Animal respiratory heartbeat double-coupling CBCT dynamic reconstruction method and system

Abstract

The application discloses an animal respiration and heartbeat double-coupling CBCT dynamic reconstruction method and system, and the method comprises the following steps: acquiring an anatomical region mask and a time sequence alignment data set output by a multi-source synchronous acquisition device; determining a phase according to a respiration waveform and an electrocardio waveform, configuring a composite phase label for CBCT projection, and generating a composite phase data set; solving a respiration global deformation field sequence according to the composite phase data set, compensating for the CBCT projection to solve a heartbeat local deformation field sequence, and forming a double-physiological deformation field group; generating a double-coupling total deformation field sequence according to the double-physiological deformation field group and the anatomical region mask; and according to the composite phase label and the double-coupling total deformation field sequence, performing motion compensation reconstruction and time sequence optimization on the CBCT projection according to the composite phase, and outputting a four-dimensional dynamic image sequence. The application is beneficial to stably presenting dynamic details of structures such as animal myocardium, large blood vessels and pulmonary hilum.

Description

Technical Field

[0001] This application relates to the fields of medical image processing and cone-beam computed tomography (CBCT), and in particular to a method and system for dynamic reconstruction of animal respiratory and cardiac rhythm using dual-coupled CBCT. Background Technology

[0002] Four-dimensional cone-beam computed tomography (CBCT) of the thoracic and abdominal regions of animals involves continuously acquiring X-ray projections from multiple angles and reconstructing volumetric data with a time dimension according to physiological movement states. This can be used to observe the dynamic morphology of tissues such as the thoracic cavity, heart, and major blood vessels. Open scanning can reduce the impact of anesthesia and fixation devices on the animal's physiological state, but spontaneous respiration, changes in body position, and rapid heartbeat will all affect the projected data.

[0003] Existing methods for animal dynamic reconstruction typically employ a single respiratory motion processing approach suited to humans. This approach struggles to adapt to the unstable respiratory cycles, rapid heart rates, and coupled cardiopulmonary movements in dogs and cats. This can easily lead to confusion between different physiological motion states, resulting in distortion estimation, organ boundary jitter, local structural blurring, and frame jumps. Therefore, accurately characterizing and correcting the superimposed motion of respiration and heartbeat in uncooperative, unconstrained animal scans is a critical problem that needs to be solved for stable reconstruction of cardiopulmonary dynamic structures. Summary of the Invention

[0004] This application proposes a method and system for dynamic reconstruction of animal respiratory and cardiac movements using dual-coupled CBCT, aiming to distinguish and coordinately correct respiratory and cardiac movements during uncooperative, unconstrained scanning of animals, thereby improving the stability of the reconstructed dynamic structures of the cardiopulmonary system. This facilitates the stable presentation of dynamic details of structures such as the animal's myocardium, major blood vessels, and pulmonary hilum.

[0005] In a first aspect, embodiments of this application provide a dynamic reconstruction method for animal respiratory and cardiac dual-coupled CBCT, applied to a CBCT reconstruction device communicating with a multi-source synchronous acquisition device, the method comprising:

[0006] Obtain the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device. The time-aligned dataset includes the corrected CBCT projection, respiratory waveform, and electrocardiogram waveform.

[0007] The phase is determined based on the respiratory waveform and the electrocardiogram waveform, and a composite phase label is configured for the CBCT projection to generate a composite phase dataset.

[0008] The global respiratory deformation field sequence is solved based on the composite phase dataset. The global respiratory deformation field sequence is then used to compensate the CBCT projection and the local heartbeat deformation field sequence is solved to form a dual physiological deformation field set.

[0009] A dual-coupled total deformation field sequence is generated based on the dual physiological deformation field group and the anatomical region mask; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region.

[0010] Based on the composite phase label and the dual-coupled total deformation field sequence, motion compensation reconstruction and temporal optimization are performed on the CBCT projection according to the composite phase to output a four-dimensional dynamic image sequence.

[0011] As can be seen, this embodiment co-characterizes the respiratory and cardiac phases of CBCT projection and solves for the global deformation caused by respiration and the local deformation caused by cardiac motion separately. The global respiratory deformation is first used to compensate for the CBCT projection to reduce the interference of respiratory motion on the solution process of the local cardiac deformation. The two types of deformation are further combined according to anatomical regions to prevent the local cardiac deformation from spreading to tissues only affected by respiration, thereby suppressing organ boundary jitter, local blurring, and inter-frame jumps.

[0012] In one possible example, generating a dual-coupled total deformation field sequence based on the dual physiological deformation field set and the anatomical region mask includes:

[0013] The partitioned deformation fields of each composite phase are determined based on the dual physiological deformation field group and the anatomical region mask;

[0014] The partitioned deformation fields of each composite phase are processed to be made into a time-series continuous process to generate the dual-coupled total deformation field sequence.

[0015] As can be seen, this example first completes the spatial partition deformation configuration of each composite phase, and then performs continuous processing on the partition deformation field along the temporal relationship of the composite phase, so that the spatial regional differences and temporal change continuity can be constrained respectively, thereby reducing the deformation abrupt changes between adjacent composite phases.

[0016] In one possible example, determining the partitioned deformation fields of each composite phase based on the dual physiological deformation field set and the anatomical region mask includes:

[0017] The regional deformation configuration of each composite phase is determined based on the dual physiological deformation field group and the anatomical region mask;

[0018] The dual physiological deformation field group is partitioned and fused according to the regional deformation configuration of each composite phase to determine the partitioned deformation field of each composite phase.

[0019] As can be seen, this example determines the regional deformation configuration corresponding to each composite phase before performing partition fusion, which can make the regions, deformation types and composite phases involved in the fusion correspond to each other, reducing the possibility of deformation fields of different regions or different phases being incorrectly combined.

[0020] In one possible example, determining the regional deformation configuration of each composite phase based on the dual physiological deformation field set and the anatomical region mask includes:

[0021] Based on the dual physiological deformation field group and the anatomical region mask, regional deformation allocation information for each composite phase is generated;

[0022] The regional deformation configuration of each composite phase is determined based on the regional deformation allocation information of each composite phase.

[0023] As can be seen, this example uses regional deformation allocation information to carry the correspondence between anatomical regions and candidate deformation fields, and then converts this correspondence into regional deformation configurations that can control partition fusion, which helps to improve the interpretability and traceability of the deformation selection process.

[0024] In one possible example, generating the regional deformation allocation information for each composite phase based on the dual physiological deformation field group and the anatomical region mask includes:

[0025] Based on the dual physiological deformation field group and the anatomical region mask, each mask voxel is associated with the global respiratory deformation field and the local cardiac deformation field corresponding to each composite phase to establish voxel deformation association data;

[0026] The region deformation allocation information for each composite phase is determined based on the voxel deformation correlation data and the anatomical region mask.

[0027] As can be seen, this example uses masked voxels as the unit of association between deformation fields and anatomical regions. This allows composite phase, voxel location, region category, and corresponding deformation field tissue to be in the same association, enabling the allocation of regional deformation to be implemented at the voxel level.

