Left atrium partitioning methods, devices and applications
By automatically identifying key feature structures in electrophysiological mapping data and constructing three-dimensional spatial rules, the standardization and repeatability issues of left atrial zoning in existing technologies have been solved, enabling standardized and refined evaluation of electrophysiological analysis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEKING UNION MEDICAL COLLEGE HOSPITAL
- Filing Date
- 2026-02-11
- Publication Date
- 2026-05-26
Smart Images
Figure CN122075010A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical image processing technology, and more particularly to a method for partitioning the left atrium, a partitioning device for the left atrium, and the application of the partitioning device for the left atrium. Background Technology
[0002] The left atrium, as the primary site of arrhythmias such as atrial fibrillation, has an anatomical structure closely related to the formation of the arrhythmia matrix. Currently, clinical and research analyses of the left atrium primarily rely on two technical approaches:
[0003] 1. Anatomical imaging-based zoning: This method relies heavily on cardiac computed tomography (CT) or cardiac magnetic resonance imaging (MRI). These techniques can non-invasively reconstruct high-precision three-dimensional anatomical models of the left atrium and have led to zoning schemes based on anatomical landmarks such as the pulmonary vein orifice and mitral valve annulus.
[0004] 2. Electrophysiological mapping-based assessment: The cardiac three-dimensional electrophysiological mapping system can acquire the spatial coordinates of points on the endocardial surface and their corresponding local potential information (voltage, activation time, etc.) in real time, forming a functional "voltage map" or "activation map". However, current analyses mostly focus on the global assessment of the entire atrium (such as the area of the overall low voltage region) or the simple reference to a pre-set anatomical line.
[0005] However, the drawbacks of existing technologies are as follows:
[0006] 1. Disconnect between functional and anatomical analysis: Existing anatomical imaging zoning standards cannot be directly applied to mapping point cloud data containing electrophysiological information. Abnormal areas observed by physicians on electrophysiological systems are difficult to quickly and systematically attribute to a specific anatomical subregion, resulting in ambiguous lesion localization descriptions.
[0007] 2. Lack of a specific electrophysiological zoning standard: The field of electrophysiological research lacks a widely accepted standard for left atrial zoning. This makes it impossible to conduct standardized comparisons and meta-analyses of data from different centers and studies, severely hindering the formation of scientific consensus.
[0008] 3. Subjective and inefficient partitioning methods: In current practice, if partitioning analysis is required, the operator usually draws the boundaries on the 3D model manually. This method is highly dependent on the operator's experience and subjective judgment, has poor repeatability, is time-consuming and laborious, and cannot achieve batch processing of large-scale data, which restricts the development of refined clinical research. Summary of the Invention
[0009] To overcome the shortcomings of existing technologies, the technical problem to be solved by this invention is to provide a method for partitioning the left atrium, which can achieve geographic calibration of electrophysiological analysis, completely avoid the subjective bias and intra- / inter-observer variation of manual partitioning, achieve 100% reproducibility, greatly improve the reliability and analysis efficiency of scientific research data, and can easily perform independent in-depth analysis of each subregion of the left atrium, realizing regional and refined depiction of the atrial fibrillation matrix.
[0010] The technical solution of this invention is: a method for partitioning the left atrium, which includes the following steps: (1) Import high-density mapping point cloud data exported from mainstream electrophysiological systems, including electrical parameters such as X, Y, Z coordinates, voltage, and activation time; (2) Identify the key feature structures in the point cloud data. The key feature structures include the boundary of the left pulmonary vein ablation circle, the boundary of the right pulmonary vein ablation circle, the boundary of the mitral valve annulus region, and the boundary of the left atrial appendage region, which are used as the basic anchor points for partitioning. (3) Construct three-dimensional spatial geometric rules and assign each data point to one of the anatomical regions without ambiguity. The anatomical regions include: anterior wall, lateral wall, superior wall, septal wall, inferior wall, and posterior wall. (4) Output a new point cloud dataset with clear region labels and generate statistical reports for each region, or visualize the 3D models of the regions with different color codes.
