Lung tissue lesion location positioning system and method based on three-dimensional reconstruction

Abstract

The invention belongs to the technical field of lung tissue lesion positioning, and particularly relates to a lung tissue lesion position positioning system and method based on three-dimensional reconstruction. According to the method, overall and local two-stage registration is combined, whole-lung expansion errors are corrected, local deformation is compensated, the lesion position and image alignment is achieved, the three-dimensional lesion position is displayed through the augmented reality technology, doctors are helped to visually master spatial distribution of the focus in the body of a patient, the safety and the success rate of operation or needle biopsy are improved, and the operation efficiency is improved. Based on a breathing model and an elastic registration method, the method can adapt to lung compliance differences and breathing movement conditions of different patients, is particularly suitable for severe or old patients difficult to keep stable breathing, reduces the risk of blind penetration or excessive resection through accurate positioning, and reduces tissue damage and the occurrence rate of postoperative complications.

Description

Technical Field

[0001] The present invention belongs to the technical field of lung tissue lesion positioning, and particularly relates to a lung tissue lesion position positioning system and method based on three-dimensional reconstruction. Background Art

[0002] Accurate localization of lung tissue lesions is of great significance for minimally invasive treatments such as thoracic surgery, interventional biopsy, and ablation. In thoracoscopic surgery, accurate localization of deep or tiny lesions in lung tissue is the key to successful surgery. The current mainstream technology relies on preoperative CT three-dimensional reconstruction to construct a virtual lesion model and achieves positioning through an intraoperative image navigation system. The traditional method relies on intraoperative visual exploration. The lung tissue deforms with respiratory movement, resulting in displacement deviations between the lesion location marked on the preoperative image and the actual anatomical structure during surgery. Tiny lesions or subsolid nodules are difficult to identify with the naked eye in collapsed lung lobes, and it is even easier to get lost in complex adhesions or anatomical variations.

[0003] Traditional methods usually involve obtaining CT sequences of the patient while they are inhaling or holding their breath before surgery, observing the approximate location of the lesion through two-dimensional slices, and then using C-arm fluoroscopy or ultrasound for reference during surgery. However, lung tissue deforms due to ventilation regulation, respiratory movement, and surgical operations. Simply relying on preoperative image annotation cannot reflect dynamic changes in the lungs during surgery. Previous navigation systems mostly used static alignment, and manual recalibration was required once the patient's breathing or body position changed, reducing work efficiency and increasing radiation exposure and surgery time. Summary of the Invention

[0004] The purpose of the present invention is to provide a system and method for locating lung tissue lesions based on three-dimensional reconstruction, which can accurately perform closed-loop ventilation control and multi-level registration compensation of anatomical landmarks, dynamically correct lung inflation errors and perform local deformation compensation, and provide more reliable technical support.

[0005] The technical solutions adopted by the present invention are as follows: A method for locating lung tissue lesions based on three-dimensional reconstruction, comprising: Obtaining CT image data of the patient's lung tissue at a preset lung inflation state, and constructing a three-dimensional virtual model of the lung tissue based on the CT image data, including the location of the lesion and anatomical landmarks, where the anatomical landmark is the junction of the azygos vein and the superior vena cava; Acquire the patient's lung tissue image in real time and, combined with the preset lung inflation state, obtain the inflation error of the anatomical landmarks; obtaining a lung tissue ventilation control strategy based on the inflation error, and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; Obtaining anatomical landmarks after adjusting to a preset lung inflation state, superimposing the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; Acquiring local structural features around anatomical landmarks after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtaining lesion offset compensation based on the local structural features; An overlap correction strategy is obtained based on lesion offset compensation, a corrected final overlap image is generated based on the overlap correction strategy, and the lesion position is located based on the final overlap image.

[0006] In a preferred embodiment, the step of acquiring a patient's lung tissue image in real time and obtaining the inflation error of an anatomical landmark based on a preset lung inflation state includes: Acquire the patient's lung tissue images in real time, and obtain real-time anatomical landmark information based on the lung tissue images; Acquiring real-time anatomical landmark spatial coordinates based on real-time anatomical landmark information; Acquiring preset anatomical landmark information based on a preset lung inflation state; Acquiring spatial coordinates of preset anatomical landmark points based on preset anatomical landmark information; The deviation of the anatomical landmark point is obtained according to the real-time anatomical landmark point spatial coordinates and the preset anatomical landmark point spatial coordinates, and is marked as an expansion error.

[0007] In a preferred embodiment, the steps of obtaining a lung tissue ventilation control strategy based on the inflation error and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy include: Acquiring a ventilation control table, wherein the ventilation control table includes a plurality of inflation error intervals and a lung tissue ventilation control strategy corresponding to each inflation error interval; Obtaining a corresponding lung tissue ventilation control strategy from a ventilation control table according to an inflation error interval corresponding to the inflation error; Based on the lung tissue ventilation control strategy, the patient's lung tissue is adjusted to the preset lung inflation state.

[0008] In a preferred embodiment, the step of adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy includes: Regulate the patient's lung tissue inflation state according to the lung tissue ventilation control strategy; obtaining an adjusted lung tissue inflation state of the patient's lung tissue, and obtaining an adjusted anatomical landmark deviation based on the adjusted lung tissue inflation state; Obtaining an adjusted anatomical landmark deviation threshold, and determining whether the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; If the adjusted anatomical landmark deviation does not exceed the adjusted anatomical landmark deviation threshold, it is determined that the patient's lung tissue inflation state meets the preset lung inflation state; If the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold, determining that the patient's lung tissue inflation state does not meet the preset lung inflation state, and obtaining an anatomical landmark excess value at which the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; Acquiring a ventilation fine-tuning table, wherein the ventilation fine-tuning table includes a plurality of anatomical landmark point excess value intervals and a ventilation fine-tuning strategy corresponding to each anatomical excess value interval; Obtaining a corresponding ventilation fine-tuning strategy from a ventilation fine-tuning table according to an anatomical landmark point excess value interval corresponding to the anatomical landmark point excess value; Fine-tune the patient's lung tissue inflation status based on ventilation fine-tuning strategies; obtaining a lung tissue inflation state of the patient after fine-tuning the lung tissue, and obtaining a fine-tuned anatomical landmark deviation based on the fine-tuned lung tissue inflation state; Determine whether the deviation of the anatomical landmarks after fine-tuning exceeds the preset conditions for fine-tuning. If not, the fine-tuning is determined to be normal. If exceeded, the fine-tuning is determined to be abnormal, and the lung tissue ventilation control strategy is re-acquired.

