Medical image reconstruction method, reconstruction device and scanning equipment

Abstract

The invention relates to a medical image reconstruction method, a reconstruction device and scanning equipment. The method comprises the following steps: acquiring scanning data of a scanning device for scanning a target object; determining a corresponding image reconstruction strategy according to the type of the reconstruction task; performing image reconstruction according to an image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object; the type of the reconstruction task comprises a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task comprises a plurality of subtasks; the image reconstruction strategy comprises the step of starting a multi-process parallel execution mechanism corresponding to the sub-tasks to perform image reconstruction. By adopting the method, the reconstruction efficiency of the medical image can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of medical image processing, and in particular to a medical image reconstruction method, reconstruction device and scanning equipment. Background Art

[0002] With the continuous development of medical imaging technology and the increasing requirements for image quality, especially with the development of dual-source CT, photon counting CT, and small-pixel CT, huge amounts of data are required in the image pre-processing process, and image resolution needs to be improved in the image post-processing process. Therefore, the current CT reconstruction system is facing tremendous performance pressure.

[0003] Traditional medical image reconstruction techniques generally include perfusion reconstruction, analytical reconstruction (AR), iterative reconstruction (IR), algebraic reconstruction technique (ART), and filtered backprojection (FBP) reconstruction. Perfusion reconstruction and artificial intelligence iterative reconstruction (AIIR) are medical image reconstruction methods. The core concept of perfusion reconstruction is to continuously scan the region of interest during a rapid intravenous bolus injection of contrast agent and reconstruct the scan data using a specific algorithm. AIIR, on the other hand, is an image reconstruction technology used in medical imaging (such as CT) that optimizes the image reconstruction process through deep learning algorithms, generating high-quality images with low radiation doses while reducing image noise and artifacts.

[0004] However, the reconstruction methods based on the above medical images have the problem of low reconstruction efficiency. Summary of the Invention

[0005] Based on this, it is necessary to provide a medical image reconstruction method, a reconstruction device and a scanning device that can improve the reconstruction efficiency of medical images in order to address the above technical problems.

[0006] In a first aspect, the present application provides a medical image reconstruction method, comprising:

[0007] Obtaining scan data of a target object scanned by a scanning device;

[0008] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0009] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0010] In one embodiment, the type of the reconstruction task is a perfusion reconstruction type, and the image reconstruction is performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object, including:

[0011] Splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, dividing the reconstruction task into a plurality of first sub-tasks, and associating the plurality of first sub-tasks with the plurality of first sub-scan data;

[0012] Allocating first idle threads to a plurality of first subtasks; each first subtask includes a plurality of first subnode tasks connected in series;

[0013] The idle threads corresponding to the first subtasks are called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0014] In one embodiment, the method further includes:

[0015] According to the processing task type corresponding to the irrigation reconstruction type, the scan data is segmented to obtain the intermediate segment data corresponding to each processing task type;

[0016] The scan data is split according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data, including:

[0017] According to the number of scan circles corresponding to the scan data and the idle resources on the scanning device, each intermediate segment of data is split to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0018] In one embodiment, calling the idle threads corresponding to the first subtasks to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object includes:

[0019] Allocate corresponding hardware resources to each idle thread according to the priority of the idle thread corresponding to the first sub-scan data of each segment; the hardware resources include CPU resources and / or GPU resources;

[0020] The hardware resources corresponding to each idle thread are called to perform image reconstruction operations to obtain a reconstructed image of the target object.

[0021] In one embodiment, the reconstruction task is of an artificial intelligence reconstruction type, and the reconstruction task includes a longitudinal task and a transverse task. The image reconstruction is performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object, including:

[0022] Splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data;

[0023] Splitting the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, dividing the longitudinal task into multiple second sub-node tasks connected in series, and associating the multiple second sub-node tasks with the multiple third sub-scan data;

[0024] Allocating a second idle thread to a plurality of second subtasks; each second subtask includes a plurality of second sub-node tasks connected in series;

[0025] The second idle threads corresponding to the second subtasks are called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0026] In one embodiment, the calling of the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object includes:

[0027] Determining a target second subtask according to the execution order of the plurality of second subtasks;

[0028] Invoking the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing;

[0029] Call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the associated scanning data, to obtain a reconstructed image of the target object.

