for hepatocellular carcinoma microvascular invasion and a construction method and application thereof, and belongs to the technical field of medical examination. Metabolite in the blood of a hepatocellular carcinoma patient is taken as a cut-in, preoperative clinical tumor indexes are combined, it is found that PIVKA-II, GGT, L-Serine, Sarcosine and tumor sizes have important significance on formation of hepatocellular carcinoma MVI, a diagnosis model based on multiple omics and multiple dimensions is established based on the diagnosis indexes, the diagnosis model is used for diagnosis of the preoperative MVI of the hepatocellular carcinoma patient, and the diagnosis model is used for diagnosis of the preoperative MVI of the hepatocellular carcinoma patient. The defects of existing diagnosis means in the aspect of preoperative diagnosis are overcome, and a more timely and accurate basis is provided for clinical decision making.

The invention relates to the technical field of medical image processing, and discloses an adversarial generative network model for three-dimensional reconstruction of a brain cell-level vascular network. The model comprises a three-dimensional blood vessel feature extraction module, a dynamic attention generator, a multi-scale discriminator and the like. Characteristics are obtained through multi-modal medical image data (magnetic resonance angiography and confocal microscope scanning data), and high-precision vascular network three-dimensional reconstruction is realized by utilizing cooperative work of all the modules. The dynamic attention generator calculates the topological connection probability of the vascular branches, and the multi-scale discriminator globally and locally evaluates the reconstruction result. Meanwhile, model training is optimized through an adversarial training controller, and the blood vessel network is perfected through a capillary network completion module. According to the method, the advantages of multi-modal data are effectively fused, the reconstruction precision is improved, and powerful support is provided for research and diagnosis of brain vascular diseases.

The invention discloses a multi-task learning-based retinal vessel segmentation and analysis method and system, and relates to the technical field of medical image processing, and the method comprises the steps: introducing an EMCAD module into an R-Unet model, and carrying out the retinal vessel segmentation; designing an area percentage quantization function; a multi-task learning framework RetiVas-MTNet is constructed and is used for auxiliary diagnosis of eye diseases. The EMCAD module fuses a channel attention CAB mechanism and a space attention SAB mechanism: 1) generating a channel weight through adaptive maximum pooling and average pooling, and strengthening key channel features; (2) context connection of local features is established through large kernel convolution, and the space attention weight is calculated; according to the invention, the EMCAD module is introduced, so that the segmentation precision of the microvessels is improved; the blood vessel area percentage is quantitatively designed, so that the crossing from qualitative segmentation to quantitative diagnosis is realized; a multi-task learning framework RetiVas-MTNet provides a comprehensive solution for automatic diagnosis of eye diseases.

The invention provides application of a monascus composition in preparation of a medicine for preventing and / or treating coronary microcirculation disturbance. The monascus composition comprises one or more of cortex mori, ramulus mori and folium mori, monascus and radix puerariae. When the composition containing the red yeast rice is applied to preparation of the medicine for preventing and / or controlling coronary microcirculation disturbance, the abnormal microvascular structure can be improved (myocardial microvascular density MVD is increased, edema around microvessels is relieved, the lumen stenosis phenomenon is improved, and the endothelial cell condition is good); 2) microvascular occlusion is improved (pathologic results show that microvascular contents are reduced); (3) the abnormal vasodilatation and contraction function of the microvessels is improved (NO is increased); 4) improving blood lipid metabolism disorder; and (5) the heart function is improved (the left ventricular ejection fraction LVEF and the left ventricular short axis shortening rate LVFS are improved, and the left ventricular pressure maximum rising rate LV + dp / dtmax and the left ventricular pressure maximum falling rate LV-dp / dtmax are increased).