[0028] In one possible example, determining the region deformation allocation information for each composite phase based on the voxel deformation association data and the anatomical region mask includes:

[0029] Based on the voxel deformation association data and the anatomical region mask, the region category of each mask voxel and its associated global respiratory deformation field and local heartbeat deformation field are extracted to form region deformation candidate information;

[0030] Based on the regional deformation candidate information, the corresponding global respiratory deformation field and local heartbeat deformation field are retained in the dual motion coupling region, and the corresponding global respiratory deformation field is retained in the single respiratory motion region, thereby determining the regional deformation allocation information for each composite phase.

[0031] The dual-motion coupling region includes at least one of the heart, great vessels, and pulmonary hilum, and the single respiratory motion region includes at least one of the non-cardiopulmonary region of the thoracic cavity and subcutaneous soft tissue.

[0032] As can be seen, this example filters candidate deformation fields corresponding to voxels by region category, retains two types of deformation in dual motion coupling regions such as the heart, great vessels, and pulmonary hilum, and excludes local deformation of the heartbeat in single respiratory motion regions such as non-cardiopulmonary regions of the thoracic cage and subcutaneous soft tissue, which can limit the scope of the heartbeat motion correction.

[0033] In one possible example, determining the phase based on the respiratory waveform and the electrocardiogram waveform, configuring composite phase labels for the CBCT projection, and generating a composite phase dataset includes:

[0034] Based on the respiratory waveform and the electrocardiogram waveform, frequency band filtering is performed to obtain the low-frequency respiratory component and the high-frequency heartbeat component;

[0035] The peaks and troughs of the low-frequency respiratory component are identified to divide the respiratory phase, and the QRS feature points of the high-frequency heartbeat component are identified to divide the heartbeat phase.

[0036] Based on the temporal correspondence in the temporal alignment dataset, the corresponding respiratory phases and heartbeat phases are combined into composite phase labels and configured for the CBCT projection to generate the composite phase dataset.

[0037] As can be seen, this example extracts physiological components suitable for phase division from the respiratory waveform and the electrocardiogram waveform respectively, and then combines the respiratory phase and heartbeat phase corresponding to the same projection time into a composite phase label, which can reduce the possibility of the two physiological motion states being mixed and represented.

[0038] In one possible example, identifying the peaks and troughs of the low-frequency respiratory component to delineate the respiratory phase, and identifying QRS feature points of the high-frequency heartbeat component to delineate the heartbeat phase, includes:

[0039] Based on the low-frequency respiratory component and the high-frequency heartbeat component, adjacent peaks and troughs and adjacent QRS feature points are detected respectively to form physiological cycle boundary information;

[0040] Based on the physiological cycle boundary information, each respiratory cycle and each heartbeat cycle are continuously segmented to obtain the respiratory phase and the heartbeat phase.

[0041] As can be seen, this example defines each physiological cycle by adjacent physiological feature points, and then makes continuous segments within each cycle, so that the respiratory phase and heartbeat phase can adapt to the changes in the length of the animal's respiratory cycle and the changes in heart rate.

[0042] In one possible example, the step of performing motion-compensated reconstruction and temporal optimization of the CBCT projection according to the composite phase based on the composite phase label and the dual-coupled total deformation field sequence, and outputting a four-dimensional dynamic image sequence, includes:

[0043] Based on the composite phase label and the dual-coupled total deformation field sequence, the CBCT projection is divided into composite phase projection groups, and each composite phase projection group is associated with a corresponding dual-coupled total deformation field to form a phase reconstruction task set;

[0044] Based on the phase reconstruction task set, motion compensation iterative reconstruction is performed under projection data fidelity constraints and deformation smoothing constraints to obtain an initial four-dimensional dynamic image sequence.

[0045] The initial four-dimensional dynamic image sequence is subjected to cross-phase grayscale alignment, temporal continuity constraint, edge-preserving filtering and grayscale normalization to output the four-dimensional dynamic image sequence.

[0046] As can be seen, this example organizes the composite phase projection group and the corresponding dual-coupled total deformation field together as a reconstruction task, and performs unified optimization on the grayscale, edge and temporal continuity of different phases after reconstruction, which is beneficial to maintain the dynamic structure boundary and reduce display fluctuations between consecutive frames.

[0047] Secondly, embodiments of this application provide an animal respiratory and cardiac dual-coupled CBCT dynamic reconstruction system, deployed in a CBCT reconstruction device communicating with a multi-source synchronous acquisition device, comprising:

[0048] The data acquisition module is used to acquire the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device. The dataset includes the corrected CBCT projection, respiratory waveform, and electrocardiogram waveform.

[0049] A phase labeling module is used to determine the phase based on the respiratory waveform and the electrocardiogram waveform, configure composite phase labels for the CBCT projection, and generate a composite phase dataset.

[0050] The deformation solving module is used to solve the global respiratory deformation field sequence based on the composite phase dataset, and to solve the local heartbeat deformation field sequence after compensating the CBCT projection with the global respiratory deformation field sequence to form a dual physiological deformation field group.

[0051] The partition fusion module is used to generate a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region.

[0052] The reconstruction output module is used to perform motion compensation reconstruction and temporal optimization on the CBCT projection according to the composite phase label and the dual-coupled total deformation field sequence, and output a four-dimensional dynamic image sequence.

[0053] Through the data transfer and processing coordination between the modules, the system can execute the aforementioned dual-coupled dynamic reconstruction method in CBCT reconstruction equipment, and form a complete processing path from synchronous data acquisition, composite phase marking, dual physiological deformation solution, partition fusion to dynamic image output.

[0054] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, a communication interface, and one or more programs, the one or more programs being stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps in the first aspect of embodiments of this application.

[0055] Fourthly, embodiments of this application provide a computer storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of this embodiment.

[0056] As can be seen, in this embodiment, compared to the existing animal 4DCBCT reconstruction method that only fits a single respiratory movement, this application co-characterizes the respiratory phase and heartbeat phase of the projection, and processes global low-frequency deformation and local high-frequency deformation separately. The CBCT reconstruction device first acquires the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device; secondly, it determines the phase based on the respiratory waveform and electrocardiogram waveform, configures a composite phase label for the CBCT projection, and generates a composite phase dataset; then, it solves the global respiratory deformation field sequence based on the composite phase dataset, and uses the global respiratory deformation field sequence to compensate for the CBCT projection before solving the local heartbeat deformation field sequence, forming a dual physiological deformation field group; thirdly, it generates a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask; finally, based on the composite phase label and the dual-coupled total deformation field sequence, it performs motion compensation reconstruction and temporal optimization on the CBCT projection according to the composite phase, and outputs a four-dimensional dynamic image sequence. Because global respiratory deformation is first used to isolate local cardiac motion, and the two types of deformation are combined in a composite phase only within the corresponding anatomical region, the cardiac deformation will not spread to the thoracic tissue that is only affected by respiration. This suppresses boundary jitter, local blurring and inter-frame jumps, which is conducive to the stable presentation of dynamic details of animal structures such as myocardium, large blood vessels and hilum. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the hardware architecture for dynamic reconstruction of animal respiration and heartbeat using dual-coupled CBCT, provided in an embodiment of this application.

