54 results about "Fracture type" patented technology

The invention discloses a tight sandstone gas reservoir logging scale fracture development evaluation method, and relates to the field of earth exploration physics, and the method comprises the steps: carrying out the inversion based on a complex pore structure high-angle fracture type rock physical model and actual logging data through employing a simulated annealing particle swarm algorithm, and obtaining two groups of fracture density parameters in a well; respectively calculating crack density identification factors under different rotation angles; based on a correlation coefficient between the Poisson impedance and the fracture density under different rotation angles, determining an optimal rotation angle when the correlation coefficient reaches a maximum value; calculating a crack density identification factor under the optimal rotation angle by combining longitudinal wave impedance and transverse wave impedance data; and applying the fracture density identification factor under the optimal rotation angle to actual logging data, and representing fracture development. The method can effectively distinguish and quantify multiple groups of high-angle fractures developed in the tight sandstone reservoir, significantly improve the prediction precision and reliability of the fractured reservoir, and provide a key basis for reservoir sweet spot prediction and well location deployment.

The invention discloses a three-section thighbone retrograde intramedullary nail, relates to the field of medical instruments, and provides a detachable and assembled three-section thighbone retrograde intramedullary nail aiming at the problems that an existing thighbone medial condyle retrograde intramedullary nail is poor in adaptability and insufficient in fixing strength. A main nail body of the intramedullary nail is composed of a head section, a middle section and a tail section, original spiral nested structures are adopted at the joints of the sections, and the three sections are tightly connected to form the complete and high-strength intramedullary nail by matching with the hollow interior design and penetrating a titanium cable into a medullary cavity for tightening and automatically rotating and matching the head section and the middle section of the main nail body. And medullary cavity structures and fracture types of different patients can be accurately matched. Compared with a traditional intramedullary nail, the intramedullary nail has the advantages that mounting convenience and structural stability are remarkably improved, intraoperative adjustment time can be shortened, the risk of injury to a medullary cavity of a patient is reduced, and an innovative solution is provided for personalized treatment of distal femur fractures.

A fracture reduction mechanism for minimally invasive pelvic fracture surgery, comprising a support assembly, a drive assembly, a self-rotating clamping assembly, and a healthy side fixation assembly, wherein the drive assembly comprises a synchronous rotation module (21), a first ball screw module (22), a second ball screw module (23), a third ball screw module (24), and an arc guide rail module (33); the mechanism forms a frame-like closed structure through connection with an operating table (1) via bedside rails (2) on both sides of the operating table (1), the support assembly and healthy side fixation assembly are slidably connected with the bedside rails (2). By manually adjusting the linear movement distances and fixed-axis rotation angles of the above mechanism, six degrees of freedom translation and rotation of the pelvic fracture block can be achieved. The mechanism features high rigidity and load-bearing capacity, good portability, simple assembly and operation, high motion precision, and its motion center is a virtual point in space, the position of which can be adjusted according to patient position and fracture type, providing convenience for surgery.

The invention relates to a method for predicting fracture development laws of carbonate rocks with different lithologies, which comprises the following steps of: performing lithology division on a target layer of a research area, and analyzing fracture development characteristics with different lithologies; simulating fracture development conditions of different lithology under different confining pressures in different burial depth environments of the research area through a rock compression experiment to obtain fracture development laws of different lithology; identifying the lithology of each single well in the research area by using a well logging curve under core calibration to obtain a lithology interpretation result of each single well; the fracture development rules of different lithology are matched with the lithology interpretation result of each single well, and the fracture development condition of each single well target layer in the research area is obtained; and predicting the fracture development condition of the target stratum of the research area by utilizing a three-dimensional geologic modeling method according to the interpretation results of different lithology of the target stratum of each single well and the fracture development condition. According to the method, multiple means of actual petrology observation, indoor rock core experiments, well logging interpretation and three-dimensional modeling are combined, the development laws of carbonate rock fractures of different lithology on the ground and underground can be accurately predicted, a fractured reservoir development area can be found easily, and a reliable basis is provided for exploration deployment.

