Anesthesia puncture visualization navigation system combining augmented reality and three-dimensional reconstruction

Abstract

The anesthesia puncture visual navigation system combining augmented reality with three-dimensional reconstruction belongs to the technical field of anesthesia puncture visual navigation, and comprises a physiological motion monitoring unit, which is used for simultaneously collecting respiratory motion cycle data and soft tissue elastic deformation information, and generating a comprehensive motion state data packet by calculating the correlation coefficient of the respiratory amplitude and the tissue elastic modulus change rate, so that the spatial deviation caused by the respiratory motion and the soft tissue deformation of the patient to the puncture navigation can be effectively compensated, the puncture precision is significantly improved compared with the traditional static navigation system under the normal respiratory state of the patient, the puncture failure rate and the complication risk are effectively reduced, the problem of further precision deterioration caused by the blind application of correction parameters of the traditional system when the registration quality deteriorates is avoided, the dependence on the personal experience of the doctor is reduced, and the problem of complete system failure caused by a single correction mode is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of anesthetic puncture visualization navigation, in particular to an anesthetic puncture visualization navigation system combining augmented reality and three-dimensional reconstruction. BACKGROUND

[0002] Traditional anesthetic puncture navigation technology mainly relies on static image data and fixed anatomical reference points for path planning and navigation. However, in actual clinical application, the following technical defects exist:

[0003] During the anesthetic puncture process, the patient will inevitably produce respiratory motion, heartbeat and slight body position changes. These physiological motions will cause real-time changes in the spatial position of the target anatomical structure. The traditional navigation system cannot effectively identify and compensate for these dynamic changes, resulting in a significant decrease in puncture accuracy and an increase in the risk of puncture failure and complications.

[0004] The prior art lacks effective processing capability for soft tissue elastic deformation. When the puncture needle contacts and compresses the soft tissue, the tissue will undergo elastic deformation to varying degrees, changing the original anatomical structure relationship. The traditional system cannot real-time perceive and compensate for this deformation, causing deviations in the navigation information and the actual anatomical position. The current registration technology lacks a dynamic quality evaluation mechanism. In the environment of patient motion interference, the quality of image registration will change with time and motion amplitude. The existing system cannot real-time evaluate the reliability of registration quality, nor can it dynamically adjust the correction strategy according to the quality changes, resulting in maintaining the original correction intensity when the registration quality decreases, further reducing the navigation accuracy.

[0005] The above information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The purpose of the present application is to provide an anesthetic puncture visualization navigation system combining augmented reality and three-dimensional reconstruction to solve the problems raised in the above background.

[0007] The technical solution of the present application is: comprising: a physiological motion monitoring unit for simultaneously collecting respiratory motion cycle data and soft tissue elastic deformation information, generating a comprehensive motion state data package by calculating the correlation coefficient of respiratory amplitude and tissue elastic modulus change rate;

[0008] A three-dimensional reconstruction processing unit is used to construct a three-dimensional anatomical model of the patient based on multi-modal medical image data, receive the comprehensive motion state data package, and dynamically adjust the three-dimensional anatomical model according to the motion parameters to generate a dynamic anatomical reference model;

[0009] The spatial position correction unit is configured to receive the integrated motion state data packet and the dynamic anatomical reference model, predict a target anatomical structure position according to the respiratory motion cycle data, evaluate a puncture needle position offset according to soft tissue elastic deformation information, and generate a correction parameter by weighted fusion of the predicted position and the offset;

[0010] The multi-modal registration quality evaluation unit is configured to receive the real-time medical image data, the dynamic anatomical reference model and the integrated motion state data packet, calculate a spatial position entropy, a texture feature entropy and a motion consistency entropy, and generate an integrated quality evaluation index by weighted fusion of the three types of entropy values.

[0011] The dynamic compensation adjustment unit is configured to receive the integrated quality evaluation index, determine a correction intensity level according to the index value, and generate a correction intensity control parameter output to the spatial position correction unit.