[0058] Figure 2 This is a schematic diagram of the animal respiratory and cardiac dual-coupled CBCT dynamic reconstruction software architecture provided in the embodiments of this application;

[0059] Figure 3 This is a schematic diagram of an animal uncooperative, unconstrained scanning scenario provided in an embodiment of this application;

[0060] Figure 4 This is a schematic diagram of the animal respiratory and cardiac dual-coupled CBCT dynamic reconstruction method provided in the embodiments of this application;

[0061] Figure 5 This is a schematic diagram illustrating the principle of dual physiological deformation field partitioning combination provided in an embodiment of this application;

[0062] Figure 6 This is a schematic diagram of the process for generating a dual-coupled total deformation field sequence provided in an embodiment of this application;

[0063] Figure 7 This is a schematic diagram of the temporal continuity processing of the partitioned deformation field provided in the embodiments of this application;

[0064] Figure 8 This is a schematic diagram of the process for determining the partitioned deformation field provided in an embodiment of this application;

[0065] Figure 9 This is a schematic diagram of the process for determining the regional deformation configuration provided in an embodiment of this application;

[0066] Figure 10 This is a schematic diagram of the process for generating regional deformation allocation information provided in an embodiment of this application;

[0067] Figure 11 This is a schematic diagram of the process for determining regional deformation allocation information provided in an embodiment of this application;

[0068] Figure 12 This is a schematic diagram of the composite phase dataset generation process provided in the embodiments of this application;

[0069] Figure 13 This is a schematic diagram of the respiratory phase and heartbeat phase division process provided in an embodiment of this application;

[0070] Figure 14 This is a schematic diagram of the motion compensation reconstruction and timing optimization process provided in an embodiment of this application;

[0071] Figure 15 This is a schematic diagram of the functional modules of the animal respiratory and cardiac dual-coupled CBCT dynamic reconstruction system provided in the embodiments of this application. Detailed Implementation

[0072] The technical solutions of this application will be clearly and completely described below with reference to various embodiments. The described embodiments are used to explain this application, and not to limit the scope of protection of this application. Where there is no technical conflict, the technical features in different embodiments can be combined with each other.

[0073] In this application, "CBCT projection" refers to the corrected CBCT projection in a time-aligned dataset. Correction processing may include at least one of detector response correction, geometric deviation correction, abnormal pixel correction, and scattering effect correction. "Time alignment" means that the CBCT projection, respiratory waveform, and ECG waveform have corresponding acquisition times or a unified time reference, and does not require the three types of data to have the same sampling frequency.

[0074] "Anatomy region mask" is data used to identify different anatomical region categories in the space to be reconstructed. The anatomical region mask can be obtained by segmenting reference volume data, or by automatically segmenting, manually correcting, or a combination of both from pre-reconstructed volume data. "Deformation field" is data used to describe the displacement correspondence between voxels or spatial positions in different physiological motion states. "Composite phase" represents the combined state of the respiratory and cardiac phases at the same projection moment.

[0075] See Figure 1 The hardware architecture includes a CBCT imaging device 100, a multi-source synchronous acquisition device 200, and a CBCT reconstruction device 300. The CBCT imaging device 100 includes an X-ray source 110, a detector 120, and a scanning gantry 130; the multi-source synchronous acquisition device 200 includes a projection acquisition component 210, a respiration acquisition component 220, an electrocardiogram acquisition component 230, and a synchronization component 240; the CBCT reconstruction device 300 includes a processor 310, a memory 320, a communication interface 330, and a display 340. The synchronization component 240 configures a unified time reference for projection, respiration, and electrocardiogram data; the communication interface 330 receives time-aligned data; the processor 310 calls the program in the memory 320 to complete the reconstruction; and the dynamic image is output through the display 340.

[0076] See Figure 2 The software architecture includes a data acquisition layer 410, a preprocessing layer 420, a phase processing layer 430, a deformation solving layer 440, a partition fusion layer 450, a reconstruction optimization layer 460, and an image output layer 470. The data acquisition layer 410 receives synchronously acquired data and anatomical region masks; the preprocessing layer 420 performs data correction and quality checks; the phase processing layer 430 generates composite phase data; the deformation solving layer 440 sequentially obtains respiratory deformation and heartbeat deformation; the partition fusion layer 450 generates the total deformation field; the reconstruction optimization layer 460 performs motion-compensated reconstruction and temporal optimization; and the image output layer 470 outputs four-dimensional dynamic images.

[0077] See Figure 3 The animal 510 is positioned within the open, free-position scanning area 520, and its chest and abdomen target area 530 is within the imaging range formed by the X-ray source 110 and the detector 120. The respiratory sensor 540 and ECG electrode 550 acquire respiratory and ECG waveforms respectively. The multi-source synchronous acquisition device 200 outputs the waveform data and CBCT projection to the CBCT reconstruction device 300 at a unified time reference. This scenario does not require the animal to actively hold its breath or maintain a fixed respiratory rhythm, and is suitable for dynamic chest and abdominal examinations of animals such as dogs and cats in their natural physiological state.

[0078] In one scenario, the animal exhibited variations in respiratory cycle length during the scan, while its heart rate cycle was significantly shorter than its respiratory cycle. Instead of directly replacing the physiological phase with a fixed time interval, the CBCT reconstruction equipment determined the phase based on simultaneously acquired respiratory and electrocardiogram waveforms. Therefore, multiple heart rate states within the same respiratory cycle, as well as corresponding heart rate states across different respiratory cycles, can be distinguished and labeled.

[0079] In another scenario, the target observation area includes the heart, great vessels, pulmonary hilum, and adjacent thoracic tissues. The heart and great vessels are simultaneously affected by respiratory and cardiac movements, the pulmonary hilum can simultaneously show pulmonary tissue traction and cardiovascular pulsation, while the chest wall and subcutaneous soft tissues mainly change with respiration. By selecting the deformation fields involved in the combination based on the anatomical regions, the inappropriate application of cardiac deformation to the chest wall or subcutaneous soft tissues can be reduced.

[0080] In one possible embodiment, see Figure 4 The dynamic reconstruction method of animal respiratory and cardiac dual-coupled CBCT can be achieved through the following steps S101 to S105.

[0081] S101, acquire the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device, the time-aligned dataset including the corrected CBCT projection, respiratory waveform and electrocardiogram waveform.

[0082] The anatomical region mask is used to identify the anatomical region to which each mask voxel belongs within the body space to be reconstructed. The anatomical region mask can at least distinguish between regions with dual motion coupling and regions with single respiratory motion, and can further subdivide the heart, great vessels, hilum, non-cardiopulmonary regions of the thoracic cage, and subcutaneous soft tissue. The mask can be generated based on reference images before scanning, or after obtaining pre-reconstructed images.