[0030] In one embodiment, the calling of the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object includes:

[0031] During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

[0032] In a second aspect, the present application further provides a medical image reconstruction method, characterized in that it is applied to a computed tomography (CT) device, comprising:

[0033] Obtaining scan data of a target object scanned by a CT device;

[0034] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0035] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0036] In a third aspect, the present application further provides a medical image reconstruction device, comprising:

[0037] An acquisition module is used to acquire scanning data of a target object scanned by a scanning device;

[0038] a determination module for determining a corresponding image reconstruction strategy based on a type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0039] The reconstruction module is used to perform image reconstruction according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0040] In a fourth aspect, the present application further provides a scanning device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0041] Obtaining scan data of a target object scanned by a scanning device;

[0042] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0043] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0044] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0045] Obtaining scan data of a target object scanned by a scanning device;

[0046] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0047] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0048] In a sixth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0049] Obtaining scan data of a target object scanned by a scanning device;

[0050] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple threads corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0051] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0052] The above-mentioned medical image reconstruction method, reconstruction device and scanning equipment realize the reconstruction task of the medical image by starting multiple processes corresponding to multiple subtasks in the reconstruction task and executing them in parallel. Compared with the existing method of realizing the reconstruction of medical images by executing serial processes, the present application greatly improves the reconstruction efficiency of medical images; in addition, it takes into account the types of different reconstruction tasks, determines the corresponding image reconstruction strategies, and realizes the reconstruction of medical images based on the image reconstruction strategies corresponding to the reconstruction tasks. Compared with the existing method of realizing the reconstruction of medical images by only providing one reconstruction strategy, the present application greatly improves the reconstruction flexibility of medical images. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 FIG. 1 is a diagram of an application environment of a medical image reconstruction method according to an embodiment;

[0055] Figure 2 is a schematic flow chart of a medical image reconstruction method in one embodiment;

[0056] Figure 3 is a flowchart of a medical image reconstruction method according to another embodiment;

[0057] Figure 4 is a flowchart of a medical image reconstruction method according to another embodiment;

[0058] Figure 5 A schematic diagram of data partitioning using a single thread in one embodiment;

[0059] Figure 6 A schematic diagram of data partitioning using multiple threads in one embodiment;

[0060] Figure 7 is a flowchart of a medical image reconstruction method according to another embodiment;

[0061] Figure 8 is a flowchart of a medical image reconstruction method according to another embodiment;

[0062] Figure 9 A schematic diagram of resource scheduling using a single thread in one embodiment;

[0063] Figure 10 A schematic diagram of resource scheduling using multiple threads in one embodiment;

[0064] Figure 11 is a flowchart of a medical image reconstruction method according to another embodiment;

[0065] Figure 12 A schematic diagram of a process for retrieving intermediate images through multiple threads in one embodiment;

[0066] Figure 13 is a flowchart of a medical image reconstruction method according to another embodiment;

[0067] Figure 14is a structural block diagram of a medical image reconstruction device in one embodiment;

[0068] Figure 15 is a structural block diagram of a medical image reconstruction device in another embodiment;

[0069] Figure 16 FIG. 4 is a diagram showing the internal structure of a scanning device in one embodiment. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0071] With the continuous development of medical imaging technology and the increasing requirements for image quality, especially with the development of dual-source CT, photon counting CT, and small-pixel CT, huge amounts of data are required in the image pre-processing process, and image resolution needs to be improved in the image post-processing process. Therefore, the current CT reconstruction system is facing tremendous performance pressure.

[0072] Traditional medical image reconstruction techniques generally include perfusion reconstruction, analytical reconstruction (AR), iterative reconstruction (IR), algebraic reconstruction technique (ART), and filtered backprojection (FBP) reconstruction. Perfusion reconstruction and artificial intelligence iterative reconstruction (AIIR) are medical image reconstruction methods. The core concept of perfusion reconstruction is to continuously scan the region of interest during a rapid intravenous bolus injection of contrast agent and reconstruct the scan data using a specific algorithm. AIIR, on the other hand, is an image reconstruction technique for medical images (such as CT) that uses deep learning algorithms to optimize the image reconstruction process, producing high-quality images with low radiation doses while reducing image noise and artifacts. However, these medical image reconstruction methods suffer from low reconstruction efficiency. This application aims to address this issue.