[0011] The beneficial technical effects of this invention compared to the prior art are as follows:
[0012] 1. Geographic calibration of electrophysiological analysis was achieved: For the first time, a standardized "geographic coordinate system" was provided for messy electrophysiological mapping point cloud data, making descriptions such as "low voltage in the anterior wall" or "slow conduction in the lateral wall" objective, quantifiable and comparable indicators.
[0013] 2. Fully automatic and highly consistent partitioning: It completely avoids the subjective bias and intra- / inter-observer variation of manual partitioning, achieving 100% repeatability and greatly improving the reliability and analysis efficiency of scientific research data.
[0014] 3. Enables refined matrix assessment: Based on automated zoning, independent in-depth analysis of each subregion of the left atrium can be easily performed. For example, the average voltage of the anterior wall versus the posterior wall can be calculated separately, and the conduction velocity dispersion of the superior wall versus the inferior wall can be compared. This enables a regionalized and refined depiction of the atrial fibrillation matrix.
[0015] A partitioning device for the left atrium is also provided, which includes: The data input module is configured to import high-density mapping point cloud data exported from mainstream electrophysiological systems, including electrical parameters such as X, Y, Z coordinates, voltage, and activation time. The data recognition module is configured to identify key feature structures in point cloud data. These key feature structures include the boundaries of the left pulmonary vein ablation zone, the right pulmonary vein ablation zone, the mitral valve annulus region, and the left atrial appendage region, which serve as the basic anchor points for partitioning. The building module is configured to construct three-dimensional spatial geometric rules, unambiguously assigning each data point to one of the anatomical regions, including: anterior wall, lateral wall, superior wall, septal wall, inferior wall, and posterior wall. The output module is configured to output new point cloud datasets with clearly labeled regions and generate statistical reports for each region, or visualize 3D models of regions with different color codes.
[0016] It also provides the application of a left atrial partitioning device, which is used in electrophysiological detection and analysis.
[0017] It also provides the application of a zoning device for the left atrium, characterized in that it is used in the regional and refined depiction of the atrial fibrillation matrix. Attached Figure Description
[0018] Figure 1 This is a flowchart of the left atrium partitioning method according to the present invention.
[0019] Figure 2 This is a schematic diagram of the partitioning method of the left atrium according to the present invention. Detailed Implementation
[0020] This invention proposes a standardized left atrial zoning scheme and a fully automated implementation system specifically designed for electrophysiological mapping data. The core of the scheme includes "one standard" and "one engine".
[0021] Based on an in-depth literature review and consensus among electrophysiology experts at Peking Union Medical College Hospital, the inventors abandoned the approach of solely relying on anatomical imaging and proposed a six-zone standard that is more in line with the perspective of electrophysiological mapping and clinical ablation practice. This standard systematically divides the left atrium into: anterior wall, lateral wall, superior wall, septal wall, inferior wall, and posterior wall.
[0022] like Figure 1 As shown, this method of dividing the left atrium includes the following steps: (1) Import high-density mapping point cloud data exported from mainstream electrophysiological systems, including electrical parameters such as X, Y, Z coordinates, voltage, and activation time; (2) Identify the key feature structures in the point cloud data. The key feature structures include the boundary of the left pulmonary vein ablation circle, the boundary of the right pulmonary vein ablation circle, the boundary of the mitral valve annulus region, and the boundary of the left atrial appendage region, which are used as the basic anchor points for partitioning. (3) Construct three-dimensional spatial geometric rules and assign each data point to one of the anatomical regions without ambiguity. The anatomical regions include: anterior wall, lateral wall, superior wall, septal wall, inferior wall, and posterior wall. (4) Output a new point cloud dataset with clear region labels and generate statistical reports for each region, or visualize the 3D models of the regions with different color codes.