[0083] A time-aligned dataset is a collection of data organized according to a unified time reference. A multi-source synchronous acquisition device can record the acquisition time of each frame of CBCT projection and the corresponding times for respiratory waveform sampling points and electrocardiogram waveform sampling points. When different data sources have different sampling frequencies, a time-series correspondence can be established based on timestamps, synchronization trigger signals, or a combination of both.

[0084] The CBCT projection specifically refers to the corrected CBCT projection in the time-aligned dataset. The purpose of correction is to reduce the impact of imaging device errors on subsequent phase grouping and deformation solving. Respiratory waveforms are used to reflect the animal's inspiration, expiration, and transitional states, while electrocardiogram waveforms are used to reflect the electrophysiological timing within the cardiac cycle.

[0085] S102, determine the phase based on the respiratory waveform and the electrocardiogram waveform, configure composite phase labels for the CBCT projection, and generate a composite phase dataset.

[0086] The respiratory phase characterizes the respiratory motion state corresponding to the CBCT projection, while the heartbeat phase characterizes the cardiac state corresponding to that projection. The composite phase label is composed of corresponding respiratory and heartbeat phase identifiers. For example, a single projection frame can simultaneously correspond to both the "expiratory phase" and the "early ventricular systole," and these two phase identifiers together constitute the composite phase label for that frame.

[0087] Composite phase datasets can be organized according to projection records, with each record associated with at least one CBCT projection, the acquisition time of that projection, and its composite phase label. Thus, projections with the same or adjacent composite phase states can be identified and grouped, rather than simply mixing projections of different heartbeat states based on a single respiratory state.

[0088] S103, solve the global respiratory deformation field sequence based on the composite phase dataset, and use the global respiratory deformation field sequence to compensate the CBCT projection to solve the local heartbeat deformation field sequence, forming a dual physiological deformation field group.

[0089] The global respiratory deformation field sequence describes the overall displacement changes of thoracic and abdominal tissues between different respiratory phases. This displacement change typically has a large range, covering the lungs, diaphragm, chest wall, and cardiopulmonary structures associated with respiratory movements. The global respiratory deformation field sequence can be obtained through registration or projection consistency relationships between data from different respiratory phases.

[0090] Compensating CBCT projection using a global respiratory deformation field sequence involves reducing or eliminating positional changes caused by respiratory motion in the projection according to the respiratory phase corresponding to the projection. This ensures that the compensated projection primarily retains the local periodic changes caused by the heartbeat. This processing sequence can reduce the possibility of large-scale respiratory displacements being misidentified as local cardiac displacements.

[0091] The local deformation field sequence of the heartbeat is used to describe the displacement changes of local structures such as the heart, great vessels, and pulmonary hilum between different heartbeat phases. During the solution process, the compensated projection can be further organized according to the heartbeat phase, and the cardiopulmonary related regions can be used as the primary solution region. The global deformation field sequence of respiration and the local deformation field sequence of the heartbeat together constitute a dual physiological deformation field set, but the two types of deformation fields can still be preserved separately at this stage.

[0092] S104, generate a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region.

[0093] The dual-coupled total deformation field sequence is a deformation field sequence organized according to the composite phase and capable of simultaneously reflecting the spatial differences between respiratory and cardiac motion. The dual-motion coupling region uses the global respiratory deformation field and the local cardiac deformation field corresponding to the current composite phase; the single respiratory motion region does not introduce the local cardiac deformation field.

[0094] As a concrete example of the principle of partitioning and combining, see [link to relevant documentation]. Figure 5 The anatomical region mask 610 includes a dual-motion coupling region 611 and a single respiratory motion region 612. The global respiratory deformation field sequence 620 provides corresponding respiratory deformation to the dual-motion coupling region 611 and the single respiratory motion region 612, while the local cardiac deformation field sequence 630 provides corresponding cardiac deformation only to the dual-motion coupling region 611. The combined result of these partitions forms the dual-coupled total deformation field sequence 640. Thus, the local cardiac deformation is limited to the area of ​​effect identified by the mask.

[0095] S105, based on the composite phase label and the dual-coupled total deformation field sequence, perform motion compensation reconstruction and temporal optimization on the CBCT projection according to the composite phase, and output a four-dimensional dynamic image sequence.

[0096] Motion compensation reconstruction based on composite phase refers to using CBCT projections with corresponding composite phase labels and a dual-coupled total deformation field corresponding to the same composite phase for volume data reconstruction. The dual-coupled total deformation field is used to correct the displacement of the target structure relative to the reference state during projection acquisition, enabling projections from different angles to participate in reconstruction under a unified motion state.

[0097] Temporal optimization is used to reduce differences in non-physiological grayscale fluctuations, boundary positions, and local noise between adjacent composite phase images. Four-dimensional dynamic image sequences, consisting of three-dimensional volume data arranged in temporal or composite phase order, can be used to continuously display dynamic changes in the myocardium, great vessels, hilum of the lung, and adjacent tissues.

[0098] As can be seen, this embodiment first compensates for the projection with global respiratory deformation and then solves for the local cardiac deformation, which can reduce the mutual interference between the two types of motion in the deformation solution stage; furthermore, by combining the two types of deformation according to the anatomical region, it can inhibit the diffusion of cardiac deformation to tissues only affected by respiration. Thus, even when the animal's breathing is irregular and the heart rate is fast, the stability of the cardiopulmonary structural boundary and continuous dynamic process can still be improved.

[0099] In one possible embodiment, see Figure 6 The process of generating the dual-coupled total deformation field sequence in step S104 can be achieved through the following steps S201 to S202.

[0100] S201, determine the partitioned deformation field of each composite phase based on the dual physiological deformation field group and the anatomical region mask.

[0101] The partitioned deformation field for each composite phase is the spatial deformation result formed for a single composite phase. For a composite phase, the CBCT reconstruction device reads the corresponding global respiratory deformation field and local cardiac deformation field, and determines the deformation type used for different voxel locations based on the anatomical region mask. This partitioned deformation field preserves the spatial characteristics of different anatomical regions affected by physiological motion to varying degrees.

[0102] For example, when a composite phase represents a combination of the expiratory phase and the ventricular systolic phase, the cardiac region can be associated with the global respiratory deformation corresponding to the expiratory phase and the local cardiac deformation corresponding to the ventricular systolic phase; while the subcutaneous soft tissue region is only associated with the global respiratory deformation corresponding to the expiratory phase.

[0103] S202, perform time-series continuous processing on the partitioned deformation fields of each composite phase to generate the dual-coupled total deformation field sequence.

[0104] The temporal continuity processing involves examining and adjusting the changes in the deformation fields between adjacent partitions along the temporal arrangement of the composite phase. The processing targets can include abrupt changes in deformation direction, discontinuous displacement changes, and local anomalous deformations. Temporal continuity processing can preserve normal periodic motion changes and reduce non-physiological jumps caused by differences in the number of projections, waveform noise, or local solution errors.