The invention relates to a retina nerve and blood vessel unit in-vitro three-culture model and a preparation method thereof, and belongs to the technical field of biology. The invention provides an in-vitro three-culture model of a retina nerve and blood vessel unit. The in-vitro three-culture model is obtained by carrying out in-vitro Transwell indirect co-culture on retina microvascular endothelial cells, retina ganglion cells and retina Muller cells. The in-vitro three-culture model of the retina nerve and blood vessel unit is based on a three-cell indirect co-culture method, retina microvascular endothelial cells, retina ganglion cells and retina Muller cells are simultaneously incorporated by utilizing Transwell, and the three cells are in indirect contact. The in-vitro three-culture model of the retina nerve and blood vessel unit can be used for more objectively simulating pathophysiological states of various cells of the retina nerve and blood vessel unit, so that common pathophysiological changes of mutual dependence and interaction of the three cells under in-vitro conditions can be observed.

The invention belongs to the technical field of medical nano-materials, and particularly relates to a lactoferrin nano-material for sonodynamic regulation and control of blood-brain barrier permeability as well as a preparation method and application of the lactoferrin nano-material. The nanomaterial is nanoparticles LCe6 formed by self-assembly of lactoferrin and chlorin e6, can target brain microvascular endothelial cells, and can temporarily, reversibly and targeting open BBB by using the sonodynamic effect of the LCe6 in combination with transcranial ultrasonic stimulation, so that the accumulation of the nanoparticles LCe6 at a focus part is improved; in addition, the nanoparticle LCe6 also has a targeting property on brain tumor cells, the tumor cells can be accurately killed by utilizing the sonodynamic effect of the nanoparticle under secondary transcranial ultrasound, and the effect of sonodynamic treatment on brain glioma is exerted to the maximum extent; the nanoparticles LCe6 have excellent biocompatibility and low toxicity; according to the invention, the intracerebral delivery efficiency of the drug can be obviously improved, and a novel method for targeting open brain disease BBB is provided.

The invention belongs to the technical field of biomedicine, and particularly relates to application of ceramide as a marker in prediction of hepatocellular carcinoma microvascular invasion risk. The invention finds that the preoperative plasma ceramide CER (d18: 1 / 16: 0) level can be used for predicting the MVI of HBV-HCC patients and evaluating the postoperative prognosis of the patients for the first time. The invention provides a preoperative non-invasive detection method based on plasma ceramide CER (d18: 1 / 16: 0), and trauma of traditional histopathology detection is avoided. According to the method, the occurrence risk of the MVI can be effectively reduced by simply detecting a blood sample of a patient. According to the method, the MVI risk of a patient is evaluated through a multi-factor prediction model, and a basis is provided for personalized treatment. The method is not only suitable for preoperative MVI risk assessment, but also can assist in making postoperative follow-up visit and treatment plans, and helps to reduce the postoperative recurrence rate and improve the survival rate of patients.

A multi-sensor remote patient monitoring platform combines multiple technologies on a patient's own device to dramatically reduce the number of devices a patient needs while also reducing device distribution logistics on health care providers. Continuous temperature monitors can be used to collect frequent temperature data over large periods of time. Additionally, a rPPG device can obtain at least one vital sign data by a remote photoplethysmogram (rPPG) comprising a plethysmogram that is optically obtained via a camera by detection of blood volume changes in the microvascular bed of tissue of the patient. The temperature data and vital sign data can be input into a remote server-side computer system that can analyze the received data for indication of a fever, and provide an alert to the patient or a clinician when the result of the analysis indicates the fever. Such alerts can be clinically beneficial for disease states in which fever can be fatal, particularly for those fevers having fast onset.

The invention provides a tissue microvessel analysis method and system for assisting drug design based on artificial intelligence. The method comprises the steps that two channel images of a blood vessel marking channel and a nano probe channel are collected; carrying out preprocessing, registration and feature fusion on the two channel images to obtain a fusion feature tensor for simultaneously coding a blood vessel-nanoparticle space coupling relationship; extracting multi-scale structure information based on the fusion feature tensor to obtain an enhanced feature tensor; constructing double models based on a U-Net + + architecture, and combining the double models with an enhanced feature tensor to obtain a blood vessel mask, a nanoprobe mask and a preprocessed probe intensity graph; and obtaining a permeability index based on the blood vessel mask, the nano probe mask and the preprocessed probe intensity graph, and obtaining a blood vessel typing and drug recommendation design strategy by adopting a clustering and random forest regression algorithm. According to the method, the quantitative analysis precision and efficiency of the vascular permeability are remarkably improved, intelligent linkage of image analysis and a nano-drug design strategy is realized for the first time, and the method has good application potential.