[0012] The augmented reality navigation unit is configured to receive the correction parameter adjusted by the correction intensity control parameter, superimpose navigation information on the real-time medical image, and perform hierarchical visual coding according to the correction confidence.

[0013] Preferably, the integrated motion state data packet includes a timestamp, a three-dimensional motion amplitude vector, a tissue elasticity change amount and a motion deformation correlation coefficient.

[0014] Preferably, the spatial position correction unit receives a correction intensity control parameter from the dynamic compensation adjustment unit to adjust the application proportion of the correction parameter, and generates a final correction parameter output to the augmented reality navigation unit.

[0015] Preferably, the dynamic compensation adjustment unit receives navigation effect evaluation data from the augmented reality navigation unit to optimize the mapping relationship between the quality index and the correction intensity.

[0016] Preferably, the physiological motion monitoring unit includes a strain-type respiration sensor and a contact-type ultrasonic transducer. The strain-type respiration sensor is arranged on the chest and abdomen of the patient to monitor respiratory motion parameters, and the contact-type ultrasonic transducer is used to detect the change of the elastic modulus of soft tissue in the puncture area.

[0017] Preferably, the weight of the spatial position entropy in the weighted fusion is 40%, the weight of the texture feature entropy is 35%, and the weight of the motion consistency entropy is 25%.

[0018] Preferably, the correction intensity level includes a high intensity level, a medium intensity level and a low intensity level. When the integrated quality evaluation index is higher than 0.8, the high intensity level is determined; when the index is between 0.5 and 0.8, the medium intensity level is determined; and when the index is lower than 0.5, the low intensity level is determined.

[0019] Preferably, the high strength level corresponds to 100% application ratio, the medium strength level corresponds to 60% application ratio, and the low strength level corresponds to 30% application ratio.

[0020] Preferably, the hierarchical visual coding includes color coding and transparency coding, high correction reliability corresponds to green solid line display, medium correction reliability corresponds to yellow dashed line display, and low correction reliability corresponds to red semi-transparent display.

[0021] The present application provides a anesthesia puncture visual navigation system combining augmented reality and three-dimensional reconstruction by improvement, compared with the prior art, has the following improvements and advantages:

[0022] By establishing a physiological motion adaptive dynamic compensation mechanism, the present application can effectively compensate for the spatial deviation caused by patient respiratory motion and soft tissue deformation to puncture navigation, and in the normal breathing state of the patient, the puncture accuracy is significantly improved compared with the traditional static navigation system, effectively reducing the puncture failure rate and complication risk.

[0023] The innovative introduction of a registration quality evaluation mechanism based on multi-modal information entropy realizes real-time quantitative monitoring of the reliability of the navigation system, can timely issue a warning when the registration quality decreases and automatically adjust the correction strategy, avoiding the further deterioration of accuracy caused by the blind application of correction parameters in the traditional system when the registration quality deteriorates.

[0024] A complete adaptive learning loop from physiological motion perception to quality evaluation to dynamic correction is established, which can automatically optimize the correction strategy according to the physiological characteristics and motion patterns of different patients, improve the navigation accuracy for specific patient groups, and significantly reduce the dependence on the personal experience of doctors.

[0025] A hierarchical visual navigation interface based on confidence is provided through augmented reality technology, doctors can intuitively judge the reliability of the navigation information and make corresponding decisions, and this intelligent human-computer interaction design greatly improves the safety and efficiency of puncture operation, and reduces the average puncture time.

[0026] The multiple-coupled adaptive mechanism ensures the stability of the system when facing various sudden motion disturbances, even in extreme cases such as severe coughing or sudden changes in body position, the system can still maintain basic navigation function through dynamic adjustment of correction strength, avoiding the complete failure of the system caused by a single correction mode. BRIEF DESCRIPTION OF DRAWINGS

[0027] The present application will be further explained in conjunction with the accompanying drawings and examples:

[0028] Figure 1 is a flowchart of the system of the present application. DETAILED DESCRIPTION

[0029] In order to make the objects, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with specific embodiments. Embodiment