[0105] In one possible implementation, see Figure 7The partitioned deformation fields 710 of each composite phase are arranged in chronological order to form adjacent phase deformation fields 720. When a discontinuous change between adjacent phases is detected, a transition phase deformation field 730 can be formed based on the effective deformation state of the preceding and following phases, and the continuous deformation of each phase is organized into a dual-coupled total deformation field sequence 640. This embodiment can solve the subdivision problem that the deformation result is prone to jumps when the projection of some composite phases is insufficient, and can improve the smoothness of continuous frame motion trajectories.

[0106] As can be seen, this embodiment sets spatial partitioning and temporal continuity as two consecutive processing stages. Spatial partitioning ensures that different anatomical regions obtain deformation fields that conform to their motion characteristics, while temporal continuity ensures that there is a reasonable dynamic transition between each composite phase, thus taking into account both local motion differences and overall temporal stability.

[0107] In one possible embodiment, see Figure 8 The process of determining the partitioned deformation field in step S201 can be achieved through the following steps S301 to S302.

[0108] S301, determine the regional deformation configuration of each composite phase based on the dual physiological deformation field group and the anatomical region mask.

[0109] The regional deformation configuration is the configuration data that controls how deformation fields are selected and combined in different anatomical regions within the current composite phase. The regional deformation configuration records the region category, the current composite phase, the available deformation types, and the corresponding deformation processing methods. This configuration is not the deformation result itself, but rather a control basis used to guide subsequent partition fusion.

[0110] For example, for voxels marked as dual-motion coupled regions, the region deformation configuration indicator calls both the global respiratory deformation field and the local heartbeat deformation field corresponding to the current composite phase; for voxels marked as single respiratory motion regions, the region deformation configuration indicator calls only the global respiratory deformation field.

[0111] S302, the dual physiological deformation field group is partitioned and fused according to the regional deformation configuration of each composite phase to determine the partitioned deformation field of each composite phase.

[0112] In this process, partition fusion combines or retains deformation data from dual physiological deformation field groups within the corresponding spatial region according to the regional deformation configuration. The fusion process maintains the correspondence between each deformation field and the current composite phase, and does not directly write deformation data from other composite phases into the current partition deformation field.

[0113] In one possible implementation, before performing partition fusion, the existence of the deformation field referenced by the current region deformation configuration can be checked, and the consistency of their composite phase identifiers can be verified. When a deformation field is unavailable, the corresponding region can be marked as a region to be continuumized and handed over to subsequent temporal continuumization processing, avoiding the direct invocation of deformation fields with mismatched phases. This implementation can solve the subdivision problem of cross-phase misfusion that easily occurs when deformation fields are missing, and is beneficial to maintaining the phase consistency of partition deformation results.

[0114] As can be seen, this embodiment uses regional deformation configuration as the control medium between the dual physiological deformation field group and the partition fusion process, so that the regions, deformation types and composite phases involved in the fusion have clear correspondences, thereby improving the controllability and verifiability of the partition fusion process.

[0115] In one possible embodiment, see Figure 9 The process of determining the regional deformation configuration in step S301 can be achieved through the following steps S401 to S402.

[0116] S401, Generate regional deformation allocation information for each composite phase based on the dual physiological deformation field group and the anatomical region mask.

[0117] The regional deformation allocation information describes the deformation fields that can be associated with each anatomical region under the current composite phase. This regional deformation allocation information can be generated separately according to the composite phase, while maintaining the association between the anatomical region identifier and the corresponding deformation fields in the dual physiological deformation field group.

[0118] Regional deformation allocation information focuses on expressing "which deformation fields are allocated to which regions". For example, the heart region corresponds to the global deformation field of respiration and the local deformation field of heartbeat, while the non-cardiopulmonary regions of the thoracic cage correspond to the global deformation field of respiration.

[0119] S402, determine the regional deformation configuration of each composite phase based on the regional deformation allocation information of each composite phase.

[0120] Among them, the regional deformation configuration is an executable configuration formed based on the regional deformation allocation information. The CBCT reconstruction equipment can perform an integrity check on the regional deformation allocation information to confirm whether the required deformation field for the current region exists, and convert the allocation relationship that passes the check into a configuration record that can be called in the partition fusion stage.

[0121] In one possible implementation, a region integrity check and a phase consistency check can be performed on the region deformation allocation information. The region integrity check confirms that all valid regions in the anatomical region mask have corresponding allocation records; the phase consistency check confirms that the global respiratory deformation field and the local cardiac deformation field referenced by the allocation record belong to the current composite phase. Records that pass the checks are written into the region deformation configuration, while records that fail are marked as records to be corrected. This implementation can solve the problem of difficulty in timely identification when allocation information is missing or phase mismatched, thus improving the reliability of the configuration data.

[0122] As can be seen, this embodiment distinguishes between regional deformation allocation information and regional deformation configuration, so that the original allocation relationship between the anatomical region and the candidate deformation field can be recorded and checked first, and then converted into configuration data used for partition fusion, which is beneficial for tracking the allocation basis of the deformation field.

[0123] In one possible embodiment, see Figure 10 The process of generating regional deformation allocation information in step S401 can be implemented through the following steps S501 to S502.

[0124] S501, based on the dual physiological deformation field group and the anatomical region mask, associate each mask voxel with the global respiratory deformation field and the local cardiac deformation field corresponding to each composite phase to establish voxel deformation association data.

[0125] Here, the mask voxel is a voxel located within the effective range of the anatomical region mask. Voxel deformation correlation data is used to record the correspondence between the mask voxel and the composite phase deformation field. Each correlation record can correspond to a mask voxel and a composite phase, and is associated with the available global respiratory deformation field and local cardiac deformation field under that composite phase.

[0126] When establishing associations, deformation data at corresponding locations in the global respiratory deformation field can be obtained first based on the spatial location of the mask voxels. Then, it can be determined whether the local heartbeat deformation field has valid deformation data at that location. The association process does not directly determine which deformation is ultimately used, but rather forms the basis for candidate associations required for subsequent region screening.

[0127] In one possible implementation, a consistency check can be performed on the association records of adjacent mask voxels. When an isolated deformation field association is missing within the same region, it can be marked as an association to be completed based on the region category of the voxel and the effective association status of adjacent voxels under the same composite phase. This implementation can solve the subdivision problem of discrete association records caused by local segmentation edges or deformation solution gaps, and can provide a stable data foundation for the subsequent formation of continuous regional deformation allocation information.

[0128] S502, determine the region deformation allocation information for each composite phase based on the voxel deformation association data and the anatomical region mask.

[0129] The CBCT reconstruction equipment reads the region category corresponding to each associated record based on the anatomical region mask and filters the assignable deformation field according to the region category. For voxels belonging to dual-motion coupled regions, the associated global respiratory deformation field and local cardiac deformation field can be retained; for voxels belonging to a single respiratory motion region, the global respiratory deformation field can be retained and the local cardiac deformation field can be excluded.