[0073] After introducing the background technology of the medical image reconstruction method provided by the embodiment of the present application, the following briefly describes the implementation environment involved in the medical image reconstruction method provided by the embodiment of the present application. The medical image reconstruction method provided by the embodiment of the present application can be applied to Figure 1The scanning device shown in FIG. The scanning device includes a data acquisition module 11 and a processor 12. The data acquisition module 11 communicates with the processor 12 via a network. The data storage system can store data that the processor 12 needs to process. The data storage system can be integrated with the processor 12 or placed on a cloud or other network server. The processor 12 can include multiple data processing processes. When reconstructing medical images, multiple data processing processes can be called to perform medical image reconstruction operations to achieve medical image reconstruction.

[0074] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the scanning device to which the solution of the present application is applied. The specific scanning device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0075] After introducing the application scenarios of the medical image reconstruction method provided by the embodiments of the present application, the following focuses on introducing the medical image reconstruction method described in the present application.

[0076] In one embodiment, Figure 2 As shown, a medical image reconstruction method is provided, which is applied to Figure 1 The following steps are used as an example to illustrate the scanning device in the figure:

[0077] S201: Obtain scanning data of a target object scanned by a scanning device.

[0078] The scanning device may be any type of imaging device, such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), or X-ray imaging (XR). The target object may be any part of the patient's body, such as the abdomen, chest, leg, or hand.

[0079] In this embodiment, when medical image reconstruction is required based on the scanning data obtained by the scanning device, the data acquisition module in the scanning device can first scan the target object to obtain the scanning data corresponding to the target object, and then the processor in the scanning device obtains the scanning data corresponding to the target object from the data acquisition module.

[0080] Optionally, the data acquisition module in the scanning device can scan the target object in advance, and after obtaining the scanning data corresponding to the target object, save the scanning data corresponding to the target object. Then, the processor in the scanning device can obtain the scanning data corresponding to the target object from the data acquisition module at any time when medical image reconstruction is required.

[0081] Optionally, after the data acquisition module in the scanning device scans the target object and obtains the scanning data corresponding to the target object, the processor in the scanning device may directly obtain the scanning data corresponding to the target object from the data acquisition module.

[0082] S202: Determine a corresponding image reconstruction strategy according to the type of reconstruction task.

[0083] Among them, the types of reconstruction tasks include perfusion reconstruction type or artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism of starting multiple processes corresponding to multiple subtasks to execute in parallel to perform image reconstruction.

[0084] In this embodiment, a user (e.g., a medical professional scanning a target object) can select a reconstruction task type within the scanning device. After receiving the reconstruction task type, the scanning device determines an image reconstruction strategy corresponding to the reconstruction task type. It should be noted that there is a one-to-one correspondence between the reconstruction task type and the image reconstruction strategy.

[0085] Optionally, the image reconstruction strategy corresponding to the infusion reconstruction type may be to segment the scan data to obtain multiple scan data segments, configure corresponding subtasks for each scan data segment, and then start a mechanism for parallel execution of multiple processes corresponding to the multiple subtasks to perform image reconstruction; the image reconstruction strategy corresponding to the artificial intelligence reconstruction type may be to segment the scan task to obtain multiple scan task segments, and start multiple processes corresponding to the multiple scan task segments according to the scan data, and perform image reconstruction by a mechanism for parallel execution.

[0086] S203 : Perform image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0087] In this embodiment, after the image reconstruction strategy and the scan data are determined as described above, image reconstruction can be performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0088] For example, when the reconstruction task is of the perfusion reconstruction type, the target object needs to be scanned multiple times, and there is a problem of a large amount of scan data. The scan data can be segmented to obtain multiple scan data segments, and corresponding subtasks can be configured for each scan data segment. Then, a mechanism of parallel execution of multiple processes corresponding to multiple subtasks is started to perform image reconstruction to obtain a reconstructed image of the target object.

[0089] Another example is that when the type of reconstruction task is an artificial intelligence reconstruction type, multiple iterative reconstructions and deep learning are required to achieve the reconstruction of medical images, and each reconstruction includes forward projection and back projection. There is a problem of large task volume. The scanning task can be segmented to obtain multiple scanning task segments, and multiple processes corresponding to the multiple scanning task segments are started according to the scanning data. The image is reconstructed by a parallel execution mechanism to obtain a reconstructed image of the target object.