[0030] Please refer to Figure 1 The present application provides a technical scheme of anesthesia puncture visualization navigation system combining augmented reality and three-dimensional reconstruction, comprising: a physiological motion monitoring unit for simultaneously collecting respiratory motion cycle data and soft tissue elastic deformation information, generating a comprehensive motion state data package by calculating the correlation coefficient of respiratory amplitude and tissue elastic modulus change rate;

[0031] A three-dimensional reconstruction processing unit is used to construct a three-dimensional anatomical model of a patient based on multi-modal medical image data, receive the comprehensive motion state data package, and generate a dynamic anatomical reference model by dynamically adjusting the three-dimensional anatomical model according to the motion parameters;

[0032] A spatial position correction unit is used to receive the comprehensive motion state data package and the dynamic anatomical reference model, predict the position of the target anatomical structure according to the respiratory motion cycle data, evaluate the position offset of the puncture needle according to the soft tissue elastic deformation information, and generate a correction parameter by weighted fusion of the predicted position and the offset;

[0033] A multi-modal registration quality evaluation unit is used to receive real-time medical image data, a dynamic anatomical reference model and a comprehensive motion state data package, calculate spatial position entropy, texture feature entropy and motion consistency entropy, and generate a comprehensive quality evaluation index by weighted fusion of the three types of entropy values;

[0034] A dynamic compensation adjustment unit is used to receive the comprehensive quality evaluation index, determine the correction intensity level according to the index value, and generate a correction intensity control parameter output to the spatial position correction unit;

[0035] An augmented reality navigation unit is used to receive the correction parameter adjusted by the corrected intensity control parameter, superimpose the navigation information on the real-time medical image, and perform hierarchical visualization coding according to the correction reliability;

[0036] In this embodiment, the workflow of the above-mentioned system is described; the core design concept of the present application is to build a physiological motion perception-quality evaluation-dynamic correction triple-coupling adaptive closed-loop mechanism; the essential difference from the prior art is that, instead of simply aggregating each functional module, the present application realizes deep cooperation and intelligent linkage between each unit through innovative data flow and control flow design, and solves the problems of navigation accuracy decline and reliability quantization caused by patient physiological motion and instrument interaction;

[0037] The system starts with the physiological motion monitoring unit, which quantifies two core aspects of patient physiological motion in real time and synchronously; it collects the period, amplitude, and phase data of respiratory motion through sensors deployed on the patient's chest and abdomen, and collects the soft tissue elastic modulus changes caused by the puncture needle pressure or respiratory motion using a probe located in the puncture area; to ensure data quality, the raw sensor signals are subjected to digital filtering to eliminate motion artifacts and noise; the innovative processing logic of this unit is that it does not independently collect two sets of data, but quantifies the coupling strength of respiratory motion to the tissue deformation in the puncture area by calculating the Pearson correlation coefficient of the two within a certain sliding time window; this generates a structured comprehensive motion state data package, which is distributed to the subsequent three processing units;

[0038] The three-dimensional reconstruction processing unit receives preoperative CT or MRI and other multi-modal medical images, uses a pre-trained convolutional neural network algorithm for semantic segmentation, and constructs a static three-dimensional anatomical model containing key structures such as blood vessels and nerves; the key role of this unit is to give the static model dynamic vitality; it receives the comprehensive motion state data package from the physiological motion monitoring unit and, based on the motion vectors and deformation parameters in it, inputs it as boundary conditions and loads into a pre-set finite element analysis model; by solving the model, the system can dynamically adjust the position and shape of the three-dimensional anatomical model in accordance with the laws of biomechanics; the output is a dynamic anatomical reference model highly synchronized with the patient's current physiological state, which is transmitted to the spatial position correction unit and the multi-modal registration quality assessment unit;

[0039] At the same time, the multi-modal registration quality assessment unit performs its core quality monitoring function; it receives three inputs: real-time medical images, the aforementioned dynamic anatomical reference model, and the comprehensive motion state data package; the processing logic is based on the principle of information entropy to assess the reliability of the current registration state from three dimensions:

[0040] By calculating the Euclidean distance probability distribution of the corresponding anatomical landmark point clusters in the model and real-time images, the spatial position entropy is quantified; the more dispersed the distribution, the higher the entropy value, representing a poorer degree of position alignment;

[0041] By calculating the normalized mutual information of the corresponding region, the texture feature entropy is quantified; the lower the mutual information, the poorer the texture similarity, and the higher the entropy value;

[0042] By comparing the difference between the motion field predicted by the model and the motion field observed from the real-time image sequence based on the optical flow method, the motion consistency entropy is quantified; the larger the difference, the higher the entropy value;

[0043] The three types of entropy values are weighted and fused to generate a single, quantitative comprehensive quality evaluation index; this index is a real-time, objective self-score of the system on its own working state reliability, and is output to the dynamic compensation adjustment unit;

[0044] The dynamic compensation adjustment unit, as the core control logic module of the system, intelligently determines the intervention strength of the correction operation according to the current registration reliability of the system; it receives the comprehensive quality evaluation index, and determines it as high, medium and low according to the preset threshold; the unit generates a correction strength control parameter according to this, and outputs it to the spatial position correction unit;

[0045] The spatial position correction unit is an execution unit that performs specific correction tasks; it receives the comprehensive motion state data packet and the dynamic anatomical reference model; the unit uses an autoregressive moving average model to predict the position of the target anatomical structure in a very short time in the future according to historical respiratory cycle data; based on the theory of continuum mechanics, combined with soft tissue elastic deformation information, the possible offset caused by the puncture needle is evaluated; the two prediction results are weighted and fused to generate a preliminary correction parameter; in the specific configuration of this embodiment, the unit receives the correction strength control parameter from the dynamic compensation adjustment unit, and uses the parameter as a gain factor to modulate the application amplitude of the preliminary correction parameter; this generates a final correction parameter that has been adjusted in strength, and outputs it to the augmented reality navigation unit;

[0046] The augmented reality navigation unit, as a human-computer interaction interface, receives the final correction parameter and converts it into visual navigation information such as graphical puncture path, target point and safety boundary; it uses graphical rendering technology to superimpose these information on real-time medical images in a graphical manner; its innovation lies in that, according to the correction confidence information contained in the correction parameter, the navigation graphics are hierarchically visualized and encoded; the encoding directly maps the correction confidence index to a preset lookup table containing color, line type and transparency attributes; this design enables the doctor to intuitively judge the reliability of the navigation information, so as to make safer clinical decisions;

[0047] Through the close cooperation and information loop of the above-mentioned units, the system constructs a complete adaptive navigation process from perception, evaluation to decision, execution, and feedback, significantly improving the accuracy and safety of anesthesia puncture navigation in a dynamic interference environment. Embodiment

[0048] The comprehensive motion state data packet contains a time stamp, a three-dimensional motion amplitude vector, a tissue elasticity change amount and a motion deformation correlation coefficient;

[0049] To further clarify the mechanism, the comprehensive motion state data packet is designed as a structured data set, which aims to comprehensively and quantitatively describe the instantaneous state of physiological motion and reveal its internal correlation, which is a technical prerequisite for achieving high-precision dynamic compensation; the time stamp is used to ensure the accurate synchronization of all data streams in the system in time, which is the basis for subsequent correlation analysis and dynamic compensation; the three-dimensional motion amplitude vector is used to represent the displacement direction and size of the target region in three-dimensional space caused by motion such as respiration, providing direct geometric input for dynamic adjustment of the three-dimensional model; the tissue elasticity change is a quantitative index measured by ultrasonic elastography technology, which is used to represent the change of soft tissue stiffness caused by the compression of the puncture needle or the movement of internal organs, and is a key parameter for evaluating non-rigid deformation; the motion deformation correlation coefficient is a value calculated in the range of -1 to 1, which represents the linear coupling strength between macroscopic respiratory motion and microscopic tissue deformation; by including information in these four dimensions, the data packet enables the system not only to separate and identify the source of motion and deformation, but also to understand the causal relationship between the two, thereby achieving more accurate and robust dynamic compensation. Embodiment

