Method for optimizing infiltrating boundary resection strategy of meningioma

Abstract

The present application relates to a meningioma infiltration boundary resection strategy optimization method, which uses the stress and strain response difference during intraoperative tissue traction and suction operation to mark the biomechanical abnormal area in space; constructs a mechanical residual atlas; extracts the local tissue compliance critical point as the physiological landmark of the infiltration boundary; identifies the high adaptability channel in the three-dimensional tissue atlas; makes partition modeling for the structural impedance of different regions; constructs a boundary evolution stability hot area map, maps the boundary prediction result back to the current operation scene, and is used to judge whether the resection range needs to be expanded; after each tissue stripping or resection, the corresponding regional tissue deformation rate, blood supply flow rate and temperature diffusion mode indicators are collected as the reflection of the intraoperative boundary stress response, which is used to evaluate whether the high-risk area has been touched by the resection behavior; once the boundary stress index exceeds the preset threshold, the system freezes the path segment, calculates the alternative path, improves the accuracy of boundary identification, and reduces the damage to the functional area and the recurrence risk.

Description

Technical Field

[0001] This invention relates to the resection of the infiltrative boundary of meningioma, specifically to an optimized method for the resection of the infiltrative boundary of meningioma. Background Technology

[0002] Currently, in optimizing resection strategies for meningiomas, although software systems and analysis methods based on MRI images, such as those described in Chinese patent CN118429266A, provide technical means for assisting diagnosis and treatment path planning, these methods still have several limitations and drawbacks. Especially in actual clinical practice, they often fail to meet the high requirements for precision and safety in meningioma surgery. Firstly, while existing software systems, based on AI segmentation and 3D reconstruction algorithms, achieve rapid tumor processing and 3D reconstruction, providing doctors with a 3D model of the tumor in the brain and helping them quickly understand the tumor's morphology and location, thereby optimizing diagnostic time and decision-making processes, this method focuses more on automated image processing and fails to effectively address the issues of accurate identification of tumor invasion boundaries and dynamic adjustments during surgery. A significant deficiency in existing technologies is that relying solely on image data for tumor segmentation and 3D reconstruction lacks a detailed assessment of subtle structural changes between tumor tissue and normal brain tissue. For example, during meningioma resection, the tumor boundaries can be very indistinct, especially when tumor cells infiltrate normal brain tissue. Traditional imaging techniques struggle to effectively identify tumor boundaries and infiltrated areas, particularly in invasive tumors that are adjacent to and significantly impact surrounding structures. In such cases, the tumor boundaries generated by AI segmentation algorithms are often inaccurate, potentially leading to misjudgments or omissions, which in turn affect decisions regarding the extent of surgical resection.

[0003] Furthermore, while the software system provides data support for tumor morphological analysis through AI algorithms and 3D reconstruction technology, it relies on static images and lacks real-time physiological feedback and dynamic adjustment mechanisms. It cannot optimize and adjust based on real-time tissue responses and boundary changes during surgery, thus failing to effectively adapt to the actual tumor invasion patterns in dynamic and complex surgical environments. Traditional imaging methods and algorithms, such as MRI-based image fusion and tumor segmentation, rely on pre-collected image data and cannot capture the behavioral characteristics of the tumor during surgery, especially the possible displacement and deformation of the tumor boundary and the physiological responses of the tissue. Moreover, although the system's tumor segmentation and image fusion modules provide surgeons with morphological features of the tumor, they lack detailed modeling of tissue physical properties, such as tissue compliance, elasticity, and structural consistency. These factors are crucial for optimizing resection strategies and making decisions during actual surgery. Therefore, although the system provides basic support for tumor diagnosis and analysis, it still has significant shortcomings in the personalized formulation and precise implementation of resection strategies. In addition, existing imaging methods cannot effectively handle the subtle structural changes between the tumor infiltration area and normal tissue, and cannot dynamically assess the physiological changes of tumor invasion. In summary, although existing technologies such as CN118429266A provide automated tools for tumor image processing and improve the efficiency of image segmentation and 3D reconstruction, they still have significant limitations in tumor boundary recognition, invasion path prediction, and real-time surgical adjustments. Summary of the Invention

[0004] The purpose of this invention is to provide an optimized resection strategy for meningioma invasion boundaries, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0005] The present invention addresses the aforementioned technical problems by employing the following technical solution: a method for optimizing the resection strategy of meningioma infiltration boundary, comprising: marking biomechanically abnormal areas in space by utilizing the differences in stress and strain response during intraoperative tissue traction and suction operations; constructing a mechanical residual map to compare the anisotropic region around the tumor with the normal meningeal layer; and extracting the local tissue compliance critical point as a physiological landmark of the infiltration boundary.

[0006] In a three-dimensional tissue atlas, highly adaptive channels including intervascular spaces, submembrane tunnels, and white matter fibrous gaps are identified; the structural impedance of different regions is modeled in a partitioned manner, and the infiltration accessibility in each direction is calculated; a boundary evolution stability heat map is constructed, and the boundary prediction results are mapped back to the current surgical scene to determine whether the resection range needs to be expanded.

[0007] After each tissue stripping or resection, the corresponding tissue deformation rate, blood flow velocity, and temperature diffusion pattern indicators are collected as a reflection of the intraoperative boundary stress response to assess whether the resection has reached a high-risk area. Once the boundary stress indicators exceed the preset threshold, the system freezes the path segment and calculates feasible alternative paths.

[0008] Furthermore, the method for extracting the local tissue compliance critical point includes the following steps:

[0009] During the operation, force feedback devices are used to collect tissue mechanical response data of the meningioma and its surrounding tissues under traction, suction and compression operations; based on the mechanical response data, a three-dimensional spatial distribution mechanical residual map is constructed to express the stress and strain differences of different regions under external force.