[0090] The medical image reconstruction method provided in the embodiment of the present application realizes the reconstruction task of the medical image by starting multiple processes corresponding to multiple subtasks in the reconstruction task for parallel execution. Compared with the existing method of realizing the reconstruction of medical images by executing serial processes, the present application greatly improves the reconstruction efficiency of medical images. In addition, the method takes into account the types of different reconstruction tasks, determines the corresponding image reconstruction strategies, and realizes the reconstruction of medical images based on the image reconstruction strategies corresponding to the reconstruction tasks. Compared with the existing method of realizing the reconstruction of medical images by only providing one reconstruction strategy, the present application greatly improves the reconstruction flexibility of medical images.

[0091] In one embodiment, Figure 2 Based on the embodiment shown, when the type of reconstruction task is the perfusion reconstruction type, as shown in FIG. Figure 3 As shown, the above S203 "performing image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object" includes:

[0092] S301. Split the scan data according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data, divide the reconstruction task into multiple first sub-tasks, and associate the multiple first sub-tasks with the multiple first sub-scan data.

[0093] The number of scan cycles refers to the number of times the data acquisition module repeatedly scans the target object in a perfusion reconstruction scenario. Once the target object is scanned and the scan data is obtained, the number of scan cycles is calculated. The number of scan cycles can be stored in the scan data or in the storage database of the scanning device. It should be noted that in a perfusion reconstruction scenario, the scan data obtained at each perfusion scan time point are not coupled and can be processed in parallel.

[0094] Among them, idle resources refer to resources that are not running in the scanning device, which may include idle threads, idle memory resources, idle graphics processing unit (GPU) resources and idle central processing unit (CPU) resources; threads refer to the execution path used for data processing in the scanning device, and the scanning device may include multiple threads, among which the threads that are performing data processing tasks are called non-idle threads, and the threads that are not performing data processing tasks are called idle threads.

[0095] In this embodiment, after obtaining the scan data of the target object scanned by the scanning device, the number of scan revolutions corresponding to the scan data and the idle resources on the scanning device can be obtained. The number of splits can be determined based on the number of scan revolutions corresponding to the scan data and the idle resources on the scanning device. The scan data can be split based on the number of splits to obtain the same number of first sub-scan data as the number of splits. The reconstruction task can be divided into the same number of first sub-tasks as the number of splits, and multiple first sub-tasks can be associated with multiple first sub-scan data. It should be noted that associating multiple first sub-tasks with multiple first sub-scan data specifically means that the first sub-scan data can be associated with the first sub-task, an associated label can be set, or a corresponding relationship can be generated; thereby, the first idle thread corresponding to the first sub-task can be called to perform image reconstruction operations in parallel based on the associated first sub-scan data.

[0096] Exemplarily, the scanning data can be split according to the number of scanning circles corresponding to the scanning data and the number of idle threads of the scanning device. For example, if the number of scanning circles corresponding to the scanning data is 10 and the number of idle threads on the scanning device is 10, then the determined number of splits is 10, and the scanning data can be split into 10 first sub-scanning data, and the reconstruction task can be divided into 10 first sub-tasks, and the 10 first sub-scanning data can be associated with the 10 first sub-tasks respectively, that is, a one-to-one correspondence is generated between the 10 first sub-scanning data and the corresponding 10 first sub-tasks, so that the first idle threads corresponding to each first sub-task can be called to perform image reconstruction operations in parallel based on the associated first sub-scanning data.

[0097] For another example, if the number of scanning circles corresponding to the scanning data is 10 and the number of idle threads on the scanning device is 5, then the determined number of splits is 5, and the scanning data can be split into 5 first sub-scanning data, and the reconstruction task can be divided into 5 first sub-tasks, and the 5 first sub-scanning data can be associated with the 5 first sub-tasks respectively, that is, a one-to-one correspondence is generated between the 5 first sub-scanning data and the corresponding 5 first subtasks, so that the first idle threads corresponding to each first subtask can be called to perform image reconstruction operations in parallel based on the associated first sub-scanning data.

[0098] For another example, if the number of scanning circles corresponding to the scanning data is 3 and the number of idle threads on the scanning device is 5, then the determined number of splits is 3, and the scanning data can be split into 3 first sub-scanning data, and the reconstruction task can be divided into 3 first sub-tasks, and the 3 first sub-scanning data can be associated with the 3 first sub-tasks respectively, that is, the 3 first sub-scanning data and the corresponding 3 first subtasks generate a one-to-one correspondence, so that the first idle thread corresponding to each first subtask can be called to perform image reconstruction operations in parallel based on the associated first sub-scanning data.