[0023] The beneficial technical effects of this invention compared to the prior art are as follows:
[0024] 1. Geographic calibration of electrophysiological analysis was achieved: For the first time, a standardized "geographic coordinate system" was provided for messy electrophysiological mapping point cloud data, making descriptions such as "low voltage in the anterior wall" or "slow conduction in the lateral wall" objective, quantifiable and comparable indicators.
[0025] 2. Fully automatic and highly consistent partitioning: It completely avoids the subjective bias and intra- / inter-observer variation of manual partitioning, achieving 100% repeatability and greatly improving the reliability and analysis efficiency of scientific research data.
[0026] 3. Enables refined matrix assessment: Based on automated zoning, independent in-depth analysis of each subregion of the left atrium can be easily performed. For example, the average voltage of the anterior wall versus the posterior wall can be calculated separately, and the conduction velocity dispersion of the superior wall versus the inferior wall can be compared. This enables a regionalized and refined depiction of the atrial fibrillation matrix.
[0027] The reason why this invention can produce this advantage is:
[0028] 1. Precision of problem definition: This invention directly addresses the pain point of the lack of dedicated analytical tools in electrophysiological research. It designs the zoning standard from the application scenario (electrophysiological mapping) rather than simply applying the concept of imaging zoning, thus ensuring the practicality and relevance of the standard.
[0029] 2. Integration of clinical expert knowledge: The establishment of standards deeply integrates the clinical practice experience of the electrophysiology center, ensuring that the zoning is not only geometrically reasonable, but also has clear electrophysiological and clinical significance, so that the analysis results can be fed back into treatment decisions.
[0030] 3. Closed-loop design of algorithms and standards: "Standards" provide clear mathematical definitions and rules for algorithms, while "algorithms" are automated tools that strictly execute those standards. This close integration of "soft standards" and "hard algorithms" is fundamental to addressing subjectivity and improving efficiency.
[0031] Furthermore, in step (2), four marker points are defined in the mitral valve annulus region, namely at the 4 o'clock, 7 o'clock, 11 o'clock and 12 o'clock positions; one marker point is defined in the left atrial appendage region, namely at the 11 o'clock position; four marker points are defined in the right pulmonary vein ablation circle, namely at the 1 o'clock, 3 o'clock, 6 o'clock and 11 o'clock positions; and three marker points are defined in the left pulmonary vein ablation circle, namely at the 1 o'clock, 6 o'clock and 11 o'clock positions.
[0032] Furthermore, in step (3), the three-dimensional spatial geometric rules are as follows: 12 marker points are set, 6 regions are defined, and a combination of several marker points represents a boundary; as shown in Table 1, the marker points include: MA 4 o'clock position marker point is MA4, MA 7 o'clock position marker is MA7, MA 11 o'clock position marker is MA11, MA 12 o'clock position marker is MA12, The LAA 11 o'clock position marker is LAA 11. 、 The LPV 6 o'clock position marker is LPV 6. 、 The 6 o'clock position marker on RPV is RPV 6. 、 RPV 11 o'clock position marker is RPV-p, LPV 1 o'clock position marker is LPV-p, RPV 1 o'clock position marker is RPV-a, LVP 11 o'clock position marker is LPV-a, RPV 3 o'clock position marker is RPV-f .
[0033] Table 1
[0034] Furthermore, in step (3), for each of the four boundaries, all points are first fitted to the optimal plane, and then the points on the plane are fitted into an ellipse to determine the point closest to a certain clock direction; the entire analysis process maintains a one-to-one mapping between data points and finally maps back to the original data points; after defining the marker points, the regions are further divided accordingly, and each region is determined by a set of specific marker points.
[0035] Furthermore, in step (3), the boundary of the front wall region is defined by five marker points: MA11, MA12, LAA11, RPV-a, and RPV-f; the boundary of the interval region is defined by... MA7 , MA11 , RPV-f , RPV6 The boundary of the sidewall region is defined by five marker points; LPV-a , LAA11 ,MA12 , MA4 , LPV6 The boundary of the top wall region is defined by five marker points; RPV- p , RPV-a , LAA11 , LPV-a , LPV-p The boundary of the rear wall region is defined by five marker points; LPV-p , RPV-p , LPV6 , RPV6 The boundary of the lower wall region is defined by five marker points; MA4 , MA7 , LPV6 , RPV6 The five markers are defined together. See Table 2 for details.