The invention provides a response liposome for treating traumatic brain injury by targeting a blood-brain barrier loaded drug and a preparation method of the response liposome, and belongs to the field of traumatic brain injury treatment.The response liposome is prepared by preparing TfR-T12 target peptide firstly, then preparing nano-particles with hydrophobic molecular drugs wrapped inside, and finally preparing the TfR-T12 target peptide with the hydrophobic molecular drugs wrapped inside. And finally, preparing the response liposome for treating traumatic brain injury by the targeted blood-brain barrier loaded medicine. According to the application, the TfR-T12 target peptide which is highly expressed in cerebral cortex microvessels and can help a nano-drug to pass through a blood brain barrier through endocytosis is prepared as a target spot, meanwhile, the liposome is subjected to polypeptide modification, and the self-assembly driving force of the nano-particles is improved through process control; the outer surface of the response lipidosome is a target peptide, and the response lipidosome is internally coated with a nano-drug; the liposome can repair the blood brain barrier after injury, improve the blood flow of the brain, improve the drainage of the colloidal lymphatic system, improve the motor and cognitive functions of mice after injury, and reduce the dosage of drugs and various possible side effects caused by the drug dosage.

The present invention relates to methods for predicting the recurrence risk of hepatocellular carcinoma (HCC). Specifically, it proposes a deep learning model capable of integrating information from different phases of CT images and clinical data to predict the risk of HCC recurrence within 1 to 5 years after treatment. Experimental results demonstrate that these models outperform traditional prediction methods based on histological microvascular invasion (MVI) in predicting HCC recurrence risk.

The invention discloses a tumor heterogeneity analysis method and equipment based on super-resolution microscopic imaging, and relates to the field of medical image analysis. The method comprises the steps that a tumor imaging result is obtained, the tumor imaging result is generated through a super-resolution microscopic imaging technology, and the tumor imaging result comprises a tumor image; determining a target area of the tumor image; based on the tumor imaging result, quantitative analysis parameters are determined, and the quantitative analysis parameters are used for quantifying capillary features of at least two dimensions of tumor capillaries in the target area; on the basis of the quantitative analysis parameters, tumor heterogeneity analysis is conducted on different target areas, heterogeneity analysis results are obtained, and different target areas belong to the same tumor or different tumors. By adopting the scheme provided by the invention, the quantitative analysis parameters are generated based on the tumor imaging result, and quantitative analysis of tumor microvessel characteristics is realized, so that the accuracy of tumor heterogeneity analysis is improved.

The treatment system includes a limb sleeve containing a heating subsystem, a pressure subsystem, and a vibration subsystem. The heating subsystem provides controlled cyclic heating to the limb, leading to dilation of the microvasculature and improved blood flow to nerve receptors. The pressure subsystem provides progressively staged, cyclic pressure to the limb, thereby improving blood circulation to the limb, reducing edema buildup, and flushing waste out the lymphatic system. The vibration subsystem provides controlled vibrotactile sensation to the limb to promote neuronal stimulation and decrease pain sensation. The treatment system allows for a multi-faceted approach in disease stabilization and assistive, personalized treatment for peripheral neuropathy.

The invention discloses a vascularized skin chip, a vascularized skin model and application of the vascularized skin chip and the vascularized skin model. The material of the vascularized skin chip does not contain polydimethylsiloxane; the vascularized skin chip comprises a dermal culture chip layer and an epidermal culture chip layer from bottom to top, wherein a vertically through culture cavity is formed between the dermal culture chip layer and the epidermal culture chip layer; the dermal culture chip layer is provided with a capillary channel, and the capillary channel is used for constructing a capillary in the dermal layer. According to the vascularized skin chip, dermis and epidermis can make direct contact, interaction between cells is truly simulated, the activity and functions of the cells are kept, good mechanical support is provided for the epidermis and the dermis layer, and cell growth and tissue form maturation are promoted. The microvessel channel can simulate microvessels of a corium layer and is used for simulating supply of nutrient substances and excretion of metabolic waste, a dynamic fluid environment is provided, and the skin structure and the microenvironment are restored more truly.