The invention provides a preoperative incision positioning device for spine minimally invasive surgery aiming at different fracture types, and belongs to the technical field of medical equipment. Comprising a mounting box, a notched gear ring is slidably connected into the mounting box, an adjusting mechanism used for adjusting the position of the notched gear ring is arranged in the mounting box, the two ends of the notched gear ring are fixedly connected with mounting bases respectively, and the lower end face of the mounting base on the upper side is fixedly connected with a detector; through the detector, the X-ray source and the positioning mechanism, real-time imaging of the spine of a patient is achieved, laser indication can be automatically conducted on an incision positioning point, medical staff can accurately position and mark the incision position, accurate positioning of the spine minimally invasive surgery incision is achieved, the positioning error is reduced, and the accuracy of the spine minimally invasive surgery incision is improved. The success rate and safety of the operation are improved, and the problems of poor positioning efficiency and precision in the prior art are solved.

The invention relates to the technical field of oil and gas field development, in particular to a shale gas well fracturing parameter optimization method and device based on pump stop pressure drop, and the method comprises the steps that fracturing construction parameters of a fracturing section are obtained; obtaining fracture network interpretation parameters and SRV parameters; obtaining production logging monitoring data of the fracturing section; according to the crack type of the fracturing section, the fracturing section is calibrated as a matrix section, a micro crack section and a bending section; the method comprises the following steps: respectively calling parameters for fracturing sections classified into a matrix section, a microfracture section and a bending section, forming an arrangement of two fracturing sections according to production logging monitoring data in the parameters and the numerical value of an SRV parameter, and screening out a plurality of fracturing sections located in a specified sorting interval in the two arrangements, and fracturing construction parameters are respectively used as preferable parameters of a matrix section, a microfracture section or a bending section. And subsequently, in a shale gas well fracturing section under the same geological background condition, fracturing parameter design can be carried out by referring to the optimal fracturing parameter range.

The application discloses a three-axis motion parallel reset mechanism for closed reset of a pelvic fracture, which comprises a sick side pelvic fracture reset robot and a healthy side fixing device, and the sick side pelvic fracture reset robot comprises a movable base, a three-axis moving device, a three-axis rotating device, a reset force detecting device and a sick side screw holding device. Through the above structure, six-degree-of-freedom translation and rotation of a fracture block can be realized, and the reset force change can be monitored in real time, thereby providing a guarantee for operation safety. The three-axis rotating device is connected in series through a U-shaped frame, has a smaller space occupation than a traditional ball pair, can be close to the upper surface of a human body during reset, has a short force arm and can provide a larger reset force. The robot has the advantages of compact structure, high precision, large load force, high flexibility, large working space, a spatial virtual point as a motion center, adjustable position according to a patient and a fracture type and convenient operation.

The invention discloses a fracturing crack type discrimination method and related equipment, and belongs to the technical field of oil exploitation, and the method comprises the steps: firstly, building a crack extension mode judgment model and a natural crack approximation angle range model according to ground stress parameters, rock strength parameters, imaging logging parameters and extension pressure judgment; according to the fracture extension mode judgment model, the fracture extension mode is judged, and then the fracturing fracture type is judged; when the fracture extension mode is judged to be a natural fracture activation mode, the type of a fracturing fracture is judged according to a natural fracture approximation angle range model; according to the method, an accurate crack extension mode and a natural crack approximation angle range model are constructed by integrating multi-dimensional parameters such as ground stress, rock strength, imaging logging and extension pressure, and the judgment precision of the fracturing crack type is effectively improved; the crack type is further refined by using the approximation angle range model, a more scientific and reasonable fracturing design basis is provided for oil and gas field development, the yield increasing effect is optimized, and the development cost is reduced.

The invention discloses a method and system for determining the fracture evolution process of hard rock containing microfractures, and belongs to the technical field of rock mechanical testing. Digital speckle images of a DIC test observation surface of a hard rock sample in the loading process of an indoor mechanical test are collected in real time by determining the DIC test observation surface of the hard rock sample; by determining a surface displacement field of the hard rock sample, a displacement vector diagram of microscopic particle points of the digital speckle image is obtained, and the type of new cracks is judged; a plurality of displacement observation points are calibrated on the two sides of the native microfracture, and the fracture types of the displacement observation points are determined by judging whether tangential displacement and normal displacement of the displacement observation points are separated or not; and according to the type of the new crack and the fracture type of the displacement observation point, determining the dynamic evolution process of the native micro-crack under the stress condition. According to the method, the dynamic evolution process of cracks from initiation to expansion can be accurately represented, and then the internal correlation between the spatial geometric characteristics of crack distribution and the hard rock fracture evolution mechanism is disclosed.