[0009] In a preferred embodiment, the steps of obtaining anatomical landmarks after adjusting the lung to a preset inflation state and superimposing the lesion position in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image include: Obtain the lesion location and anatomical landmarks in the patient's 3D virtual model of lung tissue; Obtaining anatomical landmarks of the lung tissue image after the patient's lung tissue is adjusted to a preset lung inflation state; Based on the anatomical landmarks of the lung tissue image after adjustment to a preset lung inflation state and the anatomical landmarks in the three-dimensional virtual model of lung tissue, the lesion position in the three-dimensional virtual model of lung tissue is superimposed on the lung tissue image to generate an initial overlapping image.

[0010] In a preferred embodiment, the step of obtaining local structural features around an anatomical landmark after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtaining lesion offset compensation based on the local structural features includes: Acquiring local structural features around anatomical landmarks in a lung tissue image after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology; Obtaining corresponding respiratory state values ​​and bronchial morphology values ​​based on the respiratory state and bronchial morphology respectively; Obtain corresponding respiratory weight and bronchial weight based on respiratory state and bronchial morphology respectively; Lesion offset compensation is obtained based on respiratory state value, bronchial morphology value, respiratory weight and bronchial weight.

[0011] In a preferred embodiment, after the step of obtaining lesion offset compensation based on the respiratory state value, the bronchial morphology value, the respiratory weight, and the bronchial weight, the method further includes: Obtaining a lesion offset compensation threshold, and determining whether the lesion offset compensation exceeds the lesion offset compensation threshold; If the lesion offset compensation exceeds the lesion offset compensation threshold, the lesion offset compensation is determined to be abnormal, and the time node when the compensation is determined to be abnormal is obtained and marked as the end time of the historical period; If the lesion offset compensation does not exceed the lesion offset compensation threshold, it is determined that the lesion offset compensation is normal; Obtaining a lesion offset excess value where the lesion offset compensation exceeds a lesion offset compensation threshold, and obtaining a corresponding lead-in duration based on the lesion offset excess value; Construct historical periods based on the introduction duration and the end time of the historical period; Acquire lung tissue images of multiple patients within a historical period, and acquire historical respiratory states and historical bronchial morphologies in the lung tissue images of each patient, and respectively acquire historical respiratory state values ​​and historical bronchial morphology values ​​corresponding to each historical respiratory state and historical bronchial morphology; A new lesion offset compensation is obtained based on a plurality of historical respiratory state values, a plurality of historical bronchial morphology values, a respiratory state value, a bronchial morphology value, a respiratory weight, and a bronchial weight.

[0012] In a preferred embodiment, the steps of obtaining an overlap correction strategy based on lesion offset compensation, generating a corrected final overlap image based on the overlap correction strategy, and locating the lesion position based on the final overlap image include: Obtaining an overlap correction table, wherein the overlap correction table includes a plurality of lesion offset compensation intervals and an overlap correction strategy corresponding to each lesion offset compensation interval; Obtaining a corresponding overlap correction strategy from the overlap correction table according to the lesion offset compensation interval corresponding to the lesion offset compensation; Adjusting the lesion position in the initial overlapping image based on the overlap correction strategy and generating a corrected final overlapping image; The lesion location was determined based on anatomical landmarks in the final overlapping images.

[0013] The present invention further provides a lung tissue lesion location positioning system based on three-dimensional reconstruction, which is used in the above-mentioned lung tissue lesion location positioning method based on three-dimensional reconstruction, comprising: A virtual model module is used to obtain CT image data of the patient's lung tissue in a preset lung inflation state and construct a three-dimensional virtual model of the lung tissue based on the CT image data, including the location of the lesion and anatomical landmarks, where the anatomical landmark is the junction of the azygos vein and the superior vena cava; The inflation error module is used to obtain the patient's lung tissue image in real time and obtain the inflation error of the anatomical landmarks based on the preset lung inflation state; A ventilation control module is used to obtain a lung tissue ventilation control strategy based on the inflation error, and adjust the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; An overlapping construction module is used to obtain anatomical landmarks after adjusting to a preset lung inflation state, and to superimpose the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; an offset compensation module, configured to obtain local structural features around anatomical landmarks after adjustment to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and to obtain lesion offset compensation based on the local structural features; The lesion localization module obtains an overlap correction strategy based on lesion offset compensation, generates a corrected final overlap image based on the overlap correction strategy, and locates the lesion position based on the final overlap image.

[0014] And, a lung tissue lesion location positioning terminal based on three-dimensional reconstruction, comprising: one or more processors; a storage device having one or more programs stored thereon; When one or more programs are executed by one or more processors, the one or more processors implement a lung tissue lesion location positioning method based on three-dimensional reconstruction.

[0015] The technical effects achieved by the present invention are: The present invention combines global and local two-level registration to correct the error in whole-lung expansion and compensate for local deformation, thus achieving alignment between the lesion location and the image. Augmented reality technology displays the three-dimensional lesion location, helping physicians to intuitively grasp the spatial distribution of lesions in the patient's body, thereby improving the safety and success rate of surgery or puncture biopsy. The method based on respiratory model and elastic registration can adapt to the differences in lung compliance and respiratory movement of different patients, and is particularly suitable for critically ill or elderly patients who have difficulty maintaining stable breathing. Precise positioning reduces the risk of blind puncture or excessive resection, and reduces tissue damage and the incidence of postoperative complications. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flow chart of the method provided by the present invention; Figure 2 This is a system module diagram provided by the present invention. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0018] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0019] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in a preferred embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive of other embodiments.

