Hepatic tumor radiofrequency ablation path optimization method based on hemodynamic simulation

Abstract

The present application relates to a method for optimizing radiofrequency ablation path of liver tumor based on hemodynamic simulation, comprising: acquiring collected liver image data, tumor region data, intrahepatic blood vessel data and hemodynamic simulation results, determining an expanded boundary according to the tumor region, and determining the expanded boundary region meeting a preset blood flow heat dissipation criterion as a cooling dominated region based on the hemodynamic simulation results; generating candidate radiofrequency ablation path data based on the cooling dominated region and performing anatomical safety judgment; performing thermal effective judgment on the candidate path passing the anatomical safety judgment, judging whether the predicted thermal action region covers the cooling dominated region, and sorting the candidate path according to the anatomical safety judgment result and the thermal effective judgment result and outputting the optimization result. The present application can compensate for the area affected by blood flow heat dissipation, and take into account the puncture safety, thermal field coverage sufficiency and path selection objectivity, thereby improving the clinical availability of the optimization result of the radiofrequency ablation path of liver tumor.

Description

Technical Field

[0001] This invention relates to the fields of medical image processing, preoperative planning for radiofrequency ablation of liver tumors, and hemodynamic simulation technology, specifically a method for optimizing the radiofrequency ablation path for liver tumors based on hemodynamic simulation. Background Technology

[0002] In existing technologies, radiofrequency ablation path planning for liver tumors is typically based on imaging data such as CT and MR. After reconstructing structures such as the liver, tumor, blood vessels, body surface, and bones, the path is selected based on constraints such as whether the puncture path avoids blood vessels, ribs, and other important tissues, whether the ablation zone covers the tumor, and the number of needle insertions and the needle angle. For example, Chinese patent CN113456219A discloses a method and device for planning radiofrequency ablation surgical paths for liver cancer based on CT images. It generates paths and determines the size and location of the ablation zone through organ model extraction, point set construction, path selection, Pareto optimality judgment, and integer programming. Chinese patent CN112043377A discloses a method and system for CT arbitrary section ultrasound field simulation-assisted ablation path planning. It simulates the ultrasound field based on CT data and generates a needle insertion path that satisfies the constraint of non-intersection of blood vessels. The above solutions can, to a certain extent, reduce the subjectivity caused by relying solely on physician experience for path selection and improve the automation and visualization of preoperative planning.

[0003] However, the efficacy of radiofrequency ablation for liver tumors depends not only on the geometric safety of the puncture path and the static coverage of the ablation zone over the tumor contour, but also significantly on the heat dissipation effect caused by intrahepatic vessels and blood perfusion. Tumor boundary areas adjacent to larger vessels or high-flow segments are prone to heat sinking during ablation, resulting in insufficient local temperature rise, residual active tumor, or the need for additional ablation. Existing path planning schemes often treat vessels as anatomical risks to be avoided, or only use the intersection of the path and vessels as constraints. They fail to identify the cooling-dominated area caused by blood flow obstruction or heated transport based on the positional relationship of the vessel relative to the heat source and the tumor's outer boundary, as well as blood flow and direction. Furthermore, they fail to use this cooling-dominated area as a core evaluation object for path generation, needle tip offset, and path sequencing.

[0004] Furthermore, Chinese patent CN116798638A discloses a three-dimensional temperature field simulation method for microwave ablation of liver tumors. This method generates a three-dimensional temperature field simulation result inside the tumor by combining liver image modeling, biological heat transfer equations, and boundary condition determination, thereby improving the safety of temperature judgment during surgery. While this type of temperature field simulation focuses on the ablation thermal field and boundary conditions, it is mainly used for thermal field prediction or temperature distribution display. It has not yet fully integrated the radiofrequency ablation path candidate generation process to uniformly sort and output the results of blood flow heat dissipation criteria, cooling dominance zone coverage, related vessel separation relationships, and anatomical safety judgment. Therefore, there is an urgent need for a method that can collaboratively use liver images, tumor regions, intrahepatic vessels, and hemodynamic simulation results for radiofrequency ablation path optimization to improve the reliability and thermal effectiveness of ablation planning for liver tumors adjacent to blood vessels. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the radiofrequency ablation path for liver tumors based on hemodynamic simulation, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0006] The present invention adopts the following technical solution to solve the above-mentioned technical problems:

[0007] A method for optimizing the radiofrequency ablation path for liver tumors based on hemodynamic simulation, executed by a processor, includes: acquiring acquired liver image data, tumor region data, intrahepatic vascular data and hemodynamic simulation results; determining the outer boundary based on the tumor region; and determining the outer boundary region that meets the preset blood flow heat dissipation criterion as the cooling control zone based on the hemodynamic simulation results.

[0008] Candidate radiofrequency ablation path data is generated based on the cooling control area, and the anatomical safety assessment is performed on the candidate radiofrequency ablation path data to obtain the anatomical safety assessment result.

[0009] The candidate radiofrequency ablation path data that has passed the anatomical safety determination is thermally effective. The thermal effectiveness determination includes determining whether the predicted thermal effect area corresponding to the candidate radiofrequency ablation path data covers the cooling dominance area. The candidate radiofrequency ablation path data is sorted according to the anatomical safety determination result and the thermal effectiveness determination result, and the radiofrequency ablation path optimization result is output.

[0010] Furthermore, the preset blood flow heat dissipation criterion is as follows: when the associated blood vessel segment is located between the preset heat source reference position and the outer expansion boundary region, and the blood flow of the associated blood vessel segment reaches the preset heat dissipation threshold, the outer expansion boundary region is determined as the cooling control area formed by blood flow obstruction.

[0011] Furthermore, when generating candidate radiofrequency ablation path data, the offset direction of the needle tip planning position is determined based on the position of the cooling dominance area relative to the tumor region and the blood flow direction of the associated vascular segment, and candidate radiofrequency ablation path data is generated based on the needle tip planning position.