This application provides a modular orthopedic device for fracture treatment, belonging to the field of orthopedic device technology. The device includes a fixation component, a locking component, and a connecting component. The fixation component has at least one first slot and at least one second slot, which communicate with each other. The locking component is accommodated within the first slot, and the connecting component is accommodated within the second slot and cooperates with the fixation component, allowing it to move or rotate relative to the fixation component. The locking component is used to lock the connecting component and the fixation component together. The connecting component can be a round rod, a polygonal rod, or a toothed connector, and the fixation component can include a toothed fixator, a ball-and-socket fixator, a ball-and-socket fixator, or a universal ball-and-socket fixator. This modular orthopedic device allows for flexible selection of component modules based on fracture type and soft tissue conditions, achieving precise, personalized, and minimally invasive fracture treatment, filling the technological gap between bone plate systems, intramedullary nail systems, and external fixation systems.

The invention provides a frequency-varying azimuth anisotropy seismic response characteristic analysis method for a complex fractured reservoir, and belongs to the technical field of oil and gas development, and the method comprises the steps: constructing an equivalent medium model which comprises a plurality of groups of vertical-vertical orthogonal fractures and is partially saturated with fluid in the fractures based on a Chapman multi-scale fractured medium theory; based on the model, frequency variation and azimuth anisotropy stiffness matrixes are deduced, the frequency variation azimuth stiffness matrix is embedded into the anisotropy reflectivity theory, and the PP wave frequency variation azimuth seismic forward modeling method is provided. And finally, respectively designing a single interface and thin layer reflection model, analyzing fracture geometric parameters (density, radius and porosity), partial saturated fluid parameters (saturation and plaque parameters), a PP wave frequency change reflection coefficient under a thin layer thickness change condition and a seismic azimuth response change rule thereof, and revealing control factors of frequency change azimuth response of the fractured reservoir. And a revolutionary numerical simulation method is provided for sweet spot prediction and fluid identification of the complex fractured reservoir.

The invention discloses an angle-adjustable orthopedic fracture fixing device, and belongs to the technical field of medical instruments, the orthopedic fracture fixing device comprises a telescopic rod, elastic clamping plates arranged at the two ends of the telescopic rod respectively, and universal structures arranged at the two ends of the telescopic rod respectively and connected with the correspondingly arranged elastic clamping plates; the connecting structure comprises a fixing plate arranged on the universal structure, a clamping plate arranged on the face, away from the universal structure, of the fixing plate and a locking bolt arranged on the fixing plate. According to the angle-adjustable orthopedic fracture fixing device, flexible switching between external fixation of the elastic clamping plate and internal fixation of the kirschner wire can be achieved through the connecting structure, different fracture types can be adapted without changing the whole device, the medical cost and the clinical operation difficulty are reduced, and the angle-adjustable orthopedic fracture fixing device is convenient to use through independent rotation locking of the elastic clamping plate and cooperation of the universal structure. The limb physiological angle and the fracture reduction requirement can be precisely met, and the use flexibility and the fixing precision of the device are further improved.

The invention discloses a novel intelligent steel plate screw internal fixing device and method.The novel intelligent steel plate screw internal fixing device comprises connecting steel plates, the connecting steel plates comprise straight plates and different plates, the connecting steel plates are provided with at least two screw holes in the length direction, concave-convex textures are arranged in the circumferential directions of the screw holes, and the two connecting steel plates are matched with each other through the concave-convex textures to limit relative rotation; the connecting screws comprise locking screws and lag screws, connecting holes are formed between adjacent screw holes, connecting parts are arranged on the supporting pieces and used for connecting the connecting holes, the supporting pieces are made of degradable materials, the modularized connecting steel plates are matched with the connecting screws to achieve an internal fixing system, and splicing type dynamic fine adjustment of the connecting steel plates and the connecting screws is achieved. The internal fixation shape and size can be flexibly adjusted according to different actual fracture conditions, so that the internal fixation device is suitable for various fracture types, especially complex or irregular fractures, and more accurate fixation force is provided for fracture parts.

The utility model provides a multi-row nail hole locking type metal bone fracture plate, which relates to the technical field of medical instruments and comprises a bone fracture plate a, a bone fracture plate b is symmetrically arranged on one side of the bone fracture plate a, an extension plate is fixedly connected to one end, close to the bone fracture plate b, of the bone fracture plate a, and a storage groove matched with the extension plate for use is formed in one side of the bone fracture plate b. The top and the bottom of the bone fracture plate a and the top and the bottom of the bone fracture plate b are each provided with two bone holding claws. The bone fracture plate a is pulled to drive the extension plate to move out of the storage groove, and then the bone fracture plate a and the bone fracture plate b are extended, so that the length of the bone fracture plate can be adjusted to better adapt to individual differences according to different bone sizes, fracture types and operation requirements of patients, and bolts are screwed out to relieve fixation of the sliding blocks; the bone holding claws are moved to drive the sliding blocks to move in the sliding grooves, so that the distance between the rib bone holding claws is adjusted according to the actual fracture condition, the sliding blocks are prevented from falling off from the sliding grooves through the arrangement of the guide blocks and the guide grooves, and the sliding blocks can move horizontally.