[0050] The spatial position correction unit receives the application ratio of the correction strength control parameter adjusted by the dynamic compensation adjustment unit to the correction parameter, generates a final correction parameter, and outputs it to the augmented reality navigation unit;

[0051] In the specific configuration of this embodiment, the interaction between the spatial position correction unit and the dynamic compensation adjustment unit constitutes a key adaptive control loop; the spatial position correction unit receives the motion data and the model, and calculates a theoretically optimal full correction parameter; this theoretical value is based on the ideal premise of complete registration accuracy; to cope with the situation of registration quality fluctuation in reality, the dynamic compensation adjustment unit will generate a correction strength control parameter representing the confidence according to the real-time comprehensive quality evaluation index; this parameter is passed to the spatial position correction unit, which acts as an adjustment gain to directly modulate the application ratio of the correction parameter; the spatial position correction unit will use this parameter to determine the application amplitude of the full correction parameter; for example, when the control parameter indicates high confidence, the application ratio is 1.0; when it indicates medium confidence, the application ratio is reduced to 0.6; in this way, the system dynamically links the correction intensity with the registration reliability, avoiding excessive or incorrect correction when the registration quality is poor, thereby generating a more robust and safe final correction parameter, which is output to the augmented reality navigation unit. Embodiment

[0052] The dynamic compensation adjustment unit receives navigation effect evaluation data from the augmented reality navigation unit to optimize the mapping relationship between the quality index and the correction strength;

[0053] To achieve continuous self-optimization of the system, this embodiment establishes a learning-based feedback loop from the execution end to the control logic end. The augmented reality navigation unit provides navigation services while also performing an effectiveness evaluation function. It calculates the root mean square error of navigation by comparing the system-planned puncture path with the actual puncture needle trajectory after the doctor's operation (obtained through a tracker), and generates structured navigation effectiveness evaluation data by combining this with the doctor's subjective rating. This data is fed back to the dynamic compensation adjustment unit. This unit contains a learning module based on a gradient descent algorithm. This module constructs a cost function with the goal of minimizing the navigation error. By receiving the navigation effectiveness evaluation data, this module can calculate the error gradient under the current quality index-correction intensity mapping relationship and fine-tune the parameters of the mapping function accordingly. The learning process can be set to perform an offline update after processing a certain number of cases to ensure that the system can adapt to the characteristics of different patient groups or specific surgical procedures, achieving personalized and intelligent evolution of the navigation strategy. Example

[0054] The physiological motion monitoring unit includes a strain-type respiratory sensor and a contact ultrasound transducer. The strain-type respiratory sensor is deployed on the patient's chest and abdomen to monitor respiratory motion parameters, and the contact ultrasound transducer is used to detect changes in the elastic modulus of soft tissue in the puncture area.

[0055] To achieve effective monitoring of physiological movements, the physiological movement monitoring unit in this embodiment employs a specific hardware configuration. A strain gauge respiratory sensor is designed as a flexible bandage integrating piezoelectric or fiber optic sensing elements, deployed on the patient's chest and abdomen. When the patient breathes, the rise and fall of the chest causes the bandage to stretch or contract. The sensor converts this physical deformation into an electrical signal at a sampling frequency of at least 50Hz, thereby accurately monitoring parameters such as the period, amplitude, and phase of the respiratory movement. Simultaneously, a contact ultrasonic transducer is integrated into or near the puncture probe, closely adhering to the skin of the puncture area. This transducer utilizes shear wave elastography technology, emitting sound waves and analyzing the propagation velocity of shear waves within the tissue, to detect changes in the elastic modulus of the soft tissue beneath the puncture area in real time at a frame rate of at least 20Hz. This combination of dual-modal sensors enables the system to simultaneously capture motion information at both macroscopic and microscopic levels, providing a comprehensive and accurate data foundation for subsequent dynamic compensation. Example

[0056] In the weighted fusion, the weight of spatial location entropy is 40%, the weight of texture feature entropy is 35%, and the weight of motion consistency entropy is 25%.