[0012] Furthermore, in the thermal effectiveness determination, it is determined whether the predicted thermal effect area corresponding to the candidate radiofrequency ablation path data and the cooling dominance area are separated by the associated vascular segment that satisfies the preset blood flow heat dissipation criterion; if they are separated, the thermal effectiveness score or priority of the candidate radiofrequency ablation path data is reduced.

[0013] Furthermore, when multiple cooling control zones exist, the projection position of each cooling control zone in the blood flow direction of the corresponding associated blood vessel segment is determined, and the cooling control zone whose projection position is on the downstream side of the blood flow heat and whose area or volume meets the preset screening conditions is taken as the main compensation zone. The offset direction of the needle tip planning position is determined based on the main compensation zone.

[0014] Furthermore, after determining the offset direction of the needle tip planning position, the offset amplitude of the needle tip planning position is limited; when the coverage increment of the predicted heat effect area on the cooling control area after offset is less than a preset increment threshold, and the distance from the needle tip planning position to the dissection risk area is less than a preset safety threshold, the offset amplitude of the needle tip planning position is reduced.

[0015] Furthermore, when generating candidate radiofrequency ablation path data based on the needle tip planning position, it is determined whether the heat release main axis corresponding to the candidate radiofrequency ablation path data is oriented towards the cooling control area; if the heat release main axis is away from the cooling control area, the priority of the candidate radiofrequency ablation path data is reduced.

[0016] Furthermore, before reducing the priority of the candidate radiofrequency ablation path data, it is determined whether the cooling control area is located downstream of the blood flow heat zone based on the blood flow direction of the associated vascular segment; if so, the candidate radiofrequency ablation path data is marked as a downstream obstruction path and given a higher degradation level than non-downstream obstruction paths.

[0017] Furthermore, when the predicted thermal effect area and the cooling dominance area are separated by the associated vascular segment, it is determined whether there is alternative candidate radiofrequency ablation path data that places the predicted thermal effect area and the cooling dominance area on the same side; if so, the priority of the alternative candidate radiofrequency ablation path data is increased.

[0018] Furthermore, when multiple alternative candidate radiofrequency ablation path data exist, it is determined within a preset needle tip deviation range whether the predicted thermal effect area and the cooling dominance area are still located on the same side of the associated blood vessel segment; if they are still on the same side, the priority of the corresponding alternative candidate radiofrequency ablation path data is increased; if they are no longer on the same side, the priority increase processing for the corresponding alternative candidate radiofrequency ablation path data is not performed.

[0019] This invention acquires liver imaging data, tumor region data, intrahepatic vascular data, and hemodynamic simulation results. It determines the outer boundary based on the tumor region and identifies the outer boundary region that meets the preset blood flow heat dissipation criteria as the cooling dominance zone. This allows path planning to no longer rely solely on tumor contour coverage or vascular avoidance, but to perform compensatory planning for areas significantly affected by blood flow heat dissipation, thereby reducing the risk of insufficient ablation in adjacent vascular regions.

[0020] This invention generates candidate radiofrequency ablation path data based on the cooling dominance zone and sequentially performs anatomical safety and thermal effectiveness assessments. This allows for the determination of whether the predicted thermal effect area covers the cooling dominance zone, while ensuring that candidate paths avoid anatomically risky areas. Therefore, by ranking the candidate paths according to the anatomical safety and thermal effectiveness assessment results and outputting optimized results, it is beneficial to balance puncture safety, sufficient thermal field coverage, and objectivity in path selection, thereby improving the clinical usability of optimized radiofrequency ablation paths for liver tumors. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the method for optimizing the radiofrequency ablation path for liver tumors according to the present invention.

[0022] Figure 2 This is a diagram showing the determination of blood flow heat dissipation threshold and cooling dominance area in Embodiment 1 of the present invention.

[0023] Figure 3 This is a diagram showing the determination of needle tip offset coverage increment and safety distance in Embodiment 1 of the present invention.

[0024] Figure 4 This is a ranking diagram of candidate radiofrequency ablation paths in Embodiment 1 of the present invention.

[0025] Figure 5 This is a graph showing the downgrade determination of thermal effectiveness score in Embodiment 2 of the present invention.

[0026] Figure 6 This is the spatial determination diagram of the alternative candidate paths on the same side in Embodiment 2 of the present invention.

[0027] Figure 7 This is a diagram showing the stability ratio of needle tip deviation on the same side in Embodiment 2 of the present invention. Detailed Implementation

[0028] As attached Figure 1 As shown, in one embodiment, the processor acquires pre-processed liver imaging data, tumor region data, intrahepatic vascular data, and hemodynamic simulation results. The liver imaging data may include CT imaging data, magnetic resonance imaging data, or ultrasound imaging data. The tumor region data characterizes the spatial location, morphological contour, and volume range of the tumor within the liver. The intrahepatic vascular data characterizes the spatial orientation and diameter information of the portal vein, hepatic vein, hepatic artery, or their branches. The hemodynamic simulation results characterize the blood flow direction, blood flow velocity, blood flow rate, and heat dissipation caused by blood flow in the intrahepatic vessels. Before the processor performs path optimization, the aforementioned liver imaging data, tumor region data, intrahepatic vascular data, and hemodynamic simulation results are converted to the same spatial coordinate system. The tumor region data can be a tumor segmentation mask, a three-dimensional tumor contour, or a tumor surface point set. The intrahepatic vascular data can be a vascular segmentation mask, a vascular centerline, a vascular surface model, or a list of vascular segments. The hemodynamic simulation results can establish a correspondence with the corresponding vascular segments, vascular centerline sampling points, or liver spatial grid points. The processor determines the outer boundary surrounding the tumor region based on tumor region data. The outer boundary is used to represent the safe ablation range that needs to be considered during radiofrequency ablation.