[0020] Secondly, the present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention in detail, for the sake of convenience, the schematic diagrams are only examples and should not limit the scope of protection of the present invention.

[0021] Please see the attached Figure 1 As shown, a method for locating lung tissue lesions based on three-dimensional reconstruction is provided, comprising: S1. Obtaining CT image data of the patient's lung tissue in a preset lung inflation state, and constructing a three-dimensional virtual model of the lung tissue including the lesion location and anatomical landmarks based on the CT image data, wherein the anatomical landmark is the junction of the azygos vein and the superior vena cava; S2. Acquire the patient's lung tissue image in real time and, based on the preset lung inflation state, obtain the inflation error of the anatomical landmarks; S3. obtaining a lung tissue ventilation control strategy based on the inflation error, and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; S4, obtaining anatomical landmarks after adjusting to a preset lung inflation state, and superimposing the lesion position in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; S5. Acquire local structural features around the anatomical landmark after adjustment to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtain lesion offset compensation based on the local structural features; S6. Obtain an overlap correction strategy based on lesion offset compensation, generate a corrected final overlap image based on the overlap correction strategy, and locate the lesion position based on the final overlap image.

[0022] As in steps S1 to S6 above, CT images acquired from the patient at a preset lung inflation state are used to extract the surface and internal structure of the lung tissue through image segmentation and voxel reconstruction technology, and the lesion location is marked in the three-dimensional virtual model. At the same time, the junction of the azygos vein and the superior vena cava is used as a stable anatomical landmark point. During the surgery or intervention, the patient's current lung image is acquired in real time and aligned with the landmark point position in the three-dimensional virtual model at the preset state. The position difference between the two in three-dimensional space is the inflation error. Based on the inflation error of the landmark point, the difference between the patient's current lung tissue and the preset lung tissue is obtained, and the corresponding ventilation adjustment parameters are generated. After the ventilation is adjusted to the preset state, the anatomical landmark point position is re-located, and the pre-calibrated lesion location in the three-dimensional virtual model is projected onto the real-time image to achieve preliminary lesion-image overlap display. Even if the overall lung volume matches, local tissues (such as different lung segments and bronchial shapes) may still have slight differences. Deformation, by extracting features of local structures around anatomical landmarks (respiratory state fluctuations, bronchial bifurcation angles, etc.), quantifying the offset of the lesion area relative to the model, and obtaining lesion offset compensation, the lesion offset compensation is applied to the initial overlapping images to achieve fine-tuning and correction of the lesion position, generating a final highly consistent overlapping image, and locating the lesion position based on this. Combining global and local two-level registration, it not only corrects the error of full lung expansion, but also compensates for local deformation, achieving alignment between the lesion position and the image. Augmented reality technology displays the three-dimensional lesion position, helping physicians intuitively grasp the spatial distribution of lesions in the patient's body, improving the safety and success rate of surgery or puncture biopsy. The method based on respiratory model and elastic registration can adapt to the differences in lung compliance and respiratory movement of different patients, and is especially suitable for critically ill or elderly patients who have difficulty maintaining stable breathing. Accurate positioning reduces the risk of blind puncture or excessive resection, reducing tissue damage and the incidence of postoperative complications.

[0023] In a preferred embodiment, the step of acquiring a patient's lung tissue image in real time and obtaining the inflation error of an anatomical landmark in combination with a preset lung inflation state includes: S201, acquiring a lung tissue image of the patient in real time, and acquiring real-time anatomical landmark information based on the lung tissue image; S202, acquiring real-time anatomical landmark spatial coordinates based on real-time anatomical landmark information; S203, obtaining preset anatomical landmark information based on a preset lung inflation state; S204, obtaining the spatial coordinates of the preset anatomical landmark points based on the preset anatomical landmark point information; S205 : Obtain anatomical landmark point deviations according to the real-time anatomical landmark point spatial coordinates and the preset anatomical landmark point spatial coordinates, and mark them as expansion errors.

[0024] As in steps S201 to S205 above, real-time images are used to obtain continuous frame data of the patient's current lung tissue, and the pixel area at the intersection of the azygos vein and the superior vena cava in the image is automatically identified. The position of the real-time anatomical landmark point is extracted, and the pixel position of the landmark point in the image coordinate system is converted to a three-dimensional physical space coordinate system with the help of system calibration. The anatomical landmark point information in the three-dimensional virtual model of the lung tissue is collected based on the preset lung inflation state. The coordinate position of the anatomical landmark point is calibrated in advance and converted to the physical space coordinate system identical to the real-time environment. The spatial coordinates of the anatomical landmark point obtained in real time are compared with the spatial coordinates in the preset state to obtain their difference in three-dimensional space, which is the inflation error. The calculation formula for the anatomical landmark point deviation is: Where P represents the deviation of the anatomical landmark point, x represents the X-axis coordinate point of the real-time anatomical landmark point space coordinate, y represents the Y-axis coordinate point of the real-time anatomical landmark point space coordinate, z represents the Z-axis coordinate point of the real-time anatomical landmark point space coordinate, u represents the X-axis coordinate point of the preset anatomical landmark point space coordinate, v represents the Y-axis coordinate point of the preset anatomical landmark point space coordinate, and w represents the Z-axis coordinate point of the preset anatomical landmark point space coordinate. Through multiple real-time acquisitions and preset comparisons, the expansion error can be accurately obtained, making subsequent ventilation adjustment and lesion projection more accurate. The automated landmark point detection and coordinate conversion process avoids manual alignment or manual measurement errors and improves stability and repeatability.

[0025] In a preferred embodiment, the steps of obtaining a lung tissue ventilation control strategy based on the inflation error and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy include: S301, obtaining a ventilation control table, wherein the ventilation control table includes multiple inflation error intervals and a lung tissue ventilation control strategy corresponding to each inflation error interval; S302, obtaining a corresponding lung tissue ventilation control strategy from a ventilation control table according to the inflation error interval corresponding to the inflation error; S303: Adjust the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy.