The present application relates to a kind of fracture type buried hill reservoir skin coefficient system decomposition method, comprising: step A.Combined with the calculation method of various types of dynamic and static skin coefficient of fracture type buried hill reservoir characteristics constitutes total skin;Step B.Based on the test data of test well, total skin coefficient of test well is obtained using pressure recovery well test interpretation method;Step C.According to the well condition of test well, the main dynamic and static skin coefficient constituting total skin coefficient of test well is determined;Step D.The method of step A is used to calculate the skin coefficient type determined in step C, to realize the skin decomposition of test well, and to clarify the composition of various skin coefficients of test well.The present application can accurately obtain the size of various skin coefficients of test well in fracture type buried hill reservoir, and provide guidance for further stimulation operation.

The invention relates to a method and a device for calculating a mass exchange coefficient between bedrock and a fracture in an underground reservoir, and belongs to the technical field of seepage simulation of fracture type underground reservoirs. The method for calculating the mass exchange coefficient between the bedrock and the fracture in the underground reservoir comprises the following steps: performing embedded grid division on a bedrock region and a fracture region in the underground reservoir, and determining a connection pair between a fracture grid and a bedrock grid; based on the permeability of the bedrock grid corresponding to the connection pair, determining a Green element function under a bedrock grid permeability main axis coordinate system; source item coordinates in the Green element function are used as variables, integration is conducted on the Green element function along the fracture grids belonging to the same connection pair, and a pressure response function is determined; and based on the pressure response function, determining the conductivity between the bedrock grid and the fracture grid corresponding to the connection pair. According to the method, the calculation precision of mass exchange between the bedrock and the fracture in seepage simulation of the embedded discrete fracture model can be remarkably improved.

The invention discloses a fracture development orientation prediction 90-degree ambiguity weakening method, electronic equipment, a storage medium and a fracture development orientation prediction 90-degree ambiguity weakening device. The method comprises the following steps: determining influence factors causing 90-degree fuzziness of fracture development orientation prediction based on an orientation elastic wave impedance equation; analyzing the influence factors based on a fracture orientation interpretation result of logging data or geological data and a fracture-induced orientation anisotropy model; and determining a strategy for weakening the 90-degree fuzziness of fracture development orientation prediction based on an analysis result. According to the method, starting from a root cause for generating fracture development orientation to predict 90-degree fuzziness, a mapping relation between a fracture orientation prediction result and fracture type reservoir geophysical parameters and combinations thereof is researched through model testing, and a strategy for weakening fracture development orientation to predict 90-degree fuzziness is determined through the mapping relation. Fracture orientation prediction with 90-degree ambiguity weakened is achieved, and reliability of fracture orientation seismic description is improved.

The invention relates to a formation fracturing type and fracturing performance analysis method and device and a storage medium, and relates to the technical field of oil exploration and development, and the method comprises the following steps: obtaining detection data of a target formation; calculating the hole friction according to the detection data; the theoretical critical fracture stress is calculated through the internal friction angle and cohesive force of the target stratum; calculating a theoretical critical fracture stress by using the pore pressure and the hole friction resistance of the target stratum; the stratum fracturing type is judged according to the theoretical critical fracture stress; according to the judged formation fracturing type, determining the maximum critical fracturing stress and the minimum critical fracturing stress, corresponding to the formation fracturing type, in a block where the target formation is located; and calculating fracturing evaluation parameters of the target stratum by using the maximum critical fracture stress, the minimum critical fracture stress and the critical fracture stress. Through the content of the invention, the fracture type can be determined according to the stratum condition, the stratum fracturing performance can be accurately predicted, the exploration and development cost is reduced, and the stratum development efficiency is improved.