[0029] Specifically, the processor can expand the tumor region outward according to a preset safe expansion distance, and trim or mark the expansion result based on the liver boundary, liver capsule, important blood vessels, bile ducts, and other preset anatomical risk areas to obtain an expanded boundary region composed of multiple boundary sub-regions. Subsequently, based on hemodynamic simulation results, the processor identifies areas within the expanded boundary region affected by blood flow heat dissipation, and determines the expanded boundary regions that meet preset blood flow heat dissipation criteria as cooling dominance zones. Cooling dominance zones indicate areas where heat deposition is insufficient due to heat being carried away by adjacent blood vessels during radiofrequency ablation. The processor further generates candidate radiofrequency ablation path data based on the cooling dominance zones. The candidate radiofrequency ablation path data includes at least one of the following: needle entry point, needle entry direction, needle tip planning position, and ablation range. The processor performs anatomical safety judgment on the candidate radiofrequency ablation path data to determine whether the candidate path avoids important intrahepatic blood vessels, bile ducts, liver capsule, and preset anatomical risk areas, and obtains the anatomical safety judgment result. Anatomical safety assessment can include determining whether the needle insertion segment, the area occupied by the needle body, and the planned position of the needle tip cross or excessively approach a preset anatomical risk area. It can also determine whether a candidate path passes the anatomical safety assessment based on a preset minimum safety distance. For candidate radiofrequency ablation path data that passes the anatomical safety assessment, the processor performs a thermal validity assessment.

[0030] Thermal effectiveness determination involves assessing whether the predicted thermal effect area corresponding to the candidate radiofrequency ablation path data covers the cooling dominance zone, thereby determining whether the candidate path can compensate for insufficient ablation caused by blood flow heat dissipation. The predicted thermal effect area can be determined based on the radiofrequency electrode type, effective segment length, preset ablation power, preset ablation time, tissue thermal parameters, hemodynamic simulation results, or pre-stored ablation templates, and mapped to the same spatial coordinate system as the outward boundary region. The processor comprehensively sorts the candidate radiofrequency ablation path data based on the anatomical safety determination results and the thermal effectiveness determination results, and outputs optimized radiofrequency ablation paths that simultaneously consider path safety and thermal coverage of the cooling dominance zone. The optimized radiofrequency ablation path results can include the needle insertion point, needle insertion direction, needle insertion depth, needle tip planning position, corresponding cooling dominance zone, predicted thermal effect area, and sorting results for the recommended path.

[0031] In a further implementation, the processor filters the outer boundary region based on a preset blood flow heat dissipation criterion. Specifically, the processor uses a preset heat source reference position as the reference position for the release of heat from radiofrequency ablation, and combines this with intrahepatic vascular data to determine associated vascular segments that have a heat dissipation influence relationship with the outer boundary region. The preset heat source reference position can be the geometric center of the tumor region, the conventional ablation center, the candidate needle tip planning position, or the center position of the effective action segment of the radiofrequency electrode. The processor can filter associated vascular segments from the intrahepatic vascular data according to the spatial distance between the vascular segment and the outer boundary region, the relative position of the vascular segment with respect to the heat source reference position and the outer boundary region, the vascular segment diameter, and the blood flow direction. When the associated vascular segment is located between the preset heat source reference position and the outer boundary region, it indicates that the heat generated by radiofrequency ablation may be blocked and affected by the associated vascular segment during its transfer to the outer boundary region. The processor further reads or calculates the blood flow of the associated vascular segment and compares the blood flow with a preset heat dissipation threshold. Blood flow can be directly read from hemodynamic simulation results, or obtained based on the diameter and blood flow velocity information of the associated vessel segment. The preset heat dissipation threshold can be pre-set based on historical case data, equipment parameters, physician experience, or preoperative planning requirements. When the blood flow of the associated vessel segment reaches the preset heat dissipation threshold, it is determined that the associated vessel segment has a sufficient heat dissipation effect on the corresponding outward boundary region to influence heat deposition. At this point, the processor identifies the outward boundary region as a cooling-dominated area formed by blood flow obstruction, so that the cooling-dominated area can be specifically compensated during subsequent candidate radiofrequency ablation path generation and thermal effectiveness determination.

[0032] In a further implementation, when generating candidate radiofrequency ablation path data, the processor first determines the spatial orientation of the cooling dominance area relative to the tumor region, and, in conjunction with the blood flow direction of the associated vessel segment, determines in which direction the radiofrequency ablation heat needs to be compensated. The spatial orientation can be determined based on the center position of the cooling dominance area, the distribution of boundary sub-regions, or the position of a representative point on the outward expansion boundary relative to the tumor region. Based on the aforementioned spatial orientation and blood flow direction, the processor determines the offset direction of the needle tip planning position, shifting it relative to the conventional ablation center towards the cooling dominance area, thereby improving the coverage of the predicted thermal effect area over the cooling dominance area. The conventional ablation center can be the geometric center of the tumor region, the center after merging the tumor region with the safe outward expansion range, or the ablation center determined in the preoperative planning when blood flow heat dissipation was not considered. Subsequently, the processor generates candidate radiofrequency ablation path data based on the needle tip planning position. The candidate radiofrequency ablation path data can include the needle entry point, needle entry angle, needle entry depth, needle tip planning position, and the predicted thermal effect area corresponding to the path. Specifically, the processor can select multiple needle insertion points in the body surface area or liver surface area where needle insertion is permitted, and connect each needle insertion point with the planned position of the needle tip to form a candidate needle insertion direction. Then, it generates candidate radiofrequency ablation path data based on the needle length, needle tip position, effective working segment position of the radiofrequency electrode, and preset needle insertion constraints.