[0130] In one possible implementation, voxel-associated records with the same composite phase, region category, and deformation allocation method can be merged into region-level allocation records. This merging process does not change the original correspondence between voxels and deformation fields, but it reduces the number of duplicate records that need to be processed in subsequent region deformation configurations. This implementation can solve the problem of decreased configuration generation efficiency when the amount of voxel-level associated data is large, and it is beneficial to maintain the consistency of allocation rules within the same region.

[0131] As can be seen, this embodiment establishes the association between anatomical regions and dual physiological deformation fields based on voxels, and then generates allocation information based on the anatomical region category, so that the deformation field allocation can be accurately implemented to the spatial location and can be uniformly managed according to the regional rules.

[0132] In one possible embodiment, see Figure 11 The process of determining the regional deformation distribution information in step S502 can be achieved through the following steps S601 to S602.

[0133] S601, based on the voxel deformation association data and the anatomical region mask, extract the region category of each mask voxel and its associated global respiratory deformation field and local heartbeat deformation field to form region deformation candidate information.

[0134] Among them, the regional deformation candidate information is data formed by reorganizing voxel deformation association data according to anatomical region categories. This candidate information can at least indicate whether the mask voxel belongs to a dual-motion coupled region or a single respiratory motion region, and indicate what kind of global respiratory deformation field and local cardiac deformation field the voxel is associated with in the current composite phase.

[0135] Not all deformation fields in the candidate information have been determined as the final assignment results. For example, a voxel in a single respiratory motion region can be associated with nearby local heartbeat deformation data in the voxel deformation association data, but this local heartbeat deformation data will be excluded in subsequent region filtering.

[0136] S602, based on the regional deformation candidate information, retain the corresponding global respiratory deformation field and local heartbeat deformation field in the dual motion coupling region, retain the corresponding global respiratory deformation field in the single respiratory motion region, and determine the regional deformation allocation information for each composite phase;

[0137] The dual-motion coupling region includes at least one of the heart, great vessels, and pulmonary hilum, and the single respiratory motion region includes at least one of the non-cardiopulmonary region of the thoracic cavity and subcutaneous soft tissue.

[0138] The heart, great vessels, and pulmonary hilum move with the entire chest and abdomen during respiration, and are also affected by myocardial contraction, vascular pulsation, or the movement of adjacent cardiovascular structures. Therefore, they can be identified as areas of dual-motor coupling. The non-cardiopulmonary regions of the thoracic cavity and subcutaneous soft tissues mainly change with respiration and overall body position, and usually do not require the introduction of local deformation due to heartbeat. Therefore, they can be identified as areas of single respiratory motion.

[0139] For example, in cat chest scans, voxels within the heart outline preserve both global respiratory deformation and local cardiac deformation; voxels in the subcutaneous soft tissue of the lateral chest wall preserve only global respiratory deformation. Voxels located at region boundaries are assigned according to the category in the anatomical region mask, which can reduce non-physiological aberrations between adjacent phases in subsequent temporal continuum processing.

[0140] As can be seen, this embodiment selects candidate deformation fields based on clearly defined anatomical region categories, so that the local deformation of the heartbeat mainly acts on the areas where there is indeed dual movement of breathing and heartbeat, reducing the interference of heartbeat correction on non-cardiopulmonary regions of the thoracic cavity and subcutaneous soft tissues.

[0141] In one possible embodiment, see Figure 12 The composite phase dataset generation process in step S102 can be implemented through the following steps S701 to S703.

[0142] S701, perform frequency band filtering based on the respiratory waveform and the electrocardiogram waveform to obtain the low-frequency respiratory component and the high-frequency heartbeat component.

[0143] Bandwidth filtering is used to reduce baseline drift, transient noise, and the interaction between different physiological signal frequency components. Low-frequency respiratory components are preserved to characterize the waveform trends of inspiratory and expiratory cycles, while high-frequency cardiac components are preserved to identify waveform changes at key points in the cardiac cycle. The specific filtering range can be configured based on animal species, body size, scanning status, and the actual sampling frequency.

[0144] S702, identify the peaks and troughs of the low-frequency respiratory component to divide the respiratory phase, and identify the QRS feature points of the high-frequency heartbeat component to divide the heartbeat phase.

[0145] The peaks and troughs of the low-frequency respiratory component are used to define the positions of inspiration, expiration, and the transition between them. QRS characteristic points are feature points in the electrocardiogram waveform used to locate the timing of ventricular depolarization and can serve as a time reference for dividing continuous heartbeat cycles. Phase division is based on actually detected physiological characteristic points, rather than directly assuming that the animal has a fixed respiratory cycle or a fixed heartbeat cycle.

[0146] S703, based on the temporal correspondence in the temporal alignment dataset, the corresponding respiratory phase and heartbeat phase are combined into the composite phase label and configured for the CBCT projection to generate the composite phase dataset.

[0147] The CBCT reconstruction equipment queries the respiratory and cardiac phases corresponding to each frame of CBCT projection based on the acquisition time, and then combines the two phase identifiers into a composite phase label. The composite phase label can be attached to the record of the corresponding projection, or it can be indexed with the projection identifier.

[0148] For example, multiple projections of different heartbeat phases may be acquired within the same respiratory phase. By using composite phase labels, these projections will not all be classified into the same motion state simply because they share the same respiratory phase, thus providing a temporal basis for subsequently solving the global respiratory deformation and the local heartbeat deformation separately.

[0149] As can be seen, this embodiment first separates the respiratory component and the heartbeat component suitable for phase recognition, and then combines the two phases according to a unified time relationship, which can simultaneously preserve the low-frequency respiratory motion state and the high-frequency heartbeat motion state, and reduce the physiological motion confusion caused by a single phase label.

[0150] In one possible embodiment, see Figure 13 The process of dividing the respiratory phase and heartbeat phase in step S702 can be achieved through the following steps S801 to S802.

[0151] S801, based on the low-frequency respiratory component and the high-frequency heartbeat component, detect adjacent peaks and troughs and adjacent QRS feature points respectively to form physiological cycle boundary information.

[0152] The physiological cycle boundary information includes respiratory cycle boundaries and cardiac cycle boundaries. Respiratory cycle boundaries can be determined by adjacent peaks, adjacent troughs, or combinations of peaks and troughs; cardiac cycle boundaries can be determined by adjacent QRS feature points. Each cycle boundary corresponds to the time reference used in the time-aligned dataset.

[0153] When respiratory amplitude changes, the local trend of the waveform can be used to determine whether candidate peaks or troughs constitute valid boundaries. When short-term noise exists in the ECG waveform, candidate points with obvious discontinuities can be eliminated based on the temporal continuity of adjacent QRS feature points, thereby reducing the impact of abnormal feature points on cycle division.

[0154] S802, based on the physiological cycle boundary information, each respiratory cycle and each heartbeat cycle are continuously segmented to obtain the respiratory phase and the heartbeat phase.

[0155] Continuous segmentation refers to the formation of continuous phase segments within each defined physiological cycle, according to the start and end boundaries of that cycle. The duration of different respiratory cycles may vary, and the duration of different cardiac cycles may also vary, but the phase segments within each cycle maintain a predetermined order.