[0026] As in steps S301 to S303 above, a ventilation control table is pre-set. The ventilation control table is a mapping table established based on a large amount of preoperative lung function data, lung compliance model and clinical experience. The ventilation strategy table can be continuously optimized based on historical patient data, real-time feedback effects, etc. The table divides the lung tissue expansion error into several intervals (for example: less than 2mm, 2-5mm, greater than 5mm, etc.), and sets a set of ventilation control strategies for each error interval. If the error interval is 0-2mm, the current ventilation parameters are maintained. If the error interval is 2-5mm, the tidal volume VT is increased and the positive end-expiratory pressure (PEEP) is maintained unchanged. If the error interval is greater than 5mm, the positive end-expiratory pressure (PEEP) is increased and the respiratory rate is fine-tuned. The current expansion error is obtained (for example, 3.8mm is calculated), the corresponding error interval (such as 2-5mm) is found in the control table, and the corresponding ventilation control strategy is read (for example, the tidal volume VT is adjusted to maintain the expiratory pressure). The positive end-expiratory pressure (PEEP) remains unchanged), and the selected ventilation control strategy is sent to the ventilator or artificial ventilation system, and parameters such as tidal volume (VT), positive end-expiratory pressure (PEEP), inspiration-expiration ratio, and respiratory rate are adjusted in real time to gradually restore the patient's lung tissue to the target expansion state. By continuously monitoring the position of anatomical landmarks, it is determined whether the current expansion state has reached the target. If not, the strategy lookup and adjustment can be executed again to achieve closed-loop control, and the expansion error is mapped one-to-one with the ventilation strategy to avoid manual judgment and trial and error operations, and to achieve fast, accurate, and intelligent ventilation parameter adjustment. Through ventilation adjustment, the actual state of the lungs tends to the preset state, reducing the image error and alignment offset caused by inconsistent lung expansion from the source, and providing higher precision support for subsequent lesion superposition positioning. Through error range subdivision and corresponding strategies, differentiated control can be set for patients with different body shapes, lung compliance, and lesion locations to avoid overventilation or lung tissue damage.

[0027] In a preferred embodiment, the step of adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy includes: S3031. Regulating the patient's lung tissue inflation state according to a lung tissue ventilation control strategy; S3032: Obtaining the adjusted lung tissue expansion state of the patient's lung tissue, and obtaining the adjusted anatomical landmark deviation based on the adjusted lung tissue expansion state; S3033, obtaining an adjusted anatomical landmark deviation threshold, and determining whether the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; If the adjusted anatomical landmark deviation does not exceed the adjusted anatomical landmark deviation threshold, it is determined that the patient's lung tissue inflation state meets the preset lung inflation state; If the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold, determining that the patient's lung tissue inflation state does not meet the preset lung inflation state, and obtaining an anatomical landmark excess value at which the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; S3034. Obtain a ventilation fine-tuning table, wherein the ventilation fine-tuning table includes a plurality of anatomical landmark point excess value intervals and a ventilation fine-tuning strategy corresponding to each anatomical excess value interval; S3035. Obtain a corresponding ventilation fine-tuning strategy from the ventilation fine-tuning table according to the anatomical landmark point excess value interval corresponding to the anatomical landmark point excess value; S3036. Fine-tune the patient's lung tissue inflation state based on ventilation fine-tuning strategies; S3037. Obtaining the expansion state of the patient's lung tissue after fine-tuning, and obtaining the deviation of the fine-tuned anatomical landmarks based on the fine-tuned expansion state of the lung tissue; S3038. Determine whether the deviation of the anatomical landmark point after fine-tuning exceeds the preset condition of fine-tuning. If not, determine that the fine-tuning is normal. If exceeded, determine that the fine-tuning is abnormal, and re-acquire the lung tissue ventilation control strategy.

[0028] As in steps S3031 to S3038 above, according to the ventilation control strategy selected previously (such as adjusting tidal volume, PEEP, etc.), the output parameters of the ventilation device are adjusted to affect the degree of inflation of the patient's alveoli, thereby changing the expansion state of the lung tissue. After the strategy is executed, the adjusted lung tissue image is acquired in real time, and the spatial position of the anatomical landmark point (such as the junction of the azygos vein and the superior vena cava) is extracted and compared with the preset state to calculate the adjusted deviation value. The calculation can be performed using the calculation formula of the anatomical landmark point deviation, or other calculation formulas similar to the Euclidean distance calculation formula, which will not be excessively described here. Set an adjustment deviation threshold (such as 2mm) as the judgment standard. If the adjusted deviation is less than the threshold, it means that the preset lung expansion state has been achieved. If the deviation exceeds the threshold, it means that the preset state has not been achieved and further fine-tuning is required. At the same time, the excess value of the anatomical landmark point is extracted, that is, the difference between the actual deviation and the threshold, which is used as the basis for the next fine-tuning. A ventilation fine-tuning table is constructed, which is a mapping table established based on a large amount of preoperative lung function data, lung compliance model and clinical experience. The ventilation fine-tuning table can be continuously optimized based on historical patient data, real-time feedback effects, etc., and the excess values ​​of different ranges (such as 1-2mm, 2-3mm, etc.) are mapped to the corresponding The corresponding fine-tuning strategy (such as increasing PEEP, prolonging inspiratory time, etc.) is selected according to the table of the exceeded value range. The most appropriate ventilation fine-tuning strategy is selected. Based on the fine-tuning strategy obtained from the table, the ventilation equipment settings are fine-tuned to achieve smaller and more accurate lung tissue adjustment (such as slightly increasing or decreasing tidal volume, changing respiratory rate). The aforementioned image extraction, anatomical landmark positioning and spatial registration operations are repeated to obtain the landmark deviation in the expanded state of the lung tissue after fine-tuning. The judgment standard after fine-tuning is set (such as a deviation of less than 1mm is considered a successful fine-tuning). If the standard is met, it is determined that the expansion state of the lung tissue has reached the preset state. If If the conditions are not met, the fine-tuning is deemed to have failed, and the system returns to reselecting the ventilation control strategy and re-enters the control closed loop until the conditions are met. Combining coarse adjustment with fine adjustment can not only quickly converge to the target state, but also further eliminate the residual error through fine control, and achieve millimeter or even submillimeter level lung tissue position alignment. By introducing the anatomical landmark point excess value and ventilation fine-tuning table mechanism, the ventilation strategy can be dynamically adjusted for different patients, different lung compliances, and different lesion areas to achieve highly personalized breathing control. Fine-tuning can avoid excessive expansion or compression of the alveoli due to excessive adjustment, thereby improving patient safety and comfort.