[0033] When multiple cooling dominance zones exist, the processor determines the projected position of each cooling dominance zone along the blood flow direction of its corresponding associated vessel segment, and determines the relationship between the cooling dominance zone and the direction of blood flow heating based on the projected position. The projected position can be obtained by mapping a representative point of the cooling dominance zone along the blood flow direction of the corresponding associated vessel segment. The representative point can be the center point, area center point, volume center point of the cooling dominance zone, or the boundary point closest to the associated vessel segment. If the projected position of a cooling dominance zone is downstream of the blood flow heating direction, and the area or volume of the cooling dominance zone meets the preset screening criteria, the processor designates this cooling dominance zone as the primary compensation zone. The preset screening criteria can be that the area, volume, boundary length, proximity to the tumor region, or proximity to important blood vessels of the cooling dominance zone meet the corresponding preset requirements. The primary compensation zone is used to represent the area among multiple cooling dominance zones that requires more thermal compensation through needle tip planning position offset. The processor determines the offset direction of the needle tip planning position based on the primary compensation zone, so that the candidate radiofrequency ablation path prioritizes enhancing the thermal coverage of the primary compensation zone.

[0034] After determining the offset direction of the needle tip planning position, the processor limits the offset amplitude to avoid excessive offset that could reduce path safety or cause the ablation range to deviate from the target area. The offset amplitude limit can be preset based on the tumor area size, safe expansion range, effective action segment length of the radiofrequency electrode, predicted thermal effect area size, and distance from the needle tip to the anatomical risk area. The processor can adjust the offset amplitude based on the increase in coverage of the predicted thermal effect area over the cooling dominance area after offset, and the distance from the needle tip planning position to the anatomical risk area. The coverage increase can be determined by comparing the changes in area, volume, or number of boundary sub-regions of the predicted thermal effect area covering the cooling dominance area before and after offset. When the coverage increase is less than a preset increase threshold, and the distance from the needle tip planning position to the anatomical risk area is less than a preset safety threshold, it indicates that maintaining the current offset amplitude has limited improvement on cooling dominance area coverage and may increase anatomical risk. In this case, the processor reduces the offset amplitude of the needle tip planning position to achieve a balance between thermal coverage compensation and anatomical safety.

[0035] When generating candidate radiofrequency ablation path data based on the tip planning position, the processor also determines whether the thermal release axis corresponding to the candidate radiofrequency ablation path data is oriented towards the cooling dominance region. The thermal release axis can be determined based on the axial direction of the radiofrequency electrode, the thermal field extension direction, or the long axis direction of the predicted thermal effect area. The processor can compare the thermal release axis with the direction pointing from the tip planning position to the representative point of the cooling dominance region to determine the orientation relationship of the thermal release axis relative to the cooling dominance region. When the thermal release axis is oriented towards the cooling dominance region, it indicates that the candidate radiofrequency ablation path is beneficial for thermal compensation of the cooling dominance region. When the thermal release axis is away from the cooling dominance region, it indicates that the candidate radiofrequency ablation path has weak thermal coverage capability of the cooling dominance region, and the processor reduces the priority of the candidate radiofrequency ablation path data.

[0036] Before lowering the priority of candidate radiofrequency ablation path data, the processor further determines whether the cooling dominance area is located downstream of the heat-generating blood flow zone based on the blood flow direction of the associated vessel segment. If the cooling dominance area is located downstream of the heat-generating blood flow zone, it indicates that the associated vessel segment has a more significant heat removal effect on this area, and insufficient ablation is more likely to occur when the heat release axis deviates from this area. In this case, the processor marks the candidate radiofrequency ablation path data as a downstream obstruction path and assigns this downstream obstruction path a higher degradation margin than non-downstream obstruction paths, so that the ranking results can prioritize excluding candidate radiofrequency ablation paths that do not adequately compensate for the downstream cooling dominance area. The degradation margin can be pre-defined based on the size of the cooling dominance area, the blood flow of the associated vessel segment, the distance between the associated vessel segment and the cooling dominance area, and the degree of deviation from the heat release axis. The processor adjusts the position of the candidate radiofrequency ablation path data in the comprehensive ranking according to the corresponding classification results.

[0037] In a further implementation, when determining thermal effectiveness, the processor checks whether there exists a related vascular segment that satisfies a preset blood flow cooling criterion between the predicted thermal effect area and the cooling dominance area corresponding to the candidate radiofrequency ablation path data. The processor can determine whether the related vascular segment is located on the main path from the predicted thermal effect area to the cooling dominance area based on the positional relationship of the predicted thermal effect area, the cooling dominance area, and the related vascular segment in the same spatial coordinate system. If the related vascular segment is located between the predicted thermal effect area and the cooling dominance area, the processor determines that the predicted thermal effect area and the cooling dominance area are separated by the related vascular segment. Since the blood flow in the related vascular segment carries away the heat generated by radiofrequency ablation, even if the predicted thermal effect area is spatially close to the cooling dominance area, it may still be difficult to form effective thermal coverage of the cooling dominance area due to the influence of blood flow cooling. Therefore, when the processor determines that the predicted thermal effect area and the cooling dominance area are separated by the related vascular segment, it lowers the thermal effectiveness score or priority of the candidate radiofrequency ablation path data, reducing the recommendation level of the candidate radiofrequency ablation path data in subsequent ranking. If the predicted thermal effect area covers the cooling control area, and there is no associated vascular segment between the two that meets the preset blood flow heat dissipation criterion, then the processor can determine that the candidate radiofrequency ablation path data has effective thermal coverage capability for the cooling control area.

[0038] When the predicted thermal effect area and the cooling dominance area are separated by an associated vessel segment, the processor further searches the generated candidate radiofrequency ablation path data set to determine if alternative candidate radiofrequency ablation path data exists. The predicted thermal effect area and cooling dominance area corresponding to the alternative candidate radiofrequency ablation path data are located on the same side of the associated vessel segment, ensuring that the transfer of radiofrequency ablation heat to the cooling dominance area does not need to cross the associated vessel segment, thereby reducing the impact of blood flow heat dissipation on the thermal coverage effect. "Same side" can be determined based on the vessel centerline, vessel surface model, or local segmentation plane of the associated vessel segment. The processor performs lateral judgments on representative points of the predicted thermal effect area and the cooling dominance area relative to the associated vessel segment. When they are in the same lateral region, it is determined that the predicted thermal effect area and the cooling dominance area are on the same side of the associated vessel segment. If the processor determines that alternative candidate radiofrequency ablation path data exists, it increases the priority of the alternative candidate radiofrequency ablation path data, giving it a higher ranking position in the radiofrequency ablation path optimization results.