[0156] For example, when an animal's previous respiratory cycle is longer than its subsequent respiratory cycle, the two cycles are segmented according to their respective boundaries. As a result, the subsequent cycle will not continue to use the absolute time length of the previous cycle, which can reduce the phase shift caused by changes in respiratory rate.

[0157] As can be seen, this embodiment uses the adjacent physiological feature points detected in actual testing to define the period boundary and independently performs continuous segmentation within each period, which can adapt to changes in animal respiratory rhythm and heart rate, and improve the accuracy of the temporal correspondence between respiratory phase and heartbeat phase.

[0158] In one possible embodiment, see Figure 14 The motion compensation reconstruction and temporal optimization process in step S105 can be implemented through the following steps S901 to S903.

[0159] S901, based on the composite phase label and the dual-coupled total deformation field sequence, the CBCT projection is divided into composite phase projection groups, and a corresponding dual-coupled total deformation field is associated with each composite phase projection group to form a phase reconstruction task set.

[0160] The composite phase projection group consists of CBCT projections with the same composite phase label or satisfying a predetermined adjacent phase relationship. Each composite phase projection group is associated with the dual-coupled total deformation field corresponding to its composite phase, thus forming a phase reconstruction task. Multiple phase reconstruction tasks constitute a phase reconstruction task set according to the order of the composite phases.

[0161] When the number of projections is sufficient, projection groups can be formed using the same composite phase label; when the number of projections for individual composite phases is insufficient, projections of adjacent composite phases can be introduced while maintaining temporal continuity, and they can be compensated to the target composite phase according to the corresponding deformation field.

[0162] S902, based on the phase reconstruction task set, motion compensation iterative reconstruction is performed under projection data fidelity constraints and deformation smoothing constraints to obtain an initial four-dimensional dynamic image sequence.

[0163] Among them, the projection data fidelity constraint is used to ensure that the projection results generated by the reconstructed volume data are consistent with the actual acquired CBCT projection. The deformation smoothing constraint is used to limit the deformation field from abrupt changes that lack physiological basis in the spatial neighborhood and between adjacent composite phases. The motion-compensated iterative reconstruction unifies the motion state of different projections according to the dual-coupled total deformation field during the iteration process and updates the volume data of each composite phase.

[0164] The initial four-dimensional dynamic image sequence consists of multiple three-dimensional image frames arranged in a composite phase. The sequence has been compensated for respiratory and cardiac movements, but grayscale fluctuations and local noise may still exist between different composite phase images due to differences in the number of projections or imaging conditions.

[0165] S903, perform cross-phase grayscale alignment, temporal continuity constraint, edge-preserving filtering and grayscale normalization on the initial four-dimensional dynamic image sequence, and output the four-dimensional dynamic image sequence.

[0166] Among them, cross-phase gray-level alignment is used to reduce non-physiological gray-level deviations of similar tissues in different composite phases; temporal continuity constraint is used to limit abnormal abrupt changes between adjacent image frames; edge-preserving filtering is used to preserve the structural boundaries of myocardium, large blood vessels and hilum of the lung while suppressing local noise; gray-level normalization is used to unify the display range of each image frame.

[0167] The above processes can be performed sequentially or in combination while maintaining their individual functions. The output four-dimensional dynamic image sequence is arranged in chronological or composite phase order and can be used for frame-by-frame display, loop playback, and three-dimensional observation of selected composite phases.

[0168] As can be seen, this embodiment ensures that each composite phase projection group corresponds to the corresponding dual-coupled total deformation field, while simultaneously considering the consistency of projection data and the continuity of deformation during the reconstruction phase. Through temporal optimization after reconstruction, grayscale fluctuations and boundary jitter between different phases can be further reduced, improving the continuous display effect of dynamic images.

[0169] In another embodiment, see Figure 15 The animal respiratory and cardiac dual-coupled CBCT dynamic reconstruction system is deployed in the CBCT reconstruction device 300, including a data acquisition module 810, a phase labeling module 820, a deformation solving module 830, a partition fusion module 840, and a reconstruction output module 850. Each module is connected sequentially through the data transmission path in the CBCT reconstruction device 300.

[0170] The data acquisition module 810 is used to acquire the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device. The dataset includes the corrected CBCT projection, respiratory waveform, and electrocardiogram waveform.

[0171] The data acquisition module 810 can receive data output from the multi-source synchronous acquisition device via a communication interface, and obtain the anatomical region mask from local storage space, network storage space, or segmentation processing program. The data acquisition module 810 can also check the integrity of the projection markers, acquisition time, and waveform time reference, providing a unified data input for subsequent phase marking.

[0172] The phase labeling module 820 is used to determine the phase based on the respiratory waveform and the electrocardiogram waveform, configure composite phase labels for the CBCT projection, and generate a composite phase dataset.

[0173] The phase marking module 820 may include waveform filtering, feature point detection, period division, and label configuration functions. The phase marking module 820 queries the corresponding respiratory and cardiac phases based on the projection acquisition time, and then combines the two phases and configures them for the corresponding CBCT projection.

[0174] The deformation solving module 830 is used to solve the global deformation field sequence of respiration based on the composite phase dataset, and to solve the local deformation field sequence of heartbeat after compensating the CBCT projection with the global deformation field sequence of respiration, forming a dual physiological deformation field group.

[0175] Specifically, the deformation solving module 830 first obtains the global respiratory deformation field sequence based on the overall motion relationship between respiratory phases, then uses this sequence to compensate for respiratory motion in the CBCT projection, and solves the local heartbeat deformation field sequence from the compensated projection. The two types of deformation fields are stored separately and a correspondence is established through composite phase.

[0176] The partition fusion module 840 is used to generate a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region.

[0177] The partition fusion module 840 can perform voxel deformation association, region category extraction, region deformation allocation, region deformation configuration, partition fusion, and temporal continuity processing. The partition fusion module 840 limits the range of the local deformation field of the heartbeat according to the anatomical region and organizes the partition deformation results of each composite phase into a dual-coupled total deformation field sequence.

[0178] The reconstruction output module 850 is used to perform motion compensation reconstruction and temporal optimization on the CBCT projection according to the composite phase based on the composite phase label and the dual-coupled total deformation field sequence, and output a four-dimensional dynamic image sequence.

[0179] The reconstruction output module 850 can establish a phase reconstruction task set, perform motion-compensated iterative reconstruction, and perform cross-phase grayscale alignment, temporal continuity constraints, edge-preserving filtering, and grayscale normalization on the initial four-dimensional dynamic image sequence. The reconstruction results can be output to a display device, an image storage device, or a subsequent image analysis program.

[0180] As can be seen, this embodiment achieves a complete system architecture from multi-source synchronous data access to four-dimensional dynamic image output through the collaborative processing of the data acquisition module 810, phase marking module 820, deformation solving module 830, partition fusion module 840, and reconstruction output module 850. This system can separately process global respiratory motion and local cardiac motion, and limits the combination range of the two types of deformation based on the anatomical region, thereby improving the structural stability and temporal continuity of animal cardiopulmonary dynamic CBCT images.