[0010] Anisotropic regions are identified in the mechanical residual map, and the response characteristics of normal meningeal tissue and tumor edge tissue are compared. Based on strain mutation points or compliance mutation patterns, local tissue compliance critical points representing the tumor invasion boundary are extracted as real physiological landmarks during surgery. These physiological landmarks are used to guide the determination of the resection range and path of meningioma, thereby controlling the invasion boundary.

[0011] Specifically: During the operation, the micro-force and displacement response data of the tumor edge and surrounding tissue under traction and suction surgical actions are continuously collected by the stress feedback device, and a mechanical residual map under three-dimensional spatial structure is generated. This map reflects the heterogeneous deformation response characteristics of each region under the same external force conditions. By comparing the response changes of the tumor edge and normal meningeal tissue in the residual map, the spatial regions of anisotropy of tissue structure are identified.

[0012] A non-publicly known structural mutation identification function is introduced to extract the key locations where tissues transition from structural compliance to pathological infiltration, i.e., the local compliance critical point. The following higher-order function expression is introduced:

[0013]

[0014] in:

[0015] Ξ(x,y,z): Represents the compliance rate of change function at the point (x,y,z) in three-dimensional space, i.e., whether the point is in an abrupt transition zone; Λ(x,y,z): The local residual response amplitude of the tissue at this point; Φ(θ): The anisotropy weighting coefficient of this point under the applied force direction θ, determined by the directionality of the local tissue structure arrangement; δ: The internal damping coefficient of the tissue, representing the dissipation capacity of the local microstructure; κ(x,y,z): Represents the recovery curvature characteristic factor of the tissue at this point, used to assess the degree of elastic recovery after deformation; ∈: The amplitude of unit traction disturbance, used as the reference variable for the derivative of this function;

[0016] When the value of the function Ξ(x,y,z) suddenly reaches a local maximum, it indicates that the tissue exhibits discontinuous compliance behavior in that region, i.e., a structural abrupt change point. The function is then scanned across the entire mechanical residual map space to locate a set of spatial points (x...y...z). i ,y i ,z i This set of local tissue compliance thresholds constitutes the actual physiological boundary markers of tumor invasion during surgery, which can replace or supplement traditional imaging boundaries.

[0017] Furthermore, the tissue mechanical response data includes displacement amplitude, rebound time, traction recovery ratio, and tissue compressive strength curve parameters; the mechanical residual map is obtained by constructing the difference between the actual deformation and the predicted model deformation of each unit area under standard traction strength.

[0018] Furthermore, the local tissue compliance critical point is the spatial point in the residual map where a nonlinear transition occurs, and the transition reflects the transformation from a physiologically connected structure to a pathologically relaxed structure.

[0019] Furthermore, the physiological landmarks are visually superimposed on the navigation system display interface to guide the dynamic updating of the resection boundary; the force feedback acquisition device is a miniature stress sensing module integrated into the surgical suction head, micromanipulation arm, or intraoperative navigation probe.

[0020] Furthermore, the method for calculating the infiltration accessibility in each direction includes the following steps:

[0021] S1. Acquire preoperative multimodal medical image data of meningioma and its surrounding tissues to identify local tissue structure types; based on the microstructural characteristics of brain tissue, divide the tumor-peripheral tissue region into multiple structural impedance subregions, including: white matter fiber bundles, meninges, intervascular spaces, glial tissue, and sulci.

[0022] S2. Set the structural impedance level for each sub-region to form a structural impedance distribution map; starting from the tumor boundary, project paths along multiple spatial directions and calculate the cumulative impedance value of the area traversed by each path; generate a directional accessibility assessment heatmap based on the total cumulative impedance value of each path to identify potential preferred invasion paths of the tumor in different directions.

[0023] S3. Based on the accessibility assessment results, adjust the resection strategy and perform extended pre-resection or focused intraoperative monitoring on high-accessibility areas.

[0024] Furthermore, the structural impedance level is set based on physiological parameters such as tissue arrangement density, intercellular space size, permeability, and tissue continuity; the directional accessibility heatmap visualizes the relative accessibility in each direction in three dimensions, and uses color gradients or probability values ​​to indicate different risk levels.

[0025] Furthermore, the path accessibility assessment also includes a tissue direction consistency index, which is used to enhance the weight of the penetration path along the structural direction; the resection strategy optimization includes setting a variable boundary line in the intraoperative navigation system based on the directional accessibility results, so as to dynamically adjust the resection range in real time.

[0026] The beneficial effects of this invention: Traditional meningioma boundary identification mainly relies on imaging techniques such as MRI and CT scans, which often struggle to accurately determine the subtle infiltration areas of the tumor, especially when there is a blurred transition zone between the tumor and normal tissue. This invention constructs a three-dimensional mechanical residual map using real-time intraoperative tissue mechanical response data (such as displacement amplitude, rebound time, traction recovery ratio, and tissue compressive strength curve parameters), effectively identifying the true boundaries of tumor infiltration, filling imaging blind spots, and ensuring precise boundary localization. The directional accessibility assessment heatmap of this invention can quantify the possible infiltration paths of tumor cells based on the tissue's impedance level and the continuity of its microstructure, providing quantitative analysis based on tissue physical properties. This analysis result can be superimposed in real-time on the intraoperative navigation system, allowing surgeons to dynamically adjust the resection range according to the patient's specific condition, achieving personalized resection strategies, and enabling fine-tuning of the resection boundary during surgery based on real-time feedback, avoiding over-resection or incomplete resection.