[0039] When multiple alternative candidate radiofrequency ablation (RFA) path data exist, the processor performs a stability assessment on each alternative candidate RFA path data. Specifically, within a preset needle tip deviation range, the processor simulates the deviation of the planned needle tip position corresponding to each alternative candidate RFA path data and reassesses whether the predicted thermal effect area and cooling control area are still located on the same side of the associated vessel segment. The preset needle tip deviation range can be pre-set based on RFA needle positioning error, image registration error, respiratory motion effects, or clinically permissible errors. The deviation simulation can include anterior-posterior deviation along the needle insertion direction, lateral deviation perpendicular to the needle insertion direction, and multiple discrete deviation positions around the planned needle tip position. If, within the preset needle tip deviation range, the predicted thermal effect area and cooling control area are still located on the same side of the associated vessel segment, it indicates that the alternative candidate RFA path data has good tolerance for needle tip positioning errors, and the processor increases the priority of the corresponding alternative candidate RFA path data. If, within the preset needle tip deviation range, the predicted thermal effect area and the cooling control area are no longer located on the same side of the associated blood vessel segment, it indicates that the alternative candidate radiofrequency ablation path data may be affected by blood flow separation again under the actual needle insertion deviation, and the processor will not perform priority increase processing on the corresponding alternative candidate radiofrequency ablation path data.

[0040] Example 1:

[0041] In this embodiment, the processor acquires pre-processed liver CT image data, tumor region data, intrahepatic vascular data, and hemodynamic simulation results. The tumor region data is a three-dimensional tumor region obtained from liver CT image segmentation. The intrahepatic vascular data includes the spatial orientation, centerline, and diameter information of the portal vein, hepatic vein, and their branches. The hemodynamic simulation results include the blood flow direction, velocity, and flow rate information for each vascular segment. The processor registers the above data to the same spatial coordinate system, enabling the tumor region, intrahepatic vascular segments, and hemodynamic simulation results to be determined within the same three-dimensional space.

[0042] The processor determines the outer boundary based on the tumor region. Specifically, the maximum diameter of the tumor region is 28 mm, and the processor extends the tumor region outward by a safe extension distance of 5 mm to obtain the outer boundary region surrounding the tumor region. The outer boundary region is used to represent the safe ablation range that needs to be considered during radiofrequency ablation. The processor further divides the outer boundary region into multiple boundary sub-regions to determine whether each boundary sub-region is affected by heat dissipation from adjacent blood vessels.

[0043] The processor uses the geometric center of the tumor region as a preset heat source reference location and combines intrahepatic vascular data to identify associated vascular segments that have a heat dissipation impact relationship with the outward boundary region. When an associated vascular segment is located between the preset heat source reference location and the corresponding outward boundary region, the processor determines that the heat generated by radiofrequency ablation may be blocked and affected by the associated vascular segment when it is transferred to the outward boundary region. Figure 2 The distribution of blood flow in the associated vascular segments corresponding to different outward expansion boundary sub-regions is shown. A preset heat dissipation threshold is used as the criterion to distinguish the boundary sub-regions that reach the threshold from those that do not, thus intuitively reflecting the selection criteria for the cooling control area.

[0044] The processor further calculates the blood flow rate of the associated blood vessel segment based on the blood flow velocity and diameter information. The blood flow rate calculation formula is:

[0045]

[0046] in, This indicates the blood flow in the associated vascular segment. This indicates the average blood flow velocity within the associated vascular segment. This represents the cross-sectional area of ​​the associated blood vessel segment. The formula is derived from the fact that the volume of blood passing through the cross-section of the blood vessel per unit time equals the average blood flow velocity multiplied by the cross-sectional area of ​​the blood vessel.

[0047] When the cross-section of a blood vessel is approximately circular, the processor calculates the cross-sectional area based on the vessel's inner diameter, using the following formula:

[0048]

[0049] in, Indicates the cross-sectional area of ​​the blood vessel. Indicates the inner diameter of the associated vascular segment. This represents pi (π). The formula is derived from the relationship of circular cross-sectional area calculation and is used to convert the diameter of a blood vessel into a cross-sectional area that can be used to calculate blood flow.

[0050] For example, in a certain de-identified case dataset, the processor identifies a portal vein branch near the outer boundary region, with an inner diameter of 3.2 mm and an average blood flow velocity of 0.18 m / s. Based on the obtained cross-sectional area of ​​approximately 8.04 mm², the processor converts 8.04 mm² to m² and inputs it, obtaining m³ / s, i.e., a blood flow rate of approximately 1.45 mL / s. Since the preset heat dissipation threshold is 1.20 mL / s, the blood flow rate of the associated vessel segment reaches the preset heat dissipation threshold, and this associated vessel segment is located between the preset heat source reference location and the outer boundary region. Therefore, the processor identifies the corresponding outer boundary region as the cooling-controlled area formed by blood flow obstruction. Correspondingly, Figure 2The blood flow rate at the middle boundary 3 is 1.45 mL / s, and the blood flow rate at the boundary 5 is 1.28 mL / s. Both are higher than the preset heat dissipation threshold of 1.20 mL / s, so they can be used as a graphical representation of the cooling dominance zone determination results.

[0051] When the processor determines the existence of multiple cooling dominance zones, it determines the spatial orientation of each cooling dominance zone relative to the tumor region and, based on the blood flow direction of the corresponding associated vessel segment, determines the projected position of each cooling dominance zone in the blood flow direction. If a cooling dominance zone is located downstream of the blood flow heat zone and its area or volume meets preset screening criteria, the processor designates this cooling dominance zone as the primary compensation zone. The primary compensation zone represents the area among the multiple cooling dominance zones that most requires heat compensation through needle tip positioning offset.