[0029] In a preferred embodiment, the steps of obtaining anatomical landmarks after adjusting the lung to a preset inflation state and superimposing the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image include: S401, obtaining the lesion location and anatomical landmarks in the patient's lung tissue three-dimensional virtual model; S402, obtaining anatomical landmarks of the lung tissue image after the patient's lung tissue is adjusted to a preset lung inflation state; S403 : Based on the anatomical landmarks of the lung tissue image after adjustment to the preset lung inflation state and the anatomical landmarks in the three-dimensional virtual model of lung tissue, the lesion position in the three-dimensional virtual model of lung tissue is superimposed on the lung tissue image to generate an initial overlapping image.

[0030] As described in steps S401 to S403 above, the 3D coordinates of the target lesion are accurately annotated in the reconstructed 3D virtual lung model. Furthermore, the model contains the coordinates of anatomical landmarks (the junction of the azygos vein and the superior vena cava) that are registered with the real-time image. After ventilation control is completed and the lung tissue reaches a preset inflation state, a current lung image is captured based on the real-time image. The 3D spatial coordinates of the anatomical landmarks in the image are identified and extracted. Using the landmark coordinates obtained in the adjusted image as a reference, the corresponding landmarks in the 3D model are aligned to the real-time image coordinate system. The coordinates of the entire or a portion of the model (including the lesion area) are mapped to the current image coordinate system. The lesion location is superimposed on the real-time lung tissue image in a semi-transparent, pseudo-colored, or edge-delineated manner to generate an initial overlay image. The preoperatively annotated 3D lesion is directly mapped to the intraoperative image, allowing the physician to clearly see the true projected location of the lesion in the current lung without the need for repeated manual measurement or estimation. Through precise landmark registration, high spatial consistency between the 3D model and the real-time image is ensured, reducing positioning errors caused by lung inflation differences or respiratory motion.

[0031] In a preferred embodiment, the step of obtaining local structural features around an anatomical landmark after adjustment to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtaining lesion offset compensation based on the local structural features includes: S501, obtaining local structural features around anatomical landmarks in a lung tissue image after the lung is adjusted to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology; S502, obtaining corresponding respiratory state values ​​and bronchial morphology values ​​based on the respiratory state and bronchial morphology respectively; S503, obtaining corresponding respiratory weight and bronchial weight based on respiratory state and bronchial morphology respectively; S504: Obtain lesion offset compensation based on the respiratory state value, bronchial morphology value, respiratory weight, and bronchial weight.

[0032] As in steps S501 to S504 above, after the patient's lung tissue is adjusted to a preset inflation state, the latest intraoperative image is collected, and an area within a certain radius around the anatomical landmark (the junction of the azygos vein and the superior vena cava) is selected as the local structural area. Within this area, two key features are extracted: respiratory state (which can be quantified as indicators such as lung volume change rate, diaphragm position, airway airflow velocity, or chest movement amplitude) and bronchial morphology (assessing geometric parameters such as bifurcation angle, diameter, tortuosity, and branching level of the bronchial tree in this area). The respiratory state indicators are normalized to obtain The respiratory state value describes the current respiratory movement intensity or stability of the region. Based on morphological analysis of the bronchial morphology (such as centerline extraction and curvature statistics), a scalar value, namely the bronchial morphology value, is calculated to reflect the bronchial deformation or branch complexity of the region. According to clinical experience and lung anatomical characteristics, the respiratory weight and bronchial weight are set. Usually, the sum of the respiratory weight and the bronchial weight is equal to one. The weight size can be preset according to the difference in rigidity and flexibility of different lung segments. The respiratory state value and the bronchial morphology value are combined with the weight to calculate the lesion offset compensation. The calculation formula for lesion offset compensation is: , where B represents the lesion offset compensation, H represents the respiratory state value, and Q represents the bronchial morphology value. Expressed as breathing weight, Expressed as bronchial weights, they are used to perform local correction on the lesion position in the initial overlapping images. Both respiratory motion and bronchial deformation are considered simultaneously to achieve fine-level compensation for the lesion position and reduce the registration error caused by substructure deformation. Real-time image data is called to dynamically calculate eigenvalues ​​and weights. This allows for personalized compensation for different respiratory cycles or implant positions. By weighted fusion of two independent sources of deformation, under- or over-compensation by a single factor is avoided, thereby improving overall positioning stability.

[0033] In a preferred embodiment, after the step of obtaining lesion offset compensation based on the respiratory state value, the bronchial morphology value, the respiratory weight, and the bronchial weight, the method further includes: S505: Obtain a lesion offset compensation threshold, and determine whether the lesion offset compensation exceeds the lesion offset compensation threshold; If the lesion offset compensation exceeds the lesion offset compensation threshold, the lesion offset compensation is determined to be abnormal, and the time node when the compensation is determined to be abnormal is obtained and marked as the end time of the historical period; If the lesion offset compensation does not exceed the lesion offset compensation threshold, it is determined that the lesion offset compensation is normal; S506, obtaining a lesion offset excess value where the lesion offset compensation exceeds a lesion offset compensation threshold, and obtaining a corresponding lead-in duration based on the lesion offset excess value; S507: Construct a historical period based on the introduction duration and the end time of the historical period; S508. Acquire lung tissue images of multiple patients within a historical period, and acquire a historical respiratory state and a historical bronchial morphology in the lung tissue image of each patient, and respectively acquire a historical respiratory state value and a historical bronchial morphology value corresponding to each historical respiratory state and historical bronchial morphology; S509 , obtaining a new lesion offset compensation based on multiple historical respiratory state values, multiple historical bronchial morphology values, respiratory state values, bronchial morphology values, respiratory weights, and bronchial weights.