[0027] By precisely modeling and dynamically monitoring the tissue structure in different regions, surgeons can identify which areas represent high-risk pathways for tumor expansion and which are low-risk areas, thus enabling them to adopt more precise resection paths. This not only reduces the risk of accidentally damaging brain functional areas but also maximizes the thoroughness of resection and lowers the probability of postoperative recurrence. By combining a biomechanical feedback device with a real-time navigation system, tissue response data can be acquired in real time during the surgery, and the system's dynamic feedback helps surgeons make decisions. During the surgery, if a resection path enters a high-risk area, the system will automatically update the boundaries and remind the surgeon to adjust the resection path through a variable boundary line locking function, effectively avoiding postoperative tumor residue or functional damage. Attached Figure Description

[0028] Figure 1 This is a flowchart of the method for optimizing the resection strategy of meningioma invasion boundary according to the present invention.

[0029] Figure 2 This is a diagram showing the relationship between brain tissue microstructure access pathways and tumor infiltration prediction function according to the present invention.

[0030] Figure 3 This is a graph showing the changes in intraoperative stress response and risk threshold during the procedure according to the present invention.

[0031] Figure 4 This is a visualization of the dynamic path adjustment and invasion boundary during meningioma resection according to an embodiment of the present invention.

[0032] Figure 5 This is a diagram illustrating the dynamic adjustment and accessibility assessment of the surgical resection path for meningiomas according to an embodiment of the present invention. Detailed Implementation

[0033] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0034] Combined with appendix Figure 1This invention provides an optimized resection strategy for meningioma infiltration boundaries. It constructs a mechanical characteristic spatial map using real-time intraoperative tissue physical response data to identify the physiological location of the infiltration boundary, guiding the development of personalized resection paths. During the operation, standardized traction and suction operations are performed on the meningioma margin area and surrounding normal tissue using surgical instruments (such as traction hooks, suction heads, or navigation probes). The stress and strain responses generated by the tissue during the operation are recorded simultaneously. The response data includes various physical parameters such as local displacement, rebound velocity, strain recovery curve, and deformation resistance. Based on this, a corresponding three-dimensional spatial coordinate index is established. Then, based on the deviations in stress-strain relationships exhibited by different tissue regions under the same applied force, a mechanical residual map covering the tumor and surrounding tissues is constructed. This map represents the differences in the responses of different tissue regions under the same mechanical input conditions, reflecting their structural heterogeneity and potential pathological changes. In this atlas, the system further compares the anisotropic response differences between the tumor margin region and the distal normal meningeal tissue, analyzing the differences in tissue deformation and energy dissipation characteristics induced by a unit force in the same direction. Through spatial statistics and continuity analysis, it identifies regions where the tissue response exhibits nonlinear jumps under a certain mechanical stimulation threshold. In these regions, the tissue typically transitions from structural compliance (i.e., exhibiting healthy connectivity) to pathological compliance (i.e., losing the complete structural barrier), and this aberrant behavior is extracted as the compliance threshold. The system extracts the distribution of these points in three-dimensional space in the form of a function and marks them as physiological landmarks of the invasion boundary. That is, this set of points truly represents the boundary zone of tumor microinvasion, independent of imaging or staining, but reflects the changes in the structural behavior of tissue under real physical perturbation. Finally, based on the spatial morphology presented by this physiological landmark layer, combined with the tumor shape, the distribution of surrounding functional areas, and the principle of nerve protection, a precise and individualized resection boundary plan is formulated and visualized and integrated into the intraoperative navigation system. This guides the surgeon to dynamically adjust the resection strategy in real time according to changes in tissue behavior, thereby achieving the goal of maximizing complete tumor resection while effectively avoiding the risk of recurrence caused by unnecessary damage to functional areas or insufficient resection.

[0035] Combined with appendix Figure 2By identifying and modeling the access pathways of brain tissue microstructures, the potential future invasion direction of tumors and the dynamic adjustment of resection range can be achieved. High-resolution multimodal imaging data, including T1-enhanced MRI, T2-FLAIR, and DTI sequences, are acquired preoperatively. Combined with existing anatomical atlases, three-dimensional reconstruction of the patient's brain tissue is performed, constructing a three-dimensional tissue structure atlas including vascular channels, submural structures, and white matter fiber gaps. Microchannel structures with high permeability and low tissue resistance are identified as highly adaptive channels, which are often preferred pathways for meningioma cell infiltration and migration. Subsequently, based on this three-dimensional atlas, structural impedance zoning modeling is performed on the tumor periphery according to tissue type (e.g., dense white matter, intercellular spaces, glial tissue, meningeal folds, etc.). Each tissue unit is assigned a structural impedance level based on its microstructure density, intercellular space, tissue elasticity, and fluid permeability, and further refined according to different directions. The total impedance level traversed is used to calculate the invasive accessibility of tumor cells from the tumor edge to each unit space in all directions, i.e., the degree of structural accessibility of tumor cells in each direction. The lower the value, the easier it is to penetrate. After completing the accessibility calculation in all directions, the results are mapped into a spatial heat map, called the boundary evolution stability heat map, which is used to express which areas are more likely to become breakthrough points or microinvasive development directions of meningiomas in the future. High-risk channels in this map will be marked in the form of high heat values. During the operation, this map can be fused and superimposed with the real-time navigation system, reflecting the prediction results back to the current surgical scene. This allows the surgeon to clearly understand the current visible boundary and obtain spatial judgment on the potential invasion direction, thereby assisting in the decision-making on whether to appropriately expand the resection range in a specific area to intervene in advance on the basis of recurrence in high-risk directions.