[0036] Table 2
[0037] Furthermore, in step (3), the three-dimensional data points of the left atrium are spherically fitted to obtain the center of the sphere O; based on the center of the sphere O and any two marker points, a set of spatial planes are constructed; the spatial boundaries of each anatomical region are jointly defined by a set of planes.
[0038] Furthermore, in step (3), the set of marked points for the interval region is {MA7, MA11, RPV-f, RPV6}, and the corresponding boundary planes include: O-MA7-MA11, O-MA11-RPV-f, O-MA7-RPV6 and O-RPV-f-RPV6; if a point in space is located within the quadrangular pyramid region enclosed by these four planes, it is determined to belong to the interval wall, and the corresponding partition information will be recorded in the data table.
[0039] A partitioning device for the left atrium is also provided, which includes: The data input module is configured to import high-density mapping point cloud data exported from mainstream electrophysiological systems, including electrical parameters such as X, Y, Z coordinates, voltage, and activation time. The data recognition module is configured to identify key feature structures in point cloud data. These key feature structures include the boundaries of the left pulmonary vein ablation zone, the right pulmonary vein ablation zone, the mitral valve annulus region, and the left atrial appendage region, which serve as the basic anchor points for partitioning. The building module is configured to construct three-dimensional spatial geometric rules, unambiguously assigning each data point to one of the anatomical regions, including: anterior wall, lateral wall, superior wall, septal wall, inferior wall, and posterior wall. The output module is configured to output new point cloud datasets with clearly labeled regions and generate statistical reports for each region, or visualize 3D models of regions with different color codes.
[0040] It also provides the application of a left atrial partitioning device, which is used in electrophysiological detection and analysis.
[0041] It also provides the application of a zoning device for the left atrium, characterized in that it is used in the regional and refined depiction of the atrial fibrillation matrix.
[0042] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.
Claims
1. A method of partitioning the left atrium, characterized by: It comprises the following steps: (1) importing high-density mapping point cloud data derived from mainstream electrophysiology system, containing X, Y, Z coordinates, voltage, activation time electrical parameters; (2) identifying key feature structures in point cloud data, key feature structures including left pulmonary vein ablation ring boundary, right pulmonary vein ablation ring boundary, mitral annulus region boundary, left atrial appendage region boundary, taking this as the basis anchor point of partition; (3) constructing three-dimensional space geometric rules, unambiguously attributing each data point to one of the anatomical regions, including: anterior wall, lateral wall, roof, septal wall, inferior wall, posterior wall; (4) outputting new point cloud data set with clear region label, and generating statistical report of each partition, or visualizing and displaying partition three-dimensional model with different color coding.
2. The method of partitioning the left atrium of claim 1, wherein: In step (2), the mitral annulus region defines 4 marker points, respectively 4 o'clock position, 7 o'clock position, 11 o'clock position and 12 o'clock position; the left atrial appendage region defines 1 marker point, which is 11 o'clock position; The right pulmonary vein ablation ring defines 4 marker points, respectively 1 o'clock position, 3 o'clock position, 6 o'clock position and 11 o'clock position; the left pulmonary vein ablation ring defines 3 marker points, respectively 1 o'clock position, 6 o'clock position and 11 o'clock position.
3. The method of partitioning the left atrium of claim 2, wherein: In the step (3), the three-dimensional space geometry rules are: 12 marker points are set, 6 regions are defined, and a boundary is represented by a combination of several marker points; the marker points include: MA 4 o'clock azimuth marker point is MA4、 MA 7 o'clock azimuth marker point is MA7、 MA 11 o'clock azimuth marker point is MA11、 MA 12 o'clock azimuth marker point is MA12、 LAA 11 o'clock azimuth marker point is LAA 11 、 LPV 6 o'clock azimuth marker point is LPV6 、 RPV 6 o'clock azimuth marker point is RPV 6 、 RPV 11 o'clock azimuth marker point is RPV-p、 LPV 1 o'clock azimuth marker point is LPV-p、 RPV 1 o'clock azimuth marker point is RPV-a、 LVP 11 o'clock azimuth marker point is LPV-a、 RPV 3 o'clock azimuth marker point is RPV-f .