[0052] The processor determines the offset direction of the needle tip planning position based on the location of the main compensation area relative to the tumor region and the blood flow direction of the corresponding associated vascular segment. Specifically, the processor uses the conventional ablation center determined without considering blood flow heat dissipation as a reference position, and shifts the needle tip planning position relative to the conventional ablation center towards the main compensation area. This makes the predicted thermal effect area of ​​the radiofrequency electrode closer to the main compensation area, improving the thermal coverage capability of the cooling control area.

[0053] After determining the offset direction of the needle tip planning position, the processor limits the offset amplitude. The processor compares the change in coverage of the predicted thermal effect area over the cooling-dominated area before and after the offset, and determines whether to retain the current offset amplitude based on the minimum distance from the needle tip planning position to anatomically risky areas such as the bile duct, important blood vessels, and liver capsule. The coverage increment calculation formula is:

[0054]

[0055] in, This indicates the increase in coverage of the cooling control area after tip bias. This indicates the coverage ratio of the predicted thermal effect area to the cooling dominance area after biasing. This represents the proportion of the predicted thermal effect area covering the cooling dominance area before biasing. This formula is obtained by comparing the difference in coverage proportions before and after biasing, and is used to determine whether tip biasing brings about an effective improvement in thermal coverage.

[0056] For example, the processor calculates the coverage ratio before bias. The coverage ratio after bias is 0.62. If it is 0.81, then This indicates that the coverage of the cooling control area increases by 19 percentage points after the needle tip is offset. Simultaneously, the processor measures a minimum distance of 6.5 mm from the planned needle tip position to the risk area of ​​the bile duct, which is greater than the preset safety threshold of 5.0 mm. This indicates that the current offset amplitude improves the coverage of the cooling control area without exceeding the preset anatomical safety limit; therefore, the processor retains the current needle tip offset amplitude. Figure 3 The diagram illustrates the synchronous changes in the coverage ratio of the cooling dominance area and the distance from the needle tip to the risk area before and after tip offset. The coverage ratio increases from 62% to 81%, while the safety distance remains above the preset safety threshold of 5.0 mm, indicating that the offset result simultaneously satisfies thermal coverage compensation and dissection safety constraints. If the coverage increment is less than the preset increment threshold, and the distance from the planned needle tip position to the dissection risk area is less than the preset safety threshold, the processor reduces the offset amplitude of the planned needle tip position to reduce path safety risks.

[0057] The processor generates candidate radiofrequency ablation path data based on the needle tip planning position. The candidate radiofrequency ablation path data includes the needle insertion point, insertion direction, insertion depth, needle tip planning position, and the predicted thermal effect area corresponding to the path. The processor can select multiple needle insertion points within a permissible body surface area or liver surface area, connect each insertion point with the needle tip planning position to form candidate insertion directions, and then generate multiple candidate radiofrequency ablation path data based on the needle length, the effective action segment position of the radiofrequency electrode, and preset insertion constraints.

[0058] The processor performs an anatomical safety assessment on candidate radiofrequency ablation path data, determining whether the needle insertion path avoids important intrahepatic blood vessels, bile ducts, the liver capsule, and pre-defined anatomical risk areas. For candidate radiofrequency ablation path data that passes the anatomical safety assessment, the processor further determines whether its thermal release axis is oriented towards the cooling control area. The thermal release axis can be determined based on the axial direction of the radiofrequency electrode, the long axis direction of the predicted thermal effect area, or the main extension direction of the thermal field.

[0059] When the thermal release axis faces the cooling dominance area, the processor determines that the candidate radiofrequency ablation path is beneficial for heat compensation in the cooling dominance area. When the thermal release axis moves away from the cooling dominance area, the processor determines whether the cooling dominance area is located downstream of the blood flow heat zone of the corresponding associated vessel segment. If the cooling dominance area is located downstream of the blood flow heat zone, it indicates that the area is more susceptible to the influence of blood flow heat, and the thermal release axis moving away from this area is more likely to lead to insufficient ablation coverage. The processor marks this candidate radiofrequency ablation path data as a downstream obstruction path and assigns it a higher degradation level than non-downstream obstruction paths. Figure 4 The priority changes of multiple candidate radiofrequency ablation paths before and after degradation are shown. Among them, the candidate paths marked as downstream obstruction paths have significantly decreased scores after degradation and are below the recommended screening line, which is used to reflect the constraint effect on the ranking results when the heat release axis deviates from the cooling dominance area.

[0060] The processor comprehensively sorts multiple candidate radiofrequency ablation path data based on the anatomical safety assessment results, cooling dominance area coverage, needle tip offset coverage increment, thermal release axis orientation, and downstream obstruction path marking results, and outputs the optimized radiofrequency ablation path results. The optimized radiofrequency ablation path results include the needle entry point, needle entry direction, needle entry depth, needle tip planned position, corresponding cooling dominance area, predicted thermal effect area, and sorting results of the recommended path, enabling the output path to simultaneously consider anatomical safety and thermal coverage compensation capability of the cooling dominance area.

[0061] Example 2:

[0062] In this embodiment, after generating candidate radiofrequency ablation path data and determining anatomical safety, the processor reads the set of candidate radiofrequency ablation path data that has passed the anatomical safety determination. The candidate radiofrequency ablation path data includes the needle insertion point, insertion direction, insertion depth, planned needle tip position, and the predicted thermal effect area corresponding to the candidate path. The processor maps the predicted thermal effect area, cooling distribution area, and intrahepatic vascular data to the same spatial coordinate system to further determine the thermal effectiveness of the candidate paths.

[0063] For each candidate radiofrequency ablation path that passes the anatomical safety assessment, the processor determines its corresponding predicted thermal effect area. The predicted thermal effect area can be determined based on the effective action segment location of the radiofrequency electrode, the planned needle tip location, the parameters of the radiofrequency ablation device, the tissue thermal parameters, and the hemodynamic simulation results. It represents the spatial range in which the candidate path can form an effective thermal effect under preset ablation conditions.