[0036] Combined with appendix Figure 3By monitoring the physiological and physical stress responses of tissues in real time during surgery, the system dynamically identifies high-risk areas and adjusts the resection path to improve the safety and precision of the resection. During meningioma surgery, after each tissue dissection or partial tumor resection, the system uses an integrated intraoperative monitoring module to collect multi-dimensional data on the local tissue status of the operated area, including key physiological parameters such as tissue deformation rate, changes in local blood flow velocity, and temperature diffusion patterns in the surface or deep tissue layers. Deformation rate is used to assess the integrity of the tissue structure and the degree of tension release; changes in blood flow velocity reflect whether the area involves important blood vessels or perfusion-sensitive areas; and temperature diffusion patterns indirectly determine microcirculatory stability and tissue activity. Through joint analysis of these three parameters, the system constructs an intraoperative boundary stress response index group to assess in real time whether the resection has approached or invaded functionally risky areas or areas susceptible to potential recurrence. If the system monitors... If any one or more stress indicators exceed the pre-set safety threshold, such as a sudden decrease in tissue deformation rate, a sudden reduction in blood supply, or abnormal concentration of heat diffusion, the system will determine that the resection path segment is in a high-risk boundary state and trigger the path freezing mechanism. This means suspending further resection operations in that area and calling the alternative resection path database established in the pre-operative model or recalculating a new feasible resection path based on the remaining anatomical structure. The alternative path will avoid the high-risk area and re-optimize the approach angle, resection sequence, and boundary advancement direction. The system will display the new path in real time on the navigation platform interface for the surgeon's reference and decision-making, thereby realizing a closed-loop control mechanism of intraoperative dynamic perception, risk identification, and path reconstruction, ensuring that tumor tissue is removed to the maximum extent while avoiding unnecessary tissue damage during the operation.

[0037] Example 1:

[0038] Combined with appendix Figure 4In this embodiment, during a right frontal lobe meningioma resection surgery, the patient was a 56-year-old female. Preoperative MRI showed that the tumor was located near the right middle frontal gyrus, with significant adhesion to the surrounding cortex. T1-weighted contrast enhancement showed blurred tumor boundaries and heterogeneous signal in some areas. The surgeon initially diagnosed it as an invasive meningioma and determined that the extent of resection needed to be assessed intraoperatively to avoid recurrence. Using the force feedback module in the integrated intraoperative micromanipulation arm, three standard traction forces of 0.8N, 1.2N, and 1.5N were applied to pull the tumor and adjacent meningeal regions. The micro-displacement response of the tissue under these forces was recorded at a sampling frequency of 100Hz and a recording period of 5 seconds. After data processing, the displacement response matrix of the tumor edge and adjacent white matter tissue points was obtained, and then the residual response amplitude Λ(x,y,z) of each point was calculated. In a sample region P (coordinate range 2.3≤x≤2.7, 3.1≤y≤3.4, 1.5≤z≤1.9), the residual amplitude was found to be about 0.43 mm in the tumor region and 0.09 mm in the normal meningeal region, indicating that the structural elastic properties of this region are significantly different.

[0039] The system then introduced a structural consistency factor in the local direction, obtaining the anisotropy weight coefficient Φ(θ) for each point, where θ is the angle between the surgical approach and the direction of the main fiber bundle, with a weight range of [0.5, 1.8], where a higher weight indicates that the tumor is more likely to slide or expand along that direction. The tissue structure damping coefficient δ was set, based on intraoperative acoustic impedance data and a stress attenuation model; in this case, this coefficient ranged from 0.3 to 0.7. The tissue recovery curvature characteristic factor κ(x,y,z) was calculated based on the elastic recovery rate within 1 second after stretching, and was approximately 0.2–0.35 at the tumor edge and 0.65–0.85 in the normal tissue area.

[0040] Substitute the collected data into the non-public function model:

[0041]

[0042] Taking region P as an example, let's take a point P0 with the following parameters: Λ = 0.42, Φ(θ) = 1.5, δ = 0.5, κ = 0.28, and let the unit disturbance amplitude ∈ be 0.05 (the perturbation control value). Then we have:

[0043]

[0044] Under perturbation control conditions, this value is insensitive to ε, indicating that this point still falls within the structural compliance continuity region. However, at the neighboring point P1, Λ=0.51, Φ(θ)=1.7, κ=0.21, after substituting:

[0045]

[0046] Because multiple neighboring points form a sudden step in space, the system uses continuous function scanning to detect areas where the mutation intensity exceeds 5 times the average background (in this procedure, the mutation rate is set to ΔΞ≥0.15 / mm). This area is then identified as a set of local compliance critical points, with spatial coordinates defined as (x...). i ,y i ,z i This indicates that the system uses it as a physiological marker of the true infiltration boundary.

[0047] Subsequently, the set of landmarks was mapped back to the intraoperative navigation system and superimposed on the preoperative MRI enhancement layer. The surgeon found that the area where the boundary determined by the original MRI extended less than 1 cm to the right frontal gyrus actually contained multiple compliance mutation points. After further dissection during surgery, the tissue in this area showed a state of loosening without rebound and thin blood supply. Finally, it was determined that the surgical resection path should be extended 0.8 cm in this direction to completely remove the tumor residue in the boundary area.

[0048] Three months postoperative follow-up showed no recurrence signal on enhanced MRI, and the patient's neurological function score was normal with no language or motor impairment. This demonstrates that the method can effectively identify potential infiltration areas, prevent postoperative recurrence, control the resection range, and ensure the safety of functional areas.