The invention relates to the field of artificial intelligence medical treatment, in particular to a multi-modal feature fusion analysis system and method for gynecological face diagnosis, and aims to improve the early diagnosis precision of gynecological diseases, the system fuses multispectral images and symptom text data, and captures capillary morphology and hemodynamic parameters through a microcirculation feature extraction module; the region mapping analysis module is combined with the face region-gynecological viscera knowledge graph to generate viscera abnormality scores; the endocrine cycle modeling module constructs a dynamic model based on historical data and monitors cycle abnormity; the semantic attention fusion module fuses the symptom text and the microcirculation features by using a bidirectional cross attention mechanism to form a fusion feature vector; the knowledge reasoning feedback module is used for generating a diagnosis result and a personalized conditioning scheme based on dynamic knowledge graph multi-path reasoning, and optimizing a diagnosis and treatment path. The system overcomes the limitation of a traditional two-dimensional RGB image, effectively captures the subepidermal microcirculation state, and provides an important basis for early diagnosis of gynecological diseases.

The invention discloses application of a traditional Chinese medicine composition in preparation of a medicine for relieving side effects of oxaliplatin chemotherapy. The traditional Chinese medicine composition composed of radix paeoniae alba and liquorice is used for relieving various side reactions caused by oxaliplatin, wherein the side reactions comprise peripheral neurotoxicity, hematological toxicity and systemic reaction. The traditional Chinese medicine composition plays a role in relieving peripheral neurotoxicity caused by oxaliplatin through multiple ways, on one hand, neuroinflammatory response in dorsal root ganglions is inhibited, the dorsal root ganglions are responsible for receiving and conducting sensory nerve impulse, the injury is mainly represented as sensory disorders such as numbness of hands and feet, stabbing pain and delayed feeling, and on the other hand, the neuroinflammatory response in the dorsal root ganglions is inhibited; therefore, OIPN can be relieved by inhibiting the inflammatory reaction; on the other hand, the blood flow of microvessels is promoted, and the circulation disorder of the microvessels at the tail ends of the limbs is improved to relieve OIPN. Moreover, the traditional Chinese medicine composition does not influence the anti-tumor effect of oxaliplatin, is good in safety, and does not damage important organs such as heart, liver, spleen, lung and kidney.

The present application discloses a method, device, medium and product for determining a comprehensive brain aging index, relating to the field of image data processing. The method includes: preprocessing images such as T1 and TOF-MRA to obtain brain tissue segmentation images; post-processing TOF-MRA magnetic resonance images to extract cerebrovascular imaging features; using a method combining modules such as asymmetric convolution to perform convolution feature extraction on T1 and T2-FLAIR images, and compressing the features to generate a convolution feature group; after connecting demographic indicators, cerebrovascular imaging features and the convolution feature group, using a biological age prediction model to obtain a brain aging age prediction result, and combining with the actual age to obtain a comprehensive brain aging index. The present application can identify changes in large intracranial blood vessels, and can also reflect the state of earlier microvascular aging, improve the sensitivity to microvascular aging, solve the limitations such as low efficiency and poor consistency of manual feature extraction, and achieve comprehensive quantitative evaluation of brain aging.

Disclosed are a bio-ink for promoting microangiogenesis, a method for preparing same, and use thereof. The bio-ink is prepared by the following method: 1) preparing a filamentous microgel with a diameter within 20 microns from a sacrificial-phase material A as a porogenic template; 2) seeding vascular endothelial cells and perivascular cells on the surface of the filamentous microgel to give a hydrogel microfilament with the surface loaded with cells; and 3) mixing the hydrogel microfilament with the surface loaded with cells with a solution of a matrix-phase material B to give the bio-ink for promoting microangiogenesis. The bioprinting using the bio-ink can construct a capillary-like structure and cross-scale microvessels. The present invention can achieve the preparation of a micro-channel porous hydrogel structure with an average diameter of 20 microns or less and controllable porosity and the formation of a capillary-like network structure.