[0064] The processor determines whether there exists a related vascular segment between the predicted heat-affected area and the cooling-controlled area that satisfies a preset blood flow heat dissipation criterion. The preset blood flow heat dissipation criterion includes that the related vascular segment is located on the main path from the predicted heat-affected area to the cooling-controlled area, and that the blood flow in the related vascular segment reaches a preset heat dissipation threshold. When the above conditions are met, the processor determines that the predicted heat-affected area and the cooling-controlled area are separated by the related vascular segment.

[0065] When the processor determines that the predicted thermal effect area and the cooling dominance area are separated by an associated vascular segment, it reduces the thermal effectiveness score or priority of the candidate radiofrequency ablation path data. This is because, although the predicted thermal effect area may be spatially close to the cooling dominance area, the associated vascular segment located between them carries away some heat through blood flow, weakening the heat transfer to the cooling dominance area and thus reducing the candidate path's ability to compensate for the cooling dominance area.

[0066] In one specific implementation, the processor uses a thermal effectiveness score adjustment method to downgrade candidate paths. The score adjustment formula is as follows:

[0067]

[0068] in, This indicates the adjusted thermal effectiveness score. This indicates the thermal effectiveness score before adjustment. This represents the downgraded value caused by the separation of associated vessel segments. The derivation logic of this formula is that when a heat transfer path is blocked by a vessel segment that meets the heat dissipation criterion, the original thermal effectiveness score needs to be reduced by the separation penalty value to reflect the insufficient thermal compensation of the candidate path for the cooling dominance area.

[0069] For example, in a certain de-identified case dataset, the original thermal validity score of candidate path A. The score was 86. The processor determined that there was a hepatic vein branch that met the preset blood flow heat dissipation criterion between the predicted heat-affected area and the cooling-dominant area corresponding to candidate path A. This hepatic vein branch was located on the heat transfer path, so the score was downgraded. Set the score to 18 points. The processor substitutes the above data into the scoring adjustment formula to obtain:

[0070]

[0071] As a result, the thermal effectiveness score of candidate path A decreased from 86 points to 68 points, and the processor accordingly reduced the recommendation priority of candidate path A in the candidate radiofrequency ablation path dataset. Figure 5 Using the pre-adjustment and post-adjustment scores of candidate path A as a comparison, and setting a recommendation filter line, the score changes after the partitioning and downgrading correspond to the ranking criteria; from Figure 5 It can be seen that candidate path A dropped from 86 points to 68 points, falling below the recommended screening line of 70 points, indicating that the path separated by the associated vascular segment is subject to a downgrade constraint in the thermal effectiveness ranking.

[0072] When the predicted thermal effect area and cooling dominance area are separated by an associated vessel segment, the processor further searches the already generated candidate radiofrequency ablation path data set to determine if there are alternative candidate radiofrequency ablation path data. Alternative candidate radiofrequency ablation path data refers to candidate paths whose predicted thermal effect area and cooling dominance area are located on the same side of the associated vessel segment, so that heat transfer to the cooling dominance area does not need to cross the associated vessel segment, thereby reducing the impact of blood flow heat dissipation on the thermal coverage effect.

[0073] The processor can determine the lateral positions of the predicted thermal effect area and cooling dominance area relative to the associated vessel segment based on the vessel centerline, local orientation of the vessel segment, and corresponding spatial division plane. When the representative point of the predicted thermal effect area and the representative point of the cooling dominance area are located in the same lateral region of the associated vessel segment, the processor determines that the alternative candidate radiofrequency ablation path data satisfies the same-side condition. The representative point can be the center point of the predicted thermal effect area, the center point of the cooling dominance area, or a boundary sampling point related to the direction of heat transfer.

[0074] For example, the processor retrieves an alternative candidate path B from the candidate radiofrequency ablation path dataset. The predicted thermal effect area and the cooling dominance area corresponding to candidate path B are located on the same side of the same hepatic vein branch, and heat transfer to the cooling dominance area does not need to cross this hepatic vein branch. Based on this, the processor increases the priority of alternative candidate path B, making it appear earlier in the radiofrequency ablation path optimization results compared to candidate path A, which is separated by blood vessels. Figure 6 The spatial relationship between candidate paths and associated vascular segments is represented by the horizontal and vertical positions of local planning. In path A, the predicted thermal effect area and the cooling dominance area are located on both sides of the associated vascular segment, and the heat transfer path is blocked by the blood vessel. In path B, the predicted thermal effect area and the cooling dominance area are located on the same side of the associated vascular segment, which can form a same-side transfer relationship without crossing the vascular segment. Therefore, it can be used as an alternative candidate path for priority improvement.

[0075] When multiple alternative candidate radiofrequency ablation paths exist, the processor assesses the stability of each alternative candidate path within a preset tip deviation range. Specifically, the processor uses the tip planning position of the alternative candidate path as a reference, sets multiple deviation simulation positions within a preset tip deviation range of ±2.0 mm, and redetermines the predicted thermal effect area at each deviation simulation position. Then, it determines whether the redetermined predicted thermal effect area and the cooling control area are still located on the same side of the associated blood vessel segment.

[0076] In one specific implementation, the processor uses the same-side stability ratio as an alternative candidate path to evaluate the tolerance of tip positioning errors. The stability ratio is calculated as follows:

[0077]

[0078] in, Indicates the proportion of ipsilateral stability. This indicates the number of simulations performed while maintaining the predicted thermal effect area and the cooling distribution area on the same side of the associated blood vessel segment within the preset needle tip deviation range. This represents the total number of deviation simulations. The derivation logic of this formula is that the stability ratio is equal to the proportion of deviation samples that satisfy the same-side condition to the total number of deviation samples, and is used to evaluate the ability of alternative candidate paths to maintain the same-side heat transfer relationship under the tip positioning error.