[0049] After completing the first stage of identifying the compliance threshold in the tumor margin region, a more refined mechanical residual map was constructed using a tissue mechanical response analysis mechanism to clarify the true stress state of the internal structure of the boundary. Four types of tissue response parameters were introduced, including displacement amplitude, rebound time, traction recovery ratio, and morphological characteristics of the compression curve, to comprehensively reflect the elastic compliance and structural resistance of each unit voxel region. During the procedure, guided by the navigation system, the surgeon applied mild traction to ten suspected infiltration points along the tumor margin extension direction. Each application of the standard traction force was 1.0 N, lasting for 3 seconds, followed by immediate release to record the passive recovery behavior of the tissue. The acquisition frequency was set to 200 Hz. The system recorded the displacement amplitude of each point at the moment of maximum traction, ranging from 0.21 mm to 0.55 mm. The average displacement in the tumor core region was 0.48 mm, while the distal normal meningeal layer only showed a significant difference of 0.16 mm. Further analysis of the rebound time, i.e., the time required for tissue to recover from its maximum deformation state to 80% of its original shape, revealed that the rebound time in the tumor periphery was generally between 1.2 and 2.1 seconds, while normal meningeal tissue typically completed elastic recovery in 0.4 to 0.6 seconds, indicating that tumor tissue exhibits significant elastic hysteresis in this region. The traction recovery ratio (i.e., displacement recovery ratio) was also calculated simultaneously, with an average recovery ratio of approximately 0.63 at the tumor boundary, significantly lower than the 0.89 in the meningeal-controlled area, suggesting the presence of microstructural fractures, adhesions, or pathological remodeling in this region. Finally, combined with tissue compressive strength curve analysis, the system analyzed the load-displacement response curves at each point in a semi-modal manner. Significantly early inflection points of nonlinear segments were found in the boundary region, indicating that the tissue exhibited yielding behavior at lower pressure levels. The points where the curve slope decreased were basically consistent with the signal transition areas on MRI, but with a clearer distribution.

[0050] Based on the above four types of data, a complete three-dimensional mechanical residual map was constructed. This involved calculating the deviation between the actual tissue response and the standard tissue model response under standard traction conditions for each spatial unit region, and presenting this as a color heatmap on the entire map. Red areas represent areas with significant abnormal responses, while blue areas represent areas with normal structures. In this patient, the region with the highest residual values ​​was concentrated 1.2 cm below the right middle frontal gyrus cortex. This area only showed mild T2-prolonged signal on preoperative imaging and was not considered a priority resection area. Referring to the residual map, the surgeon identified this area as a latent infiltration hotspot and performed traction again in this area for verification. Abnormal tissue deformation was observed with no significant rebound. Therefore, the resection path was extended by 5.5 mm in this direction, ultimately resulting in complete resection of the latent structural softening area. Postoperative pathology confirmed the presence of tumor cell extravasation and stromal infiltration in the extended area, confirming the accuracy of the residual map localization. One month postoperatively, MRI showed no local residual tissue, and functional assessment revealed no new neurological deficits.

[0051] After obtaining the dynamic mechanical response of the tissue under traction and suction operations using an intraoperative real-time force feedback and mapping system, the surgeon identified the local tissue compliance critical point as a key reference marker for the resection boundary. The surgeon focused on the moderately abnormal area marked in the previous residual map, located at a depth of 1.0 cm to 1.6 cm below the right frontal cortex, with spatial coordinates ranging from x = 2.4–2.8, y = 3.3–3.7, and z = 1.2–1.8. In this area, the mechanical residual value changed significantly. The surgeon further conducted a high-density micromanipulation traction test, applying a standard force of 0.9 N, and repeatedly collecting data at each point three times, with a 10-second interval between each acquisition to eliminate tissue fatigue interference. The data was processed in real time by the mapping system to generate local mechanical response profiles. In this cross-section, the surgeon noticed a set of continuous spatial points where the tissue displacement response curves no longer exhibited a standard linear recovery pattern. Instead, they showed inflection point lag and abrupt change segments. Specifically, the tissue, which should have produced approximately 0.25 mm displacement under 0.9 N traction, actually displaced 0.41 mm at these points, far exceeding the model prediction. Simultaneously, the rebound behavior was significantly delayed, with a recovery delay exceeding 2 seconds after traction cessation and a plateau segment appearing in the recovery curve. Further calculations of the deformation response rate and recovery offset rate in this area revealed that in the surrounding tissue, the deformation response rate (displacement / time) was 0.12-0.15 mm / s, and the recovery offset rate was <10%. However, in this abnormal area, the rate increased to 0.26 mm / s, and the recovery offset rate reached 31-44%. This indicates that the tissue in this area had lost its healthy connectivity and exhibited structural relaxation characteristics; that is, the physiological connectivity structure had transformed into a pathological compliance structure.

[0052] Based on the residual map calculation mechanism, these points were marked as nonlinear transition points. Their transition morphology, reflected in the spatial profile, is a transition wall where a continuous, stable residual band suddenly jumps to a highly variable region. The identification criteria set by the map system are: if the residual change between two adjacent points exceeds 2.5 times the average residual value and is accompanied by an increase in rebound time of more than 1 second, it is determined to be an compliant transition point and included in the critical point set. In this example, the system identified a total of 12 transition points, concentrated in the intersection of the anterior edge of the tumor and the subfrontal sulcus. Based on this, the surgeon adjusted the surgical strategy, no longer using the MRI signal enhancement boundary as the resection limit, but mapping the intraoperatively identified transition points to the three-dimensional visual interface through the navigation system, updating the resection path, including intraoperatively highly elastic abnormal points in the surgical field, and extending the resection range anteriorly by approximately 6 mm.

[0053] Postoperative pathological examination revealed extensive dissemination of meningioma cells in the area previously unenhanced on MRI, particularly with focal vascular entrapment and interstitial space expansion within this transitional region, further confirming that the transitional region was indeed the boundary of a genuine pathological infiltration. Simultaneously, the surgeon effectively avoided the originally planned mis-cut brain functional areas using this method, preventing postoperative neurological deficits.

[0054] After identifying multiple tissue compliance thresholds and potential microinvasive tumor areas using mechanical residual mapping and updating the resection strategy accordingly, the surgeon further utilizes navigation integration technology to visualize and overlay extracted physiological landmarks intraoperatively, ensuring that boundary identification results can be applied immediately in resection decisions. In this example, the acquired mechanical data comes from an integrated micro-force feedback acquisition system, which is implemented by a stress sensing module at the end of the intraoperative micromanipulation arm and simultaneously embedded in the suction head and the neuronavigation probe head. Its stress accuracy is ±0.05N, and the maximum resolution can reach 0.01N / μm per square millimeter. The surgeon does not need to change instruments when routinely pulling the tumor edge and exploring the degree of tissue loosening. All mechanical responses are transmitted in real time to the intraoperative navigation host via a wireless module. This host is equipped with a mechanical data analysis module, which can complete 1000-point residual analysis, rebound time comparison, and elastic modeling calculation within 5 seconds during surgery, and generate a three-dimensional structural mechanical field map.