The invention discloses a thoracolumbar vertebra fracture detection method and system based on sequence modeling and a prior graph, and the method comprises the steps: obtaining thoracolumbar vertebra CT volume data and fracture type truth value label data, carrying out the preprocessing, and constructing a training set and a test set; and constructing a fracture detection multi-branch classification network model, training and testing the constructed fracture detection multi-branch classification network model by using the training set and the test set, preprocessing thoracolumbar vertebra CT body data of a to-be-detected patient, inputting the preprocessed thoracolumbar vertebra CT body data into the trained fracture detection multi-branch classification network model, and outputting a fracture type classification prediction result. By designing an end-to-end framework of an image feature branch, a label relation branch and a branch fusion module, the problems that in the prior art, continuity of a vertebral body structure is deficient, long-range dependence is insufficient, feature fusion is insufficient and the like are solved, and accurate detection and refined classification of thoracolumbar vertebral fractures are achieved.

The invention relates to the technical field of orthopedic trauma operation internal fixation, and discloses an orthopedic reduction nail locking device which comprises a first steel plate and a second steel plate which are symmetrically arranged. A plurality of groups of local threaded holes are formed in the first steel plate and the second steel plate, and screws are arranged on the local threaded holes; the orthopedic reduction screw locking device has the remarkable advantages that screws are accurately matched with local threaded holes, positioning is achieved through the conical bottom, screw deviation is eradicated, and pressurization failure and harmful stress are avoided; the independent pressurization moving assembly is matched with the L-shaped limiting plate, pressurization force transmission is uniform and controllable, and it is guaranteed that the fracture broken end is accurately attached and stably fixed; local threaded holes can be vertically or obliquely designed to adapt to different fracture types, the hexagonal groove design enables operation to be convenient and labor-saving, operation difficulty is greatly reduced, and a safe and efficient fixing scheme is provided for orthopedic trauma operations.

The invention discloses a fractured carbonate rock heat storage area selection evaluation method and a fractured carbonate rock heat storage area selection evaluation system. The heat storage selected area evaluation method comprises the following steps: predicting fractured carbonate rock heat storage key factors of a target area to obtain parameter data of each heat storage key factor; the heat storage key factors comprise heat storage burial depth, fracture development, water-rich characteristics and temperature distribution; determining an optimal evaluation area based on the parameter data of the heat storage burial depth; performing quantitative processing on parameter data of fracture development, water-rich characteristics and temperature distribution of the optimal evaluation area, and setting weight coefficients of the fracture development, the water-rich characteristics, the temperature distribution and the parameters; and based on a quantitative processing result, a weight coefficient and an evaluation standard, carrying out crack type carbonate rock heat storage comprehensive evaluation on the optimal evaluation region. The method can rapidly, effectively and accurately complete heat storage selection area evaluation, improves evaluation efficiency, improves the success rate of geothermal well drilling, and has the characteristics of prominent key parameters, simple calculation, convenience and rapidness.

The application provides a volcanic rock matrix-fracture double medium type oil reservoir evaluation method and system, and belongs to the technical field of geological engineering, oil reservoir technology and well logging interpretation.The application establishes a matrix porosity calculation chart by core calibration logging, adds a fracture pore as a conductive channel in an Archie sandstone conductive model, deduces fracture porosity, and finally combines the matrix pore, the fracture pore, the oil saturation and the oil test result to form a matrix-fracture type double medium oil and gas reservoir quantitative identification method.The application fills the gap of the research on the matrix and the fracture at the same time, improves the oil reservoir identification precision, and is more strict and reasonable in the identification standard, and is high in reliability.

The invention relates to a design method of a proximal fibula intramedullary nail based on computer simulation. Comprising the following steps: constructing a three-dimensional digital model of near-end and surrounding tissues of fibula based on CT scanning data of a patient, and accurately marking common peroneal nerve walking, skeleton morphological characteristics and fracture types; simulating stress distribution data of the fracture part by using a finite element analysis method; in combination with the three-dimensional model and the stress distribution data, a personalized intramedullary nail model matched with the proximal curvature, length and internal structure of the individual fibula is generated through parameterized design; and planning a minimally invasive implantation path of the intramedullary nail in a computer simulation environment, and generating matched implantation path data to guide an operation. Compared with the prior art, three-dimensional anatomical modeling, finite element analysis and parametric design technologies are creatively integrated, a complete personalized intramedullary nail design system is established, and the matching precision, fixing stability and operation safety of instruments and bones are effectively improved.