[0034] As in steps S505 to S509 above, a lesion offset compensation threshold is pre-set to evaluate the current lesion offset compensation. If the threshold is exceeded, the current compensation is determined to be abnormal, and the current time is recorded as the end time of the historical period. If the threshold is not exceeded, the current compensation is determined to be normal, and the next step can be continued or directly applied to image correction. The difference between the lesion offset compensation and the lesion offset compensation threshold, that is, the lesion offset excess value, is calculated, and the corresponding introduction time is determined based on this excess value. Each value corresponds to an interval, and each interval corresponds to an introduction time. Based on the introduction time and the end time of the historical period, the start time is calculated to determine a historical period for determining the duration of obtaining historical data. Real-time lung tissue image data of multiple patients are collected within the historical period. For each patient image, the respiratory state and bronchial morphological features within the period are repeatedly extracted and converted into a series of historical respiratory state values ​​and historical bronchial morphological values. A new lesion offset compensation is weighted and comprehensively calculated based on multiple historical respiratory state values, multiple historical bronchial morphological values, respiratory state values, bronchial morphological values, respiratory weights, and bronchial weights. The calculation formula for the new lesion offset compensation is: , where represents the new lesion offset compensation, i represents the number of multiple historical respiratory state values ​​and the number of multiple historical bronchial morphology values, Represented as the i-th historical respiratory state value, It is represented as the i-th historical bronchial state value, H is represented as the respiratory state value, Q is represented as the bronchial morphology value, Expressed as breathing weight, It is expressed as bronchial weight to replace single abnormal compensation. Through threshold judgment and historical period review, it can identify and correct single abnormal compensation to prevent sudden large deviations in lesion positioning caused by erroneous compensation. The introduction time is dynamically allocated according to the excess value, so that the length of the historical period is proportional to the degree of compensation abnormality. It can more specifically trace back the data of key moments, refer to the physiological characteristics of multiple patients at the same time in the historical period, learn from the group rules, and improve the error correction ability of patients' abnormal conditions.

[0035] In a preferred embodiment, the steps of obtaining an overlap correction strategy based on lesion offset compensation, generating a corrected final overlap image based on the overlap correction strategy, and locating the lesion position based on the final overlap image include: S601, obtaining an overlap correction table, wherein the overlap correction table includes a plurality of lesion offset compensation intervals and an overlap correction strategy corresponding to each lesion offset compensation interval; S602, obtaining a corresponding overlap correction strategy from an overlap correction table according to the lesion offset compensation interval corresponding to the lesion offset compensation; S603, adjusting the lesion position in the initial overlapping image based on the overlap correction strategy, and generating a corrected final overlapping image; S604: Locate the lesion position based on the anatomical landmarks of the final overlapping image.

[0036] As in steps S601 to S604 above, an overlap correction strategy mapping table corresponding to the lesion offset compensation interval is constructed in advance based on a large amount of clinical and simulation data. The table divides offset compensation of different magnitudes (such as 0-1mm, 1-3mm, 3-5mm, and greater than 5mm) into several intervals, and defines corresponding overlap correction operations for each interval, including translation correction (small adjustments in the X, Y axis or depth direction), rotation correction (fine-tuning the projection angle to match the local anatomical direction), and scaling correction (locally enlarging or reducing the lesion marked area to match the actual size of the tissue). The lesion offset compensation is matched with the interval in the table, and the corresponding overlap correction strategy entry is quickly located, and the parameters of the strategy (such as translation amount, rotation angle, scaling factor) are read. etc.), based on the obtained strategy parameters, the lesion projection layer in the initial overlapping image is geometrically transformed, the semi-transparent and pseudo-color rendering properties of the overlay layer are maintained, the distinguishability of the real tissue and the annotation is guaranteed, and the final corrected overlapping image is output. Its lesion annotation should be highly consistent with the target tissue position. The position of the anatomical landmark point is re-identified on the final overlapping image, and the relative position of the landmark point and the corrected lesion projection is combined to obtain the three-dimensional coordinates of the lesion in the real lung tissue. The coordinates are used as the target point for intraoperative navigation or instrument guidance (such as puncture needle trajectory). By partitioning the offset compensation, the most appropriate geometric correction method is adopted for different degrees of deviation, avoiding general adjustments, eliminating the inconsistency caused by projection and tissue deformation, and significantly improving the geometric consistency between the lesion projection and the real tissue.

[0037] Please see the attached Figure 2 As shown, the present invention further provides a lung tissue lesion location positioning system based on three-dimensional reconstruction, which is used in the above-mentioned lung tissue lesion location positioning method based on three-dimensional reconstruction, comprising: A virtual model module is used to obtain CT image data of the patient's lung tissue in a preset lung inflation state and construct a three-dimensional virtual model of the lung tissue based on the CT image data, including the location of the lesion and anatomical landmarks, where the anatomical landmark is the junction of the azygos vein and the superior vena cava; The inflation error module is used to obtain the patient's lung tissue image in real time and obtain the inflation error of the anatomical landmarks based on the preset lung inflation state; A ventilation control module is used to obtain a lung tissue ventilation control strategy based on the inflation error, and adjust the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; An overlapping construction module is used to obtain anatomical landmarks after adjusting to a preset lung inflation state, and to superimpose the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; an offset compensation module, configured to obtain local structural features around anatomical landmarks after adjustment to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and to obtain lesion offset compensation based on the local structural features; The lesion localization module obtains an overlap correction strategy based on lesion offset compensation, generates a corrected final overlap image based on the overlap correction strategy, and locates the lesion position based on the final overlap image.