[0055] In this surgery, the navigation system interface used augmented reality overlay technology to overlay identified compliance thresholds as red and yellow semi-transparent spheres onto the original preoperative MRI image. Red represented typical pathological structures with high mutation intensity and disrupted physiological connections; yellow indicated marginal regions with obvious structural variation trends but not yet fully mutated. This overlay view allowed the surgeon to customize transparency and layer priority, and could be displayed side-by-side with brain functional localization data (such as fMRI mapping of the language area or intraoperative electrical stimulation points). During the actual resection, the surgeon first planned the resection route based on the physiological landmark layers in the navigation system. When the microscopic findings matched the system's indicated locations, the red sphere area was prioritized for deep resection to avoid retaining tumor-containing remnants in structurally relaxed areas, ensuring maximum resection rate. For yellow sphere areas near the functional cortex or deep white matter tracts, a conservative strategy was adopted, supplemented by electrophysiological monitoring and cold light irradiation for discrimination. If functional activity signals were unclear, some boundaries were preserved, and postoperative radiotherapy was used for control.

[0056] Approximately 90 minutes into the procedure, the navigation system updated the atlas in real time based on newly acquired data. By recalculating the distribution of compliance transition zones, it identified a new mutation point that was not previously visible on the preoperative enhanced MRI but exhibited significantly abnormal mechanical behavior. This mutation point was located 5.2 mm downwards from the end of the original resection path. The navigation interface then overlaid a new spherical marker on the area and displayed a risk warning. After review, the surgeon decided to extend the resection range by 0.6 cm to completely remove the abnormal zone and avoid the risk of postoperative recurrence.

[0057] Postoperative MRI and enhanced sequence examination after the operation showed that the tumor resection margin was highly consistent with the physiological landmark map generated by the intraoperative navigation system, with no obvious residue. The patient recovered well after the operation, with no speech dysfunction or motor function decline. The 3-month follow-up showed that the tumor did not recur and there was no abnormal enhancement in the surgical area.

[0058] Example 2:

[0059] Combined with appendix Figure 5 Based on Example 1, after completing the compliance threshold identification and mechanical boundary visualization overlay in the aforementioned right middle frontal gyrus meningioma surgery case, the surgeon continued to apply the directional invasion accessibility assessment mechanism to dynamically optimize the remaining resection strategy. The specific implementation process is as follows: Preoperatively, the patient underwent high-resolution T1-weighted enhanced MRI, T2-FLAIR, DTI (diffusion tensor imaging), and MR angiography. All images were uniformly registered into the preoperative three-dimensional tissue model system. After multimodal reconstruction, the system divided the 3cm area around the tumor into five structural impedance sub-regions based on the microstructural characteristics of brain tissue: white matter fiber tracts (such as the superior frontal corona radiata), meningeal layer, vascular interstitial channels, glial tissue, and sulci. In the three-dimensional spatial model, the system assigned a structural impedance level to each sub-region, with a numerical range of 0.2 to 1.0, where the vascular interstitial channel was assigned 0.2 (lowest impedance), the meningeal layer 0.5, the white matter fiber tract 0.7, the glial tissue 0.6, and the sulci 0.3. The system then projects equal-length paths (initially set to 20 mm) in 360 directions (with an accuracy of 1 degree of angle) from the tumor boundary grid points as starting points. For each voxel along the path, the system calculates the tissue types it traverses and accumulates the impedance value. Taking a representative starting point P at the anterior edge of the tumor as an example, the path traversing the tissues in the left anterolateral direction is meninges → intervascular space → sulcus → glial layer → white matter → meninges, with a corresponding impedance accumulation of 0.5 + 0.2 + 0.3 + 0.6 + 0.7 + 0.5 = 2.8. However, the path traversing the tissues in the posterior direction is meninges → white matter → glial layer → white matter → white matter → meninges, with a corresponding total impedance of 0.5 + 0.7 + 0.6 + 0.7 + 0.7 + 0.5 = 3.7. Clearly, the anterolateral direction is a more accessible direction.

[0060] Based on the above calculations, the system generates a three-dimensional directional accessibility heatmap, displaying low-impedance directions (i.e., paths where the tumor is more likely to infiltrate) in red-orange and high-impedance directions in blue-green. In the navigation interface, this heatmap is projected in real-time onto the three-dimensional surgical view, allowing the surgeon to clearly observe the outward expansion path and direction of the tumor. In this patient, the system identified three high-accessibility paths, pointing to the anterior border of the subfrontal sulcus, the anterior branch region of the internal capsule, and the cribriform plate region of the skull base, respectively. Among these, the subfrontal sulcus direction, due to the presence of continuous sulci and superficial vascular pathways, had a cumulative impedance below 2.6, and was therefore classified by the system as a priority infiltration pathway at risk level 1. Based on this, the surgeon adjusted the previously untreated anterior border portion in the intraoperative strategy, employing a pre-extension resection approach extending it anteriorly by 8mm to completely remove the area. Simultaneously, intraoperative dynamic electrical stimulation monitoring was performed on the path pointing towards the cribriform plate, and resection in this direction was paused to avoid damage to the olfactory conduction tract.

[0061] Postoperative imaging evaluation revealed a small area of ​​non-enhanced abnormal signal in the subfrontal sulcus direction. Pathological examination confirmed the presence of a small amount of tumor cell extravasation in this area, indicating that pre-dilation resection achieved clinical value. Conversely, no tumor residue or infiltration was observed in the high-impedance area, demonstrating the good predictive efficacy of accessibility thermography. Ultimately, this case, through a three-step mechanism of structural impedance modeling, accessibility path quantification, and intraoperative strategy adjustment, accurately identified the potential invasion direction of the tumor, achieving targeted path optimization and risk control, significantly outperforming the traditional strategy that uses the range of image enhancement as the sole boundary.