The embodiment of the invention provides a real-time processing method and system for rapid identification and filing of emergency patients. Wherein the intelligent medical fabric is worn on the skin surface of a burn patient, reflected light intensity distribution and microvascular pulsation waveforms are captured, and the light transmittance and the sampling frequency are automatically adjusted; the reflected light intensity distribution and the microvascular pulsation waveform are converted into combined parameters and matched, and a grading alarm signal is triggered when a skin tolerance dynamic interval is exceeded; packaging the graded alarm signal and the real-time physical sign data of the patient into a cross-hospital transmission data packet, and sending a data sharing request to a cooperative hospital; performing feature space alignment on the burnt skin allergy feature spectrum to generate a cross-mechanism allergy reaction prediction model; and automatically reconstructing an allergy monitoring file of the burnt patient during detection and transfer. According to the technical scheme provided by the embodiment of the invention, intelligent early warning and dynamic filing of the allergic reaction of the burn patient are realized, and the recognition accuracy and filing efficiency of the emergency patient in an emergency scene are remarkably improved.

The invention discloses a cardiovascular and cerebrovascular disease information management method and system, and relates to the technical field of genomics, and the method comprises the steps: collecting health information and a biological sample of a patient, and generating a health file; performing genomics analysis by using the generated health archive, and calculating a genetic risk score of the patient; according to the genetic risk score of the patient, microvascular lesion detection is carried out on the high-risk patient, microvascular anomaly indexes are quantified, and an individualized health risk score is generated; metabonomics analysis is carried out based on the health risk score, and the metabolic state and metabolic capability of the patient are evaluated; combining the metabolism assessment result, the genetic risk score and the health risk score to formulate an individualized intervention scheme; the problems of data isolation and intervention generalization in the prior art are solved, and powerful technical support is provided for early prevention and accurate management of cardiovascular and cerebrovascular diseases.

The present invention relates to a method for predicting the risk of recurrence of hepatocellular carcinoma (HCC). Specifically, the present invention proposes a deep learning model capable of integrating information of a multi-phase CT image and clinical data to predict an HCC recurrence risk within 1-5 years after treatment. Experimental results show that the models are superior to a traditional histological microvascular invasion (MVI)-based prediction method in the aspect of predicting the HCC recurrence risk.

The invention discloses a construction method of a deep learning model for predicting hepatocellular carcinoma microvascular invasion through preoperative magnetic resonance and a probability prediction method, and belongs to the technical field of image processing. Collecting multi-sequence medical images of a plurality of hepatocellular carcinoma patients, and preprocessing the multi-sequence medical images; based on the preprocessed multi-sequence medical image, using a Resnet 18 architecture to construct a deep learning model for predicting hepatocellular carcinoma capillary invasion by preoperative magnetic resonance; a gradient weighted activation mapping Grad-CAM visualization method is applied to a preoperative magnetic resonance prediction hepatocellular carcinoma microvascular invasion deep learning model. The prediction performance of the fusion model integrating the multi-modal information is superior to that of a single-sequence model and a multi-sequence model; the fusion model based on the mask and the bounding box has no significant statistical difference on the prediction performance of the MVI, and the use of the bounding box can reduce manual operation steps while maintaining the accuracy of clinical application.

The invention belongs to the technical field of experiments and the field of anatomical research, and particularly relates to a self-developing vascular casting agent and a preparation method thereof. The invention discloses a preparation method of a self-developing vascular casting agent. The self-developing vascular casting agent is obtained by uniformly mixing iopromide, natural latex, II-type barium sulfate and pigment, wherein in parts by weight, the iopromide accounts for 30-40%; 25%-30% of natural latex; 40% of type II barium sulfate; and 5%-10% of pigment. The blood vessel casting agent developed by the invention is excellent in formula flowability (smooth perfusion), does not cause blockage, can avoid reverse osmosis of a capillary microenvironment, and realizes arteriovenous differentiated perfusion; and meanwhile, high-contrast development is realized, and complete coverage of capillary vessels is basically realized. The optimized blood vessel casting agent formula is completely cured (at room temperature) within 24 hours, the shrinkage rate is low, and blood vessel deformation is avoided.