[0038] The above-mentioned virtual model module collects high-resolution CT sequences of patients under preset lung inflation states (such as maximum inspiration or constant tidal volume), extracts lung parenchyma and airway structure, and reconstructs a gridded three-dimensional virtual lung model. The three-dimensional coordinates of the center point of the lesion volume and the anatomical landmark point (the junction of the azygos vein and the superior vena cava) are accurately marked in the model. The expansion error module uses a real-time ultrasound or optical tracking system to collect the current lung image, automatically detect the anatomical landmark points in the real-time image, and align their coordinates with the coordinates of the corresponding landmark points in the model. The difference between the two coordinates is the expansion error, which is used to quantify the current lung The expansion state deviates from the pre-operative preset state to a certain extent and direction. The ventilation control module searches for the established error range and ventilation parameter adjustment mapping table based on the size of the expansion error, sends the selected tidal volume, PEEP, respiratory rate and other adjustment strategies to the ventilator or ventilation system, automatically adjusts the patient's alveolar inflation degree, continuously monitors the error and repeats the table lookup until the expansion error converges to an acceptable range, overlaps the construction module, and after the ventilation control is completed, identifies the anatomical landmark coordinates in the real-time image again. Based on the landmark coordinates, the three-dimensional virtual model as a whole or the lesion area is aligned with the real-time image. Coordinate system rigid registration, projecting the lesion position in the model onto the real-time image in a semi-transparent or pseudo-color form to generate a preliminary overlapping view, offset compensation module, extracting respiratory state values ​​(such as local volume change rate) and bronchial morphology values ​​(such as branch curvature) in the area around the landmark point, combining respiratory and bronchial features according to preset weights, generating lesion offset compensation, which is used to correct position errors caused by local tissue deformation, lesion positioning module, searching for compensation intervals and geometric transformation mapping tables based on the size of lesion offset compensation, obtaining translation, rotation, scaling and other correction parameters, and micro-scaling the initial superimposed layer. Adjust and output the final overlapping image after correction. In the final image, the relative coordinates of the lesion projection and the anatomical landmarks are read, and its three-dimensional absolute position in the patient's body is calculated for puncture, biopsy or ablation navigation. Preoperative overall alignment, intraoperative ventilation closed-loop adjustment and intraoperative offset compensation are three-level corrections to ensure that the lesion projection is consistent with the real tissue. From error detection to ventilation control to superposition correction, the entire process is automated and repeatable, adapting to the patient's respiratory fluctuations in real time. Accurate positioning can reduce complications caused by blind exploration and multiple punctures, shorten operation time, and reduce patient radiation and anesthesia risks.

[0039] And, a lung tissue lesion location positioning terminal based on three-dimensional reconstruction, comprising: one or more processors; a storage device having one or more programs stored thereon; When one or more programs are executed by one or more processors, the one or more processors implement a lung tissue lesion location positioning method based on three-dimensional reconstruction.

[0040] The foregoing is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained herein shall, unless otherwise specified or limited, be implemented in accordance with conventional means in the art.

Claims

1. A method for locating lung tissue lesions based on three-dimensional reconstruction, characterized in that: include: Obtaining CT image data of the patient's lung tissue at a preset lung inflation state, and constructing a three-dimensional virtual model of the lung tissue based on the CT image data, including the location of the lesion and anatomical landmarks, where the anatomical landmark is the junction of the azygos vein and the superior vena cava; Acquire the patient's lung tissue image in real time and, combined with the preset lung inflation state, obtain the inflation error of the anatomical landmarks; obtaining a lung tissue ventilation control strategy based on the inflation error, and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; Obtaining anatomical landmarks after adjusting to a preset lung inflation state, superimposing the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; Acquiring local structural features around anatomical landmarks after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtaining lesion offset compensation based on the local structural features; An overlap correction strategy is obtained based on lesion offset compensation, a corrected final overlap image is generated based on the overlap correction strategy, and the lesion position is located based on the final overlap image.

2. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 1, characterized in that: The steps of acquiring a patient's lung tissue image in real time and obtaining the inflation error of an anatomical landmark based on a preset lung inflation state include: Acquire the patient's lung tissue images in real time, and obtain real-time anatomical landmark information based on the lung tissue images; Acquiring real-time anatomical landmark spatial coordinates based on real-time anatomical landmark information; Acquiring preset anatomical landmark information based on a preset lung inflation state; Acquiring spatial coordinates of preset anatomical landmark points based on preset anatomical landmark information; The deviation of the anatomical landmark point is obtained according to the real-time anatomical landmark point spatial coordinates and the preset anatomical landmark point spatial coordinates, and is marked as an expansion error.

3. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 1, characterized in that: The steps of obtaining a lung tissue ventilation control strategy based on the inflation error and adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy include: Acquiring a ventilation control table, wherein the ventilation control table includes a plurality of inflation error intervals and a lung tissue ventilation control strategy corresponding to each inflation error interval; Obtaining a corresponding lung tissue ventilation control strategy from a ventilation control table according to an inflation error interval corresponding to the inflation error; Based on the lung tissue ventilation control strategy, the patient's lung tissue is adjusted to the preset lung inflation state.

4. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 3, characterized in that: The steps of adjusting the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy include: Regulate the patient's lung tissue inflation state according to the lung tissue ventilation control strategy; obtaining an adjusted lung tissue inflation state of the patient's lung tissue, and obtaining an adjusted anatomical landmark deviation based on the adjusted lung tissue inflation state; Obtaining an adjusted anatomical landmark deviation threshold, and determining whether the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; If the adjusted anatomical landmark deviation does not exceed the adjusted anatomical landmark deviation threshold, it is determined that the patient's lung tissue inflation state meets the preset lung inflation state; If the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold, determining that the patient's lung tissue inflation state does not meet the preset lung inflation state, and obtaining an anatomical landmark excess value at which the adjusted anatomical landmark deviation exceeds the adjusted anatomical landmark deviation threshold; Acquiring a ventilation fine-tuning table, wherein the ventilation fine-tuning table includes a plurality of anatomical landmark point excess value intervals and a ventilation fine-tuning strategy corresponding to each anatomical excess value interval; Obtaining a corresponding ventilation fine-tuning strategy from a ventilation fine-tuning table according to an anatomical landmark point excess value interval corresponding to the anatomical landmark point excess value; Fine-tune the patient's lung tissue inflation status based on ventilation fine-tuning strategies; obtaining a lung tissue inflation state of the patient after fine-tuning the lung tissue, and obtaining a fine-tuned anatomical landmark deviation based on the fine-tuned lung tissue inflation state; Determine whether the deviation of the anatomical landmarks after fine-tuning exceeds the preset conditions for fine-tuning. If not, the fine-tuning is determined to be normal. If exceeded, the fine-tuning is determined to be abnormal, and the lung tissue ventilation control strategy is re-acquired.

5. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 1, characterized in that: The steps of obtaining anatomical landmarks adjusted to a preset lung inflation state and superimposing the lesion position in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image include: Obtain the lesion location and anatomical landmarks in the patient's 3D virtual model of lung tissue; Obtaining anatomical landmarks of the lung tissue image after the patient's lung tissue is adjusted to a preset lung inflation state; Based on the anatomical landmarks of the lung tissue image after adjustment to a preset lung inflation state and the anatomical landmarks in the three-dimensional virtual model of lung tissue, the lesion position in the three-dimensional virtual model of lung tissue is superimposed on the lung tissue image to generate an initial overlapping image.

6. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 1, characterized in that: Acquiring local structural features around the anatomical landmark after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and obtaining lesion offset compensation based on the local structural features, including: Acquiring local structural features around anatomical landmarks in a lung tissue image after adjusting to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology; Obtaining corresponding respiratory state values ​​and bronchial morphology values ​​based on the respiratory state and bronchial morphology respectively; Obtain corresponding respiratory weight and bronchial weight based on respiratory state and bronchial morphology respectively; Lesion offset compensation is obtained based on respiratory state value, bronchial morphology value, respiratory weight and bronchial weight.

7. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 6, characterized in that: After the step of obtaining lesion offset compensation based on the respiratory state value, the bronchial morphology value, the respiratory weight, and the bronchial weight, the method further includes: Obtaining a lesion offset compensation threshold, and determining whether the lesion offset compensation exceeds the lesion offset compensation threshold; If the lesion offset compensation exceeds the lesion offset compensation threshold, the lesion offset compensation is determined to be abnormal, and the time node when the compensation is determined to be abnormal is obtained and marked as the end time of the historical period; If the lesion offset compensation does not exceed the lesion offset compensation threshold, it is determined that the lesion offset compensation is normal; Obtaining a lesion offset excess value where the lesion offset compensation exceeds a lesion offset compensation threshold, and obtaining a corresponding lead-in duration based on the lesion offset excess value; Construct historical periods based on the introduction duration and the end time of the historical period; Acquire lung tissue images of multiple patients within a historical period, and acquire historical respiratory states and historical bronchial morphologies in the lung tissue images of each patient, and respectively acquire historical respiratory state values ​​and historical bronchial morphology values ​​corresponding to each historical respiratory state and historical bronchial morphology; A new lesion offset compensation is obtained based on a plurality of historical respiratory state values, a plurality of historical bronchial morphology values, a respiratory state value, a bronchial morphology value, a respiratory weight, and a bronchial weight.

8. The method for locating lung tissue lesions based on three-dimensional reconstruction according to claim 1, characterized in that: The steps of obtaining an overlap correction strategy based on lesion offset compensation, generating a corrected final overlap image based on the overlap correction strategy, and locating the lesion position based on the final overlap image include: Obtaining an overlap correction table, wherein the overlap correction table includes a plurality of lesion offset compensation intervals and an overlap correction strategy corresponding to each lesion offset compensation interval; Obtaining a corresponding overlap correction strategy from the overlap correction table according to the lesion offset compensation interval corresponding to the lesion offset compensation; Adjusting the lesion position in the initial overlapping image based on the overlap correction strategy and generating a corrected final overlapping image; The lesion location was determined based on anatomical landmarks in the final overlapping images.

9. A system for locating lung tissue lesions based on three-dimensional reconstruction, applied to the method for locating lung tissue lesions based on three-dimensional reconstruction according to any one of claims 1 to 8, characterized in that: include: A virtual model module is used to obtain CT image data of the patient's lung tissue in a preset lung inflation state and construct a three-dimensional virtual model of the lung tissue based on the CT image data, including the location of the lesion and anatomical landmarks, where the anatomical landmark is the junction of the azygos vein and the superior vena cava; The inflation error module is used to obtain the patient's lung tissue image in real time and obtain the inflation error of the anatomical landmarks based on the preset lung inflation state; A ventilation control module is used to obtain a lung tissue ventilation control strategy based on the inflation error, and adjust the patient's lung tissue to a preset lung inflation state based on the lung tissue ventilation control strategy; An overlapping construction module is used to obtain anatomical landmarks after adjusting to a preset lung inflation state, and to superimpose the lesion location in the three-dimensional virtual model of lung tissue onto the lung tissue image to obtain an initial overlapping image; an offset compensation module, configured to obtain local structural features around anatomical landmarks after adjustment to a preset lung inflation state, wherein the local structural features include respiratory state and bronchial morphology, and to obtain lesion offset compensation based on the local structural features; The lesion localization module obtains an overlap correction strategy based on lesion offset compensation, generates a corrected final overlap image based on the overlap correction strategy, and locates the lesion position based on the final overlap image.

10. A lung tissue lesion location positioning terminal based on three-dimensional reconstruction, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon; When one or more programs are executed by one or more processors, the one or more processors implement the method for locating lung tissue lesions based on three-dimensional reconstruction as described in any one of claims 1 to 8.

Images