[0062] After initially identifying the high-risk infiltration direction of the tumor into the subfrontal sulcus using mechanical residual mapping and directional accessibility thermography, the surgeon further utilized structural impedance modeling to quantitatively analyze the permeability of various tissue regions along the access path, thereby improving the physiological accuracy and clinical interpretability of the directional accessibility thermography. First, based on the patient's preoperative MRI and DTI images, combined with a tissue partitioning database, the system performed a detailed classification of the tissue structures within a 3cm radius around the tumor, including: dense white matter fibers (such as the anterior branch of the frontal corona radiata), loose glial tissue, meningeal complex, vascular spaces, and sulci. Physiological parameters corresponding to each tissue region in the anatomical specimen database were then sequentially sampled and obtained. The system retrieved tissue density values ​​(in cells / mm²). 3 ), average intercellular width (μm), liquid permeability (ml / mm) 2 The structural impedance level coefficient was set according to the following range: when the packing density is >400,000 cells / mm², the structural impedance level coefficient was set as follows: (The text then lists the values ​​for cells / mm² and their corresponding values ​​for structural impedance level coefficients.) 3Furthermore, when the permeability coefficient is <0.05, an impedance level of 1.0 is assigned (e.g., dense white matter); when the arrangement density is 150,000–250,000, the intercellular space is >8 μm, and the tissue continuity is <0.4, a value of 0.2–0.4 is assigned (e.g., intervascular spaces or fissure tissues); based on this standard, the system ultimately forms a complete three-dimensional structural impedance distribution map of the surrounding tissues, in which the average impedance value of the subfrontal sulcus region is 0.33, while that of the anterior branch region of the internal capsule is 0.71, and that of the cribriform plate region of the skull base is 0.84.

[0063] Next, using the tumor boundary as the starting plane, the system projects 720 radial channel paths in three dimensions at a resolution of 1°. Impedance accumulation is performed on the tissue regions traversed by each path, and the results are standardized to generate a directional accessibility heatmap. This heatmap is displayed on the intraoperative navigation platform using a spherical projection, employing a color gradient from red (high accessibility) to blue (low accessibility) to represent the relative penetration potential of each direction. The accessibility index is also numerically indicated, ranging from 0 to 1, where ≥0.8 is defined as a first-level high-risk infiltration direction, 0.6-0.8 as medium risk, and <0.6 as a low-risk area. During surgical navigation, the system overlays this three-dimensional heatmap onto the tumor boundary outer shell layer, forming a wraparound risk prediction view. The surgeon can directly observe which directions correspond to tissue channels with high structural permeability and high cell migration potential.

[0064] During the patient's surgery, accessibility thermography showed an accessibility index of 0.86 towards the subfrontal sulcus, classifying it as a high-risk pathway (Level 1); 0.58 towards the internal capsule (low risk); and 0.63 towards the skull base (intermediate risk). Based on this, the surgeon decided to adopt the recommended pre-expansion strategy, resecting an additional 0.7 cm towards the subfrontal sulcus and performing intraoperative electrical stimulation monitoring in the skull base region to determine if any functional pathways were accessible. Intraoperative real-time feedback showed that the tissue release in the subfrontal sulcus region was complete after resection, and no residue was found on the contrast-enhanced MRI scan during intraoperative re-scan. However, there was no clear olfactory response in the skull base region upon stimulation, and the surgeon ultimately adhered to the predetermined preservation strategy and did not resect this region.

[0065] Postoperative pathology confirmed interstitial infiltration in the extended area of ​​the subfrontal sulcus. Follow-up MRI showed no recurrence, and the patient recovered well postoperatively. This suggests that the method of combining structural impedance level with directional accessibility thermography can effectively predict the actual infiltration path and provide highly individualized decision support for resection strategy.

[0066] After completing the three-dimensional structural impedance modeling and generating directional accessibility heatmaps, the surgeon further applied tissue directional consistency indices to enhance the assessment of access pathways, identifying high-risk pathways for tumor cell migration along structural continuity directions, and ultimately achieving dynamic adjustment of the resection boundary in the intraoperative navigation system. This step was completed by the system's built-in directional consistency analysis module. First, the dominant orientation of the patient's white matter fiber bundles was extracted from the preprocessed DTI images, and the main structural orientations within a 2cm region around the tumor were encoded using a tensor field consistency scoring model, with a score of 0-1, where 1 represents extremely consistent orientation and high structural continuity, and 0 represents anisotropic disorder and structural obstruction. Taking the left anterior margin of the tumor as an example, its adjacent area contains some superficial branches of the corona radiata and prefrontal cortex fibers, with a DTI directional consistency score as high as 0.92, indicating that the tissue fibers in this direction are arranged very regularly. If there is cell infiltration dynamics, the tumor is very likely to extend forward along the structure. Conversely, the area at the posteroinferior margin of the tumor is a confluence area of ​​intersecting fiber bands, with a directional consistency of only 0.41, indicating strong structural barriers that are not conducive to continuous cell diffusion.