4. The method of partitioning the left atrium of claim 3, wherein: In step (3), for each of the four boundaries, first fit all points to the optimal plane, and then fit the points on the plane into an ellipse, so as to determine the point closest to the o'clock direction; The whole analysis process maintains one-to-one mapping between data points, and finally maps back to the original data points; after defining the marker points, the region is further divided according to the marker points, and each region is determined by a group of specific marker points.
5. The method of partitioning the left atrium of claim 4, wherein: In the step (3), the boundary of the front wall region is defined by the five mark points of MA11, MA12, LAA11, RPV-a, RPV-f; the boundary of the interval region is defined by the five mark points of MA7 , MA11 , RPV-f , RPV6 the five mark points of RPV-b, RPV-c, RPV-d, RPV-e, LAA12; the boundary of the side wall region is defined by the five mark points of LPV-a , LAA11 , MA12 , MA4 , LPV6 the five mark points of RPV-g, RPV-h, RPV-i, RPV-j, LAA13; the boundary of the top wall region is defined by the five mark points of RPV-p , RPV-a , LAA11 , LPV-a , LPV-p the five mark points of RPV-k, RPV-l, RPV-m, RPV-n, LAA14; the boundary of the back wall region is defined by the five mark points of LPV-p , RPV-p , LPV6 , RPV6 the five mark points of RPV-o, RPV-p, RPV-q, RPV-r, LAA15; the boundary of the lower wall region is defined by the five mark points of MA4 , MA7 , LPV6 , RPV6 the five mark points of RPV-s, RPV-t, RPV-u, RPV-v, LAA16.
6. The method of partitioning the left atrium of claim 5, wherein: In step (3), spherical fitting is performed on the left atrial three-dimensional data points to obtain the sphere center O; based on the sphere center O and any two marker points, a group of space planes are constructed; the spatial boundaries of each anatomical partition are jointly defined by a group of planes.
7. The method of partitioning the left atrium of claim 6, wherein: In step (3), the marker point set of the septal region is {MA7, MA11, RPV-f, RPV6}, and the corresponding boundary planes include: O-MA7-MA11, O-MA11-RPV-f, O-MA7-RPV6 and O-RPV-f-RPV6; if a point in space is located in the four-prism region surrounded by the four planes, it is determined to belong to the septal wall, and the corresponding partition information will be recorded in the data table.
8. Apparatus for partitioning the left atrium, characterized by: It comprises: A data input module configured to import high-density mapping point cloud data derived from mainstream electrophysiology system, containing X, Y, Z coordinates, voltage, activation time electrical parameters; A data recognition module configured to recognize key feature structures in the point cloud data, the key feature structures including a left pulmonary vein ablation ring boundary, a right pulmonary vein ablation ring boundary, a mitral annulus region boundary, and a left atrial appendage region boundary, as a basis anchor point for partitioning; A construction module configured to construct three-dimensional spatial geometric rules, and unambiguously attribute each data point to one of the anatomical regions, the anatomical regions including: an anterior wall, a lateral wall, an apical wall, a septal wall, an inferior wall, and a posterior wall; An output module configured to output a new point cloud data set with clear region labels, and generate a statistical report of each partition, or visually display a partitioned three-dimensional model with different color coding.
9. Use of the device for partitioning the left atrium according to claim 8, characterized in that: It is applied to electrophysiological detection and analysis.
10. Use of the device for partitioning the left atrium according to claim 8, characterized in that: It is applied to the regionalization and fine depiction of atrial fibrillation substrates.