[0079] For example, the processor sets nine deviation simulation positions for alternative candidate path B. In eight of these deviation simulation positions, the predicted thermal effect area and cooling dominance area are still located on the same side as the hepatic vein branch. The processor substitutes this data into the stable proportion calculation formula to obtain:

[0080]

[0081] Since the stability ratio threshold is 0.75, and the same-side stability ratio of alternative candidate path B is 0.89, which is higher than the stability ratio threshold, the processor determines that alternative candidate path B has good same-side stability within the preset tip deviation range, and further increases the priority of alternative candidate path B. Figure 7 The ipsilateral determination status of the nine simulated needle tip deviation positions is discretized and displayed. Eight simulated positions remain ipsilateral, and one simulated position does not remain ipsilateral. An ipsilateral stability ratio of 0.89 and a stability ratio threshold of 0.75 are simultaneously provided, allowing the processor to determine whether alternative candidate path B has sufficient needle tip deviation tolerance. If another alternative candidate path no longer maintains the predicted thermal effect area and cooling dominance area on the same side of the associated blood vessel segment within the preset needle tip deviation range, the processor will not perform priority enhancement processing for that alternative candidate path.

[0082] The processor comprehensively ranks candidate radiofrequency ablation path data based on whether candidate paths are separated by associated vascular segments that meet preset blood flow and heat dissipation criteria, whether there are ipsilateral alternative candidate paths, and the ipsilateral stability ratio of alternative candidate paths within a preset needle tip deviation range. It then outputs the optimized radiofrequency ablation path results. The optimized results include the candidate path's needle entry point, needle entry direction, planned needle tip position, predicted thermal effect area, correspondence with the cooling control area, thermal effectiveness score, and ranking result. This approach, while ensuring anatomical safety, further reduces the priority of candidate paths separated by blood vessels and lacking sufficient thermal compensation, and increases the priority of alternative candidate paths with ipsilateral thermal coverage of the cooling control area and good needle tip deviation stability.

Claims

1. A method for optimizing the radiofrequency ablation path for liver tumors based on hemodynamic simulation, characterized in that, Executed by the processor, the process includes: acquiring acquired liver image data, tumor region data, intrahepatic vascular data and hemodynamic simulation results; determining the outer boundary based on the tumor region; and determining the outer boundary region that meets the preset blood flow heat dissipation criteria as the cooling control zone based on the hemodynamic simulation results. Candidate radiofrequency ablation path data is generated based on the cooling control area, and the anatomical safety assessment is performed on the candidate radiofrequency ablation path data to obtain the anatomical safety assessment result. The candidate radiofrequency ablation path data that has passed the anatomical safety determination is thermally effective. The thermal effectiveness determination includes determining whether the predicted thermal effect area corresponding to the candidate radiofrequency ablation path data covers the cooling dominance area. The candidate radiofrequency ablation path data is sorted according to the anatomical safety determination result and the thermal effectiveness determination result, and the radiofrequency ablation path optimization result is output.

2. The method according to claim 1, characterized in that, The preset blood flow heat dissipation criterion is as follows: when the associated blood vessel segment is located between the preset heat source reference position and the outer expansion boundary region, and the blood flow of the associated blood vessel segment reaches the preset heat dissipation threshold, the outer expansion boundary region is determined as the cooling control area formed by blood flow obstruction.

3. The method according to claim 1, characterized in that, When generating candidate radiofrequency ablation path data, the offset direction of the needle tip planning position is determined according to the position of the cooling dominance area relative to the tumor region and the blood flow direction of the associated vascular segment, and candidate radiofrequency ablation path data is generated based on the needle tip planning position.

4. The method according to claim 1, characterized in that, In the thermal effectiveness determination, it is determined whether the predicted thermal effect area corresponding to the candidate radiofrequency ablation path data and the cooling dominance area are separated by the associated vascular segment that meets the preset blood flow heat dissipation criterion; if they are separated, the thermal effectiveness score or priority of the candidate radiofrequency ablation path data is reduced.

5. The method according to claim 3, characterized in that, When multiple cooling control zones exist, the projection position of each cooling control zone in the blood flow direction of the corresponding associated blood vessel segment is determined, and the cooling control zone whose projection position is on the downstream side of the blood flow heat and whose area or volume meets the preset screening conditions is taken as the main compensation zone. The offset direction of the needle tip planning position is determined based on the main compensation zone.

6. The method according to claim 3, characterized in that, After determining the offset direction of the needle tip planning position, the offset amplitude of the needle tip planning position is limited; when the coverage increment of the predicted heat effect area on the cooling control area after offset is less than a preset increment threshold, and the distance from the needle tip planning position to the dissection risk area is less than a preset safety threshold, the offset amplitude of the needle tip planning position is reduced.

7. The method according to claim 3, characterized in that, When generating candidate radiofrequency ablation path data based on the needle tip planning position, it is determined whether the heat release main axis corresponding to the candidate radiofrequency ablation path data is oriented towards the cooling control area; if the heat release main axis is away from the cooling control area, the priority of the candidate radiofrequency ablation path data is reduced.

8. The method according to claim 7, characterized in that, Before reducing the priority of the candidate radiofrequency ablation path data, it is determined whether the cooling control area is located downstream of the blood flow heat zone based on the blood flow direction of the associated vessel segment; if so, the candidate radiofrequency ablation path data is marked as a downstream obstruction path and given a higher downgrade magnitude than non-downstream obstruction paths.

9. The method according to claim 4, characterized in that, When the predicted thermal effect area and the cooling dominance area are separated by the associated vascular segment, it is determined whether there is alternative candidate radiofrequency ablation path data that places the predicted thermal effect area and the cooling dominance area on the same side; if so, the priority of the alternative candidate radiofrequency ablation path data is increased.

10. The method according to claim 9, characterized in that, When multiple alternative candidate radiofrequency ablation path data exist, it is determined whether the predicted thermal effect area and the cooling dominance area are still located on the same side of the associated blood vessel segment within a preset needle tip deviation range. If they are still on the same side, the priority of the corresponding alternative candidate radiofrequency ablation path data is increased. If they are no longer on the same side, the priority increase processing for the corresponding alternative candidate radiofrequency ablation path data is not performed.

Images