[0067] The system integrates the directional consistency index with the previously generated accessibility heatmap through a weighted multiplication process to construct an enhanced accessibility index. Directional consistency is used as a weighting factor in the calculation formula to participate in path weight allocation, thereby further highlighting high-risk pathways with not only low tissue impedance but also strong structural guidance in the navigation heatmap. On the patient navigation interface, the superimposed heatmap clearly shows an enhanced red path extending from the tumor to the left anterior (along the fiber bundle direction), with its comprehensive accessibility index increasing from 0.79 to 0.91. The system marks this direction as a primary priority area and, in conjunction with the resection range planning tool within the navigation system, suggests that the surgeon consider dynamically extending the resection boundary in this direction by 6-10 mm. Following this suggestion, the surgeon manually adjusted the boundary line in this direction within the intraoperative navigation system, moving the originally planned boundary line 0.8 cm to the left anterior. Simultaneously, guided by a microscope, layer-by-layer resection was performed to this area. During resection, the tissue appeared soft and the boundaries were indistinct. Intraoperative rapid frozen section analysis revealed small clusters of tumor cells in this area, consistent with the system's prediction.

[0068] Furthermore, the system provides a variable boundary line locking function on the navigation interface, allowing the surgeon to dynamically adjust the resection boundary by selecting points on the navigation interface based on force feedback, visual judgment, and tissue response during real-time operation. For example, when the surgeon finds that the tissue has extremely strong resilience, no obvious loosening sensation, and low structural guidance (directional consistency <0.3) during resection in the direction of the skull base, the resection boundary line in that direction can be immediately retracted by 0.5cm to avoid accidentally entering non-invasive areas or damaging functional areas. Ultimately, postoperative MRI and pathological comparison showed that the left anterior extended area was indeed a high-risk infiltration area, containing tumor cell extravasation zones, while no residual tissue or lesions were found in the avoided area of ​​the skull base.

[0069] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An optimized resection strategy for meningioma infiltration boundaries, characterized in that... Includes the following steps: By utilizing the differences in stress and strain response during intraoperative tissue traction and suction operations, biomechanically abnormal areas are marked in space; a mechanical residual map is constructed to compare the anisotropic regions around the tumor with the normal meningeal layer; and the local tissue compliance critical point is extracted as a physiological landmark of the invasion boundary. In a three-dimensional tissue atlas, highly adaptive channels including intervascular spaces, submembrane tunnels, and white matter fibrous gaps are identified; the structural impedance of different regions is modeled in a partitioned manner, and the infiltration accessibility in each direction is calculated; a boundary evolution stability heat map is constructed, and the boundary prediction results are mapped back to the current surgical scene to determine whether the resection range needs to be expanded. After each tissue stripping or resection, the corresponding area's tissue deformation rate, blood flow velocity, and temperature diffusion pattern indicators are collected as a reflection of intraoperative boundary stress response to assess whether the resection has reached a high-risk area. Once the boundary stress indicators exceed a preset threshold, the system suspends further resection operations in the corresponding area and re-optimizes the approach angle, resection sequence, and boundary advancement direction based on the remaining anatomical structures to avoid high-risk areas. The method for extracting the critical point of local tissue compliance includes the following steps: During the operation, force feedback devices were used to collect tissue mechanical response data of the meningioma and surrounding tissues under traction, suction, and compression. Based on the mechanical response data, a three-dimensional spatial distribution of mechanical residual maps was constructed to express the stress and strain differences in different regions under external force. Anisotropic regions were identified in the mechanical residual maps, and the response characteristics of normal meningeal tissue and tumor edge tissue were compared. Based on strain mutation points or compliance mutation patterns, local tissue compliance critical points representing the tumor invasion boundary were extracted as real physiological landmarks during the operation. These physiological landmarks were used to guide the determination of the meningioma resection range and path, thereby controlling the invasion boundary. The method for calculating the infiltration accessibility in each direction includes the following steps: S1. Acquire preoperative multimodal medical image data of meningioma and its surrounding tissues to identify local tissue structure types; based on the microstructural characteristics of brain tissue, divide the tumor-peripheral tissue region into multiple structural impedance subregions, including: white matter fiber bundles, meninges, intervascular spaces, glial tissue, and sulci. S2. Set the structural impedance level for each sub-region to form a structural impedance distribution map; starting from the tumor boundary, project paths along multiple spatial directions and calculate the cumulative impedance value of the area traversed by each path; generate a directional accessibility assessment heatmap based on the total cumulative impedance value of each path to identify potential preferred invasion paths of the tumor in different directions. S3. Based on the accessibility assessment results, adjust the resection strategy and perform extended pre-resection or focused intraoperative monitoring on high-accessibility areas.

2. The method for optimizing the resection strategy of meningioma infiltration boundary according to claim 1, characterized in that... The tissue mechanical response data includes displacement amplitude, rebound time, traction recovery ratio, and tissue compressive strength curve parameters; the mechanical residual map is obtained by constructing the difference between the actual deformation and the predicted model deformation of each unit area under standard traction strength.

3. The method for optimizing the resection strategy of meningioma infiltration boundary according to claim 1, characterized in that... The local tissue compliance critical point is the spatial point in the residual map where a nonlinear transition occurs, and the transition reflects the transformation from a physiologically connected structure to a pathologically relaxed structure.

4. The method for optimizing the resection strategy of meningioma infiltration boundary according to claim 1, characterized in that... The physiological landmarks are overlaid on the navigation system display interface in a visual manner to guide the dynamic updating of the resection boundary; the force feedback acquisition device is a miniature stress sensing module integrated into the surgical suction head, micromanipulation arm or intraoperative navigation probe.

5. The method for optimizing the resection strategy of meningioma infiltration boundary according to claim 1, characterized in that... The structural impedance level is set based on physiological parameters such as tissue arrangement density, intercellular space size, permeability, and tissue continuity; the directional accessibility heatmap visualizes the relative accessibility in each direction in three dimensions, and uses color gradients or probability values ​​to indicate different risk levels.

6. The method for optimizing the resection strategy of meningioma infiltration boundary according to claim 1, characterized in that... The accessibility assessment of the path includes a tissue direction consistency index, which is used to enhance the weight of the penetration path along the structural direction; the resection strategy optimization includes setting a variable boundary line in the intraoperative navigation system based on the directional accessibility results, so as to dynamically adjust the resection range in real time.

Images