[0099] S302: Allocate first idle threads to multiple first subtasks.

[0100] Each first subtask includes a plurality of first sub-node tasks connected in series, and each first sub-node task is executed in sequence.

[0101] In this embodiment, after the multiple first subtasks are determined as described above, first idle threads can be assigned to the multiple first subtasks respectively, and each first subtask includes multiple first sub-node tasks connected in series. When the first idle thread corresponding to the first subtask is executed, each first sub-node task in the first idle thread corresponding to the first subtask is executed in sequence. When each first sub-node task is completed, the first subtask is completed.

[0102] Optionally, different first idle threads can be assigned to multiple first subtasks to ensure that each first subtask corresponds to a first idle thread, thereby allowing each first subtask to be executed in parallel with the corresponding first idle thread, thereby improving image reconstruction efficiency. Optionally, the same first idle thread can also be assigned to multiple first subtasks, and each first subtask can be executed sequentially in one first idle thread.

[0103] S303 : Calling the idle threads corresponding to the first subtasks to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0104] In this embodiment, after assigning a different first idle thread to each first subtask and associating multiple first subtasks with multiple first sub-scan data, the idle thread corresponding to each first subtask can be called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0105] In an embodiment of the present application, the scan data is split to obtain multiple segments of first sub-scan data, and the reconstruction task is split to obtain multiple segments of first sub-tasks, and each segment of first sub-scan data is associated with each segment of first sub-task, and the idle thread corresponding to each segment of first sub-task is called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object, thereby improving the reconstruction efficiency of medical images.

[0106] In one embodiment, Figure 3 Based on the embodiment shown, Figure 4 As shown, the above method also includes:

[0107] S304 , segmenting the scan data according to the processing task type corresponding to the irrigation reconstruction type to obtain intermediate segment data corresponding to each processing task type.

[0108] Among them, the processing task types corresponding to the implantation reconstruction type include monitoring task type, filtered back-projection (FBP) task type and artifact removal task type.

[0109] In this embodiment, when the perfusion reconstruction type is used to reconstruct the medical image, the scan data can be segmented according to the monitoring task type, FBP task type and artifact removal task type corresponding to the perfusion reconstruction type to obtain the corresponding intermediate segment data of each processing task type.

[0110] Exemplarily, the scan data may be divided into: an intermediate data segment corresponding to a monitoring task type, an intermediate data segment corresponding to an FBP task type, and an intermediate data segment corresponding to an artifact removal task type.

[0111] The above-mentioned S301 "splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data" includes:

[0112] S301 : Split each intermediate segment of data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0113] In this embodiment, after obtaining the scanning data of the target object scanned by the scanning device as mentioned above, the number of scanning circles corresponding to the scanning data and the idle resources on the scanning device can be obtained, and the number of splits can be determined based on the number of scanning circles corresponding to the scanning data and the idle resources on the scanning device. Each intermediate segment data can be split according to the number of splits to obtain the first sub-scan data of multiple segments corresponding to each processing task type.

[0114] In this embodiment, in the irrigation reconstruction scenario, the scan data is segmented based on the processing task type to obtain multiple intermediate segment data, and the intermediate segment data is split according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data corresponding to each processing task type, which facilitates subsequent data processing of the first sub-scan data according to the processing task type and efficiently realizes the reconstruction of medical images.

[0115] In the irrigation reconstruction scenario, in traditional technology, see Figure 5 , the scan data can be segmented according to the monitoring task type, the filtering back projection task and the artifact removal task, and the intermediate segment data corresponding to the monitoring task, the intermediate segment data corresponding to the filtering back projection task, and the intermediate segment data corresponding to the artifact removal task are obtained, and each intermediate segment data includes the scan data shot1-shotN of multiple scanning time points. In this solution, see Figure 6 , each intermediate segment data can be split into two segments to obtain two segments of first sub-scan data corresponding to each processing task type, so that medical image reconstruction can be performed based on these two segments of first sub-scan data, thereby improving the efficiency of medical image reconstruction.