[0181] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0182] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0183] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0184] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0185] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0186] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0187] If the aforementioned integrated units are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0188] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0189] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for dynamic reconstruction of animal respiratory and cardiac rhythm using dual-coupled CBCT, characterized in that, A CBCT reconstruction device used in communication with a multi-source synchronous acquisition device, the method comprising: Obtain the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device. The time-aligned dataset includes the corrected CBCT projection, respiratory waveform, and electrocardiogram waveform. The phase is determined based on the respiratory waveform and the electrocardiogram waveform, and a composite phase label is configured for the CBCT projection to generate a composite phase dataset. The global respiratory deformation field sequence is solved based on the composite phase dataset. The global respiratory deformation field sequence is then used to compensate the CBCT projection and the local heartbeat deformation field sequence is solved to form a dual physiological deformation field set. Generating a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask includes: determining the partitioned deformation field of each composite phase based on the dual physiological deformation field group and the anatomical region mask; performing temporal continuity processing on the partitioned deformation fields of each composite phase to generate the dual-coupled total deformation field sequence; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region; Based on the composite phase label and the dual-coupled total deformation field sequence, motion compensation reconstruction and temporal optimization are performed on the CBCT projection according to the composite phase to output a four-dimensional dynamic image sequence.

2. The method according to claim 1, characterized in that, The step of determining the partitioned deformation field of each composite phase based on the dual physiological deformation field group and the anatomical region mask includes: The regional deformation configuration of each composite phase is determined based on the dual physiological deformation field group and the anatomical region mask; The dual physiological deformation field group is partitioned and fused according to the regional deformation configuration of each composite phase to determine the partitioned deformation field of each composite phase.

3. The method according to claim 2, characterized in that, The determination of the regional deformation configuration of each composite phase based on the dual physiological deformation field group and the anatomical region mask includes: Based on the dual physiological deformation field group and the anatomical region mask, regional deformation allocation information for each composite phase is generated; The regional deformation configuration of each composite phase is determined based on the regional deformation allocation information of each composite phase.

4. The method according to claim 3, characterized in that, The generation of regional deformation allocation information for each composite phase based on the dual physiological deformation field group and the anatomical region mask includes: Based on the dual physiological deformation field group and the anatomical region mask, each mask voxel is associated with the global respiratory deformation field and the local cardiac deformation field corresponding to each composite phase to establish voxel deformation association data; The region deformation allocation information for each composite phase is determined based on the voxel deformation correlation data and the anatomical region mask.

5. The method according to claim 4, characterized in that, The step of determining the region deformation allocation information for each composite phase based on the voxel deformation correlation data and the anatomical region mask includes: Based on the voxel deformation association data and the anatomical region mask, the region category of each mask voxel and its associated global respiratory deformation field and local heartbeat deformation field are extracted to form region deformation candidate information; Based on the regional deformation candidate information, the corresponding global respiratory deformation field and local heartbeat deformation field are retained in the dual motion coupling region, and the corresponding global respiratory deformation field is retained in the single respiratory motion region, thereby determining the regional deformation allocation information for each composite phase. The dual-motion coupling region includes at least one of the heart, great vessels, and pulmonary hilum, and the single respiratory motion region includes at least one of the non-cardiopulmonary region of the thoracic cavity and subcutaneous soft tissue.

6. The method according to any one of claims 1 to 5, characterized in that, The step of determining the phase based on the respiratory waveform and the electrocardiogram waveform, configuring composite phase labels for the CBCT projection, and generating a composite phase dataset includes: Based on the respiratory waveform and the electrocardiogram waveform, frequency band filtering is performed to obtain the low-frequency respiratory component and the high-frequency heartbeat component; The peaks and troughs of the low-frequency respiratory component are identified to divide the respiratory phase, and the QRS feature points of the high-frequency heartbeat component are identified to divide the heartbeat phase. Based on the temporal correspondence in the temporal alignment dataset, the corresponding respiratory phases and heartbeat phases are combined into composite phase labels and configured for the CBCT projection to generate the composite phase dataset.

7. The method according to claim 6, characterized in that, The step of identifying the peaks and troughs of the low-frequency respiratory component to divide the respiratory phase, and identifying the QRS feature points of the high-frequency heartbeat component to divide the heartbeat phase, includes: Based on the low-frequency respiratory component and the high-frequency heartbeat component, adjacent peaks and troughs, as well as adjacent QRS feature points, are detected to form physiological cycle boundary information. Based on the physiological cycle boundary information, each respiratory cycle and each heartbeat cycle are continuously segmented to obtain the respiratory phase and the heartbeat phase.

8. The method according to any one of claims 1 to 5, characterized in that, The step of performing motion-compensated reconstruction and temporal optimization of the CBCT projection according to the composite phase label and the dual-coupled total deformation field sequence, and outputting a four-dimensional dynamic image sequence, includes: Based on the composite phase label and the dual-coupled total deformation field sequence, the CBCT projection is divided into composite phase projection groups, and each composite phase projection group is associated with a corresponding dual-coupled total deformation field to form a phase reconstruction task set; Based on the phase reconstruction task set, motion compensation iterative reconstruction is performed under projection data fidelity constraints and deformation smoothing constraints to obtain an initial four-dimensional dynamic image sequence. The initial four-dimensional dynamic image sequence is subjected to cross-phase grayscale alignment, temporal continuity constraint, edge-preserving filtering and grayscale normalization to output the four-dimensional dynamic image sequence.

9. A dual-coupled CBCT dynamic reconstruction system for animal respiration and heartbeat, used to implement the dual-coupled CBCT dynamic reconstruction method for animal respiration and heartbeat as described in any one of claims 1 to 8, characterized in that, CBCT reconstruction equipment deployed in communication with multi-source synchronous acquisition devices includes: The data acquisition module is used to acquire the anatomical region mask and the time-aligned dataset output by the multi-source synchronous acquisition device. The dataset includes the corrected CBCT projection, respiratory waveform, and electrocardiogram waveform. A phase labeling module is used to determine the phase based on the respiratory waveform and the electrocardiogram waveform, configure composite phase labels for the CBCT projection, and generate a composite phase dataset. The deformation solving module is used to solve the global respiratory deformation field sequence based on the composite phase dataset, and to solve the local heartbeat deformation field sequence after compensating the CBCT projection with the global respiratory deformation field sequence to form a dual physiological deformation field group. The partition fusion module is used to generate a dual-coupled total deformation field sequence based on the dual physiological deformation field group and the anatomical region mask; wherein, the corresponding global respiratory deformation field and local cardiac deformation field are fused in the dual motion coupling region identified by the anatomical region mask, and the global respiratory deformation field is retained in a single respiratory motion region. The reconstruction output module is used to perform motion compensation reconstruction and temporal optimization on the CBCT projection according to the composite phase label and the dual-coupled total deformation field sequence, and output a four-dimensional dynamic image sequence.

Images