[0057] In the multimodal registration quality assessment unit, to generate a comprehensive index that accurately reflects the overall registration quality, this embodiment employs a specific weighted fusion strategy for the three types of entropy values. The weight of spatial location entropy is set to the highest at 40%. The rationale for this setting is that the alignment of anatomical landmarks in space is the most direct and important indicator for assessing the success of registration, directly affecting the geometric accuracy of navigation. The weight of texture feature entropy is set to 35%. This reflects the consistency between real-time images and the reference model in terms of tissue texture details and is an important supplement to assessing the fineness of registration, playing a significant role in identifying soft tissues lacking clear boundaries. The weight of motion consistency entropy is set to 25%. This weight is relatively low because it mainly assesses the consistency of dynamic processes, serving as a verification and supplement to the static location and texture assessment. These weight coefficients are not set arbitrarily but are determined through multiple regression analysis and principal component analysis of large-scale clinical data, aiming to maximize the correlation between the comprehensive assessment index and the actual navigation accuracy annotated by experts. Example

[0058] The strength grades include high strength, medium strength and low strength. A comprehensive quality assessment index above 0.8 is defined as high strength, an index between 0.5 and 0.8 is defined as medium strength, and an index below 0.5 is defined as low strength.

[0059] High strength grade corresponds to 100% application rate, medium strength grade corresponds to 60% application rate, and low strength grade corresponds to 30% application rate.

[0060] To transform quantitative quality assessment into explicit control strategies, this embodiment specifies in detail the adjustment mechanism of the dynamic compensation adjustment unit; the unit establishes a three-level correction intensity level system and maps the comprehensive quality assessment index (whose value range is 0 to 1) to these levels;

[0061] When the comprehensive quality assessment index is higher than 0.8, the system determines that the current registration state is at a high intensity level, i.e., a high reliability state. In this state, the system has high confidence in its own registration results, and the dynamic compensation adjustment unit will output a control parameter corresponding to 100% application ratio. This allows the spatial position correction unit to fully apply its calculated correction amount to compensate for the errors caused by motion and deformation to the greatest extent.

[0062] When the index is between 0.5 and 0.8, the system determines that the registration status is at a medium intensity level; this indicates that the registration is basically reliable, but there is some uncertainty. In order to strike a balance between improving accuracy and avoiding risks, a more prudent strategy is adopted, reducing the application ratio of correction parameters to 60%; this moderate compensation can largely correct the deviation, while avoiding excessive amplification of minor errors in registration.

[0063] When the index is below 0.5, the system determines that the registration status is at a low intensity level, i.e., a low reliability state. This usually means that there are problems such as violent motion or severe degradation of image quality. At this time, the risk of blindly making large corrections is very high. The system will reduce the application ratio to 30%, no longer pursuing perfect accuracy compensation, but prioritizing the stability and safety of navigation, and only providing basic, directional guidance.

[0064] It should be understood that the specific values ​​of the two thresholds, 0.8 and 0.5, are not fixed, but rather are the optimal operating points determined by receiver operating characteristic curve analysis of large-scale clinical datasets. The aim is to balance the sensitivity and specificity of the correction, thereby optimizing the overall navigation performance. Example

[0065] The hierarchical visual coding includes color coding and transparency coding. High calibration confidence corresponds to a solid green line, medium calibration confidence corresponds to a dashed yellow line, and low calibration confidence corresponds to a semi-transparent red line.

[0066] In order to intuitively convey the confidence status of the system to the operating doctor, the augmented reality navigation unit in this embodiment adopts a hierarchical visualization coding mechanism that integrates color coding, line coding and transparency coding.

[0067] When the received correction parameters have high correction confidence (corresponding to a high intensity level), the navigation path will be rendered as a clear solid green line; green generally represents safety and passability, while a solid line represents certainty and stability; this visual presentation provides doctors with a clear visual cue regarding the high fidelity of the navigation data;

[0068] When the confidence level is medium (corresponding to medium intensity level), the navigation path is displayed as a yellow dashed line; yellow is usually used as a warning color, while the dashed line suggests that the path is not entirely certain; this combination is intended to remind doctors that the system is providing corrective suggestions, but there is some uncertainty, and they should be used with caution.