[0116] In one embodiment, Figure 3 Based on the embodiment shown, Figure 7 As shown, the above S303 "calling the idle threads corresponding to the first subtasks to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object" includes:

[0117] S401 : Allocate corresponding hardware resources to each first idle thread according to the priority of the first idle thread corresponding to the first sub-scanning data of each segment.

[0118] The hardware resources include CPU resources and / or GPU resources.

[0119] In this embodiment, the priority of the first sub-scan data of each segment can be determined according to the importance of the first sub-scan data of each segment, and the priority of the first idle thread corresponding to the first sub-scan data of each segment can be determined according to the priority of the first sub-scan data of each segment, and according to the priority of the first idle thread corresponding to the first sub-scan data of each segment, the GPUmask method and the CPUmask method are used to allocate corresponding CPU resources and / or GPU resources to each first idle thread.

[0120] Optionally, if the importance of the first sub-scan data is high, the priority of the first idle thread corresponding to the first sub-scan data is determined to be high, and when allocating hardware resources, the corresponding CPU resources and / or GPU resources can be allocated to the first idle thread corresponding to the first sub-scan data first; optionally, if the importance of the first sub-scan data is low, the priority of the first idle thread corresponding to the first sub-scan data is determined to be low, and when allocating hardware resources, the corresponding CPU resources and / or GPU resources can be allocated to the first idle thread corresponding to the first sub-scan data last.

[0121] S402: Calling hardware resources corresponding to each first idle thread to perform an image reconstruction operation to obtain a reconstructed image of the target object.

[0122] In this embodiment, after corresponding hardware resources are allocated to each first idle thread, the hardware resources corresponding to each first idle thread may be called to perform an image reconstruction operation to obtain a reconstructed image of the target object.

[0123] In this embodiment, the priority of the first idle thread corresponding to the first sub-scan data of each segment is taken into consideration, and corresponding hardware resources are allocated to each first idle thread according to the priority, thereby avoiding the resource competition problem when the first idle threads are executed in parallel and ensuring the reconstruction efficiency of medical images.

[0124] In one embodiment, Figure 2 Based on the embodiment shown, when the reconstruction task type is an artificial intelligence reconstruction type, the reconstruction task includes a longitudinal task and a transverse task, such as Figure 8 As shown, the above S203 "performing image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object" includes:

[0125] S501 , splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data.

[0126] Among them, horizontal tasks refer to tasks that can be performed in parallel with artificial intelligence reconstruction tasks, such as tomography reconstruction tasks.

[0127] In this embodiment, when the reconstruction task type is determined to be an artificial intelligence reconstruction type, the number of lateral tasks corresponding to the artificial intelligence reconstruction type is determined, and the scan data is split according to the number of lateral tasks to obtain the same number of second sub-scan data as the number of lateral tasks. The reconstruction task is then divided into the same number of second sub-tasks as the number of lateral tasks, and the multiple second sub-tasks are associated with the multiple second sub-scan data. It should be noted that associating the multiple second sub-tasks with the multiple second sub-scan data specifically means associating the second sub-scan data with the second sub-tasks, setting an association tag, or generating a corresponding relationship; thereby, the second idle threads corresponding to the second sub-tasks can be called to perform image reconstruction operations in parallel based on the associated second sub-scan data.

[0128] S502. Split the second scan data associated with each second subtask according to the vertical task to obtain multiple segments of third sub-scan data, divide the vertical task into multiple second sub-node tasks connected in series, and associate the multiple second sub-node tasks with multiple third sub-scan data.

[0129] Among them, longitudinal tasks refer to tasks that can only be performed serially during artificial intelligence reconstruction tasks, such as spiral scanning reconstruction tasks.

[0130] In this embodiment, after determining the second scan data associated with the second subtask, the second scan data associated with each second subtask can be split according to the longitudinal task to obtain multiple segments of third sub-scan data corresponding to the second scan data associated with each second subtask. The longitudinal task can then be divided into multiple second sub-node tasks connected in series, and the multiple second sub-node tasks can be associated with the multiple third sub-scan data. It should be noted that associating multiple second sub-node tasks with multiple third sub-scan data specifically means associating the third sub-scan data with the second sub-node tasks, setting an association label, or generating a corresponding relationship; thereby, the graphics card resources corresponding to the second sub-node tasks can be called to perform image reconstruction operations in parallel based on the associated third sub-scan data.