A method for auxiliarily planning a fracture reduction path based on an augmented reality technology includes: obtaining fracture image data for a fracture area of a patient, the fracture image data including computer tomography (CT) data and X-ray data; reconstructing a three-dimensional model of a fracture part based on the fracture image data; matching and calibrating the three-dimensional model of the fracture part based on the fracture area of the patient in an actual scene; inputting a calibrated three-dimensional model of the fracture part and a corresponding fracture type and fracture surgery requirement into a fracture reduction path planning model to determine at least one corresponding optimized fracture reduction path, where the fracture reduction path planning model includes a cascaded reinforcement learning (RL) module and a multi-physics field simulation optimization module; and rendering each fracture reduction path, and determining a target fracture reduction path according to a detected interactive instruction.

The invention provides a characterization method for development degrees of natural fractures of different scales, relates to the field of oil exploitation, and aims to comprehensively and accurately characterize the development degrees of the natural fractures of different scales, especially to solve the problem of small and micro fractures in shale oil reservoirs. The method is realized through the following steps of: firstly, dividing crack types based on core observation in combination with a focused ion beam-scanning electron microscope technology; secondly, analyzing response characteristics of the cracks in electric imaging logging, array acoustic logging and conventional logging; thirdly, constructing a crack identification sample library containing the features; and finally, constructing an intelligent crack evaluation model by applying a grey correlation algorithm to obtain a crack development index. According to the method, large-scale structural cause cracks and tiny non-structural cause cracks are distinguished, crack classification can be refined, the recognition capacity is complemented through multiple well logging technologies, and the recognition precision is improved. Compared with the prior art, the method has the advantages that the accuracy of crack identification is greatly enhanced, and reservoir assessment and fracturing design are optimized.

This invention provides a non-equilibrium grouting method based on the multi-dimensional characteristics of tunnel rock mass. By constructing a technical system that integrates multi-dimensional geological feature identification and model-based decision-making, it achieves precise and intelligent grouting control for complex fractured rock masses in tunnels. First, fractures are comprehensively classified based on their width, connectivity, and filling state, forming spatial distribution data. This transforms geological conditions into structured, spatially locatable digital information, providing precise target objects for grouting construction. Then, through a grouting parameter decision model, the current fracture type is automatically mapped to a strictly matching combination of grout mix ratio, grouting pressure, and grouting rate, achieving intelligent and standardized decision-making from geological features to construction parameters. This process solidifies mature experience and theoretical understanding into repeatable rules, ensuring a high degree of compatibility between the initial construction plan and the inherent hydraulic characteristics of the fractures, thereby significantly improving the targeting and controllability of grouting construction.

This application discloses a method and system for predicting fold-related fractures based on the principle of stress superposition, belonging to the field of geological analysis technology. The method includes: acquiring past 3D seismic data and past regional stress tensors of a region, and constructing a high-precision structural grid cell on the top surface of the target reservoir; based on the high-precision structural grid cell, determining the fracture type and fracture orientation according to the sliding window-least squares surface fitting joint method, the principle of stress superposition, and rock fracture criteria, thus obtaining a fracture prediction model; acquiring the current 3D seismic data and current regional stress tensor of the target region, and inputting them into the fracture prediction model to obtain the current fracture type and current fracture orientation. This application achieves accurate prediction of the type and orientation of fold-related fractures.

The application relates to a proximal humerus fracture internal support repair system, which comprises a humeral head support piece, a humeral neck support piece, a humeral shaft support piece and a fixing piece, and realizes reduction fixation of proximal humerus fracture blocks and structural support of a humeral head-humeral shaft. Through the above scheme, the proximal humerus fracture internal support repair system can effectively treat proximal humeral head fractures, so as to realize better reduction and support effects and greater fixation strength, and the repair effect is better, especially for complex proximal humeral head fracture types, such as comminuted fractures, shortening deformities and the like.

The invention discloses a fracture X-ray film AI auxiliary diagnosis system. The system adopts a deep learning architecture based on a convolutional neural network, and comprises an image preprocessing module, a deep learning recognition module, a diagnosis result output module and a user interaction interface module. The image preprocessing module carries out de-noising processing by combining adaptive median filtering with a bilateral filtering technology, and an improved Canny edge detector is combined with a deep learning edge enhancement network to identify skeleton features. The deep learning recognition module constructs a feature extraction network based on a ResNet-50 pre-training model, and adopts a weighted cross entropy loss function and an Adam optimizer for training. And the diagnosis result output module generates a visual fracture line label by applying a gradient weighted class activation mapping technology, and automatically generates a structured report containing fracture types, severity evaluation and diagnosis and treatment suggestions. The system supports DICOM and JPG format image input, the diagnosis accuracy reaches 78%-80%, and the fracture diagnosis efficiency and accuracy are remarkably improved.