[0069] When the confidence level is low (corresponding to a low intensity level), navigation information (such as the boundary of a danger zone) will be displayed in a red semi-transparent manner. Red is a strong warning signal, while the semi-transparent effect visually reduces the intensity of its guidance, implying that its reference value is limited. This visual presentation is a warning to doctors, indicating that the system's current reliability is low and judgment should be based primarily on the original images and clinical experience.

[0070] Through this intelligent hierarchical visualization design, the system transforms complex internal quality assessment results into an intuitive visual language that doctors can easily understand, greatly improving the efficiency of human-computer interaction and the safety of puncture procedures.

[0071] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A visualization and navigation system for anesthesia puncture combining augmented reality and 3D reconstruction, characterized in that: include: The physiological motion monitoring unit is used to simultaneously collect respiratory motion cycle data and soft tissue elastic deformation information, and generate a comprehensive motion status data package by calculating the correlation coefficient between respiratory amplitude and the rate of change of tissue elastic modulus. The 3D reconstruction processing unit is used to construct a 3D anatomical model of the patient based on multimodal medical image data, receive comprehensive motion state data packets, and dynamically adjust the 3D anatomical model according to motion parameters to generate a dynamic anatomical reference model. The spatial position correction unit is used to receive the comprehensive motion state data packet and the dynamic anatomical reference model, predict the position of the target anatomical structure based on the respiratory motion cycle data, evaluate the position offset of the puncture needle based on the soft tissue elastic deformation information, and generate correction parameters by weighted fusion of the predicted position and offset. The multimodal registration quality assessment unit is used to receive real-time medical image data, dynamic anatomical reference models and comprehensive motion state data packets, calculate spatial location entropy, texture feature entropy and motion consistency entropy, and perform weighted fusion of the three types of entropy values ​​to generate a comprehensive quality assessment index. The dynamic compensation adjustment unit is used to receive the comprehensive quality assessment index, determine the correction intensity level based on the index value, and generate correction intensity control parameters to output to the spatial position correction unit. The augmented reality navigation unit receives the correction parameters adjusted by the correction intensity control parameters, overlays the navigation information onto the real-time medical image, and performs hierarchical visual encoding based on the correction reliability.

2. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The comprehensive motion status data package includes timestamps, three-dimensional motion amplitude vectors, tissue elasticity changes, and motion deformation correlation coefficients.

3. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The spatial position correction unit receives the correction intensity control parameters from the dynamic compensation adjustment unit, adjusts the application ratio of the correction parameters, and generates the final correction parameters to be output to the augmented reality navigation unit.

4. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The dynamic compensation adjustment unit receives navigation performance evaluation data from the augmented reality navigation unit and uses it to optimize the mapping relationship between the quality index and the correction intensity.

5. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The physiological motion monitoring unit includes a strain-type respiratory sensor and a contact ultrasound transducer. The strain-type respiratory sensor is deployed on the patient's chest and abdomen to monitor respiratory motion parameters, while the contact ultrasound transducer is used to detect changes in the elastic modulus of soft tissue in the puncture area.

6. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, In the weighted fusion, the weight of spatial location entropy is 40%, the weight of texture feature entropy is 35%, and the weight of motion consistency entropy is 25%.

7. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The strength grades include high strength, medium strength and low strength. A high strength grade is defined as an index above 0.8, a medium strength grade is defined as an index between 0.5 and 0.8, and a low strength grade is defined as an index below 0.

5.

8. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 7, characterized in that, High strength grade corresponds to 100% application rate, medium strength grade corresponds to 60% application rate, and low strength grade corresponds to 30% application rate.

9. The anesthesia puncture visualization and navigation system combining augmented reality and three-dimensional reconstruction according to claim 1, characterized in that, The hierarchical visual encoding includes color encoding and transparency encoding. High calibration confidence corresponds to a solid green line, medium calibration confidence corresponds to a dashed yellow line, and low calibration confidence corresponds to a semi-transparent red line.

Images