[0131] In this embodiment, in the artificial intelligence reconstruction scenario, the second scan data associated with each second sub-task is split based on the longitudinal task to obtain multiple segments of third sub-scan data, and the longitudinal task is divided into multiple second sub-node tasks, and then each second sub-node task is associated with each third scan data, so as to facilitate subsequent data processing of the first sub-scan data according to the processing task type and efficiently realize the reconstruction of medical images.

[0132] S503: Allocate second idle threads to the plurality of second subtasks.

[0133] Each second subtask includes a plurality of second sub-node tasks connected in series, and each second sub-node task is executed in sequence.

[0134] In this embodiment, after the multiple second subtasks are determined as described above, second idle threads can be allocated to the multiple second subtasks respectively, and each second subtask includes multiple second sub-node tasks connected in series. When the second idle thread corresponding to the second subtask is executed, each second sub-node task in the second idle thread corresponding to the second subtask is executed in sequence. When each second sub-node task is completed, the second subtask is completed.

[0135] Alternatively, different second idle threads can be assigned to multiple second subtasks, ensuring that each second subtask corresponds to a second idle thread. This allows each second subtask to be executed in parallel with the corresponding second idle thread, thereby improving image reconstruction efficiency. Alternatively, the same second idle thread can be assigned to multiple second subtasks, and each second subtask can be executed sequentially in one second idle thread.

[0136] S504 : Call the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0137] In this embodiment, after assigning a different second idle thread to each second subtask and associating multiple second subtasks with multiple second sub-scan data, the second idle thread corresponding to each second subtask can be called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0138] In an embodiment of the present application, the scan data is split to obtain multiple segments of second sub-scan data, and the horizontal task is split to obtain multiple second sub-tasks, and each segment of the second sub-scan data is associated with each segment of the second sub-task, and the idle thread corresponding to each segment of the second sub-task is called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object, thereby improving the reconstruction efficiency of medical images.

[0139] In the AIIR reconstruction scenario, in traditional technology, see Figure 9, the reconstruction task can be split into task 1, task 2 and task 3, and GPU resources can be uniformly allocated to task 1, task 2 and task 3 to obtain the GPU resources corresponding to task 1, the GPU resources corresponding to task 2, and the GPU resources corresponding to task 3. Based on the GPU resources corresponding to task 1, the GPU resources corresponding to task 2, the GPU resources corresponding to task 3, and the thread corresponding to the reconstruction task, image reconstruction can be implemented. For example, 40% of GPU resources 1 can be allocated to task 1, 30% of GPU resources 1 can be allocated to task 2, and 30% of GPU resources 1 can be allocated to task 3; 40% of GPU resources 2 can be allocated to task 1, 30% of GPU resources 2 can be allocated to task 2, and 30% of GPU resources 2 can be allocated to task 3; 40% of GPU resources 3 can be allocated to task 1, 30% of GPU resources 3 can be allocated to task 2, and 30% of GPU resources 3 can be allocated to task 3; 40% of GPU resources 4 can be allocated to task 1, 30% of GPU resources 4 can be allocated to task 2, and 30% of GPU resources 4 can be allocated to task 3, and one thread corresponding to the reconstruction task can be used to implement image reconstruction.

[0140] In this application, see Figure 10 , GPU resource 1 can be assigned to thread 1, GPU resource 2 to thread 2, and GPU resources 3 and 4 to thread 3. Threads 1, 2, and 3 are executed in parallel to achieve efficient image reconstruction. Furthermore, the tasks within GPU resources 1, 2, and 3 are executed sequentially. That is, in the thread corresponding to CPU resource 1, 30% of task 2 is executed after 40% of task 1 is completed, and 30% of task 2 is executed after 30% of task 2 is completed. In the thread corresponding to CPU resource 2, 30% of task 2 is executed after 40% of task 1 is completed, and 30% of task 3 is executed after 30% of task 2 is completed. The threads corresponding to CPU resource 1 and CPU resource 2 can execute in parallel. In the thread corresponding to CPU resource 3, 30% of task 2 is executed after 40% of task 1 is completed, 30% of task 2 is executed after 30% of task 2 is completed, and 30% of task 3 is executed after 30% of task 3 is completed. After 40% of task 1 is completed, 30% of task 2 is executed, and 30% of task 3 is executed after 30% of task 2 is completed. The threads corresponding to CPU resource 1 and CPU resource 2 can execute in parallel. In the thread corresponding to CPU resource 3, 30% of task 2 is executed after 40% of task 1 is completed, 30% of task 2 is executed after 30% of task 2 is completed, and 40% of task 1 is executed after 40% of task 1 is completed, and 30% of task 2 is executed after 30% of task 2 is completed.

[0141] In one embodiment, Figure 8 Based on the embodiment shown, Figure 11 As shown, the above S503 "calling the idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object" includes:

[0142] S601: Determine a target second subtask according to the execution order of multiple second subtasks.

[0143] In the embodiment of the present application, after the above splitting to obtain multiple second subtasks, the second subtask that is executed first can be determined as the target second subtask according to the execution order of each second subtask.

[0144] S602 : Calling the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image.

[0145] The first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing.

[0146] In this embodiment, after the target second subtask is determined as described above, the idle thread corresponding to the target second subtask is called to perform the first image reconstruction operation, obtain the reconstructed image corresponding to the first image, and generate pre-processing intermediate data and post-processing intermediate data in the process of executing the first image reconstruction operation.

[0147] S603. Call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the associated scanning data, to obtain a reconstructed image of the target object.

[0148] The other second subtasks refer to the other second subtasks among the multiple second subtasks except the target second subtask.

[0149] In this embodiment, after the target second subtask executes the first image reconstruction operation and generates pre-processing intermediate data and post-processing intermediate data, the idle threads corresponding to other second subtasks are called to obtain the pre-processing intermediate data and / or post-processing intermediate data generated during the target second subtask's execution of the first image reconstruction operation, and the image reconstruction operation is performed in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the scanning data associated with other second subtasks, to obtain a reconstructed image of the target object.

[0150] In this embodiment, according to the execution order of each second subtask, the idle thread corresponding to the target second subtask is called to perform the image reconstruction operation, and pre-processing intermediate data and post-processing intermediate data are generated in the process of executing the first image reconstruction operation. The idle threads corresponding to other second subtasks are called to obtain the pre-processing intermediate data and / or post-processing intermediate data to execute the image reconstruction operation in parallel, avoiding repeated execution of intermediate data, reducing the amount of repeated calculations in the image reconstruction process, and thereby improving the reconstruction efficiency of medical images.

[0151] See also Figure 12 After adding multiple reconstruction tasks online and offline, you can first obtain the intermediate data in the offline reconstruction task, and execute task 1 among the multiple reconstruction tasks based on thread 1, that is, gradually execute raw data acquisition, algorithm 1, algorithm 2, algorithm 3, obtain pre-processing intermediate data, continue to execute algorithm 4, algorithm 5 and algorithm 6, obtain intermediate projection data, and then perform image post-processing to obtain reconstructed image 1. When thread 2 executes task 2, you can directly obtain the pre-processing intermediate data from thread 1, continue to execute algorithm 4, algorithm 5 and algorithm 6, obtain intermediate projection data, and then perform image post-processing to obtain reconstructed image 2. When thread 3 executes task 3, you can directly obtain intermediate projection data from thread 1, continue to execute image post-processing, and obtain reconstructed image 3. In addition, for task 2 corresponding to thread 2 that has been suspended, when executing, you can directly obtain pre-processing intermediate data from thread 1, continue to execute algorithm 4, algorithm 5 and algorithm 6, obtain intermediate projection data, and then perform image post-processing to obtain reconstructed image 2.

[0152] In one embodiment, Figure 8 Based on the embodiment shown, the above S503 "calling the idle threads corresponding to the respective second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object" includes:

[0153] During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

[0154] In this embodiment, after the multiple second subtasks are determined as described above, for a single second subtask, during the execution of the second idle thread corresponding to the single second subtask, multiple graphics card resources corresponding to the second idle thread corresponding to the single second subtask are scheduled, and image reconstruction operations are performed in parallel on the third sub-scan data associated with the multiple second sub-node tasks contained in the second subtask to obtain a reconstructed image of the target object.

[0155] In this embodiment, during the execution of a second subtask, multiple graphics card resources corresponding to the second idle thread corresponding to the second subtask are scheduled to perform image reconstruction operations in parallel to obtain a reconstructed image of the target object, thereby efficiently realizing the reconstruction of medical images.

[0156] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0157] Based on the same inventive concept, embodiments of the present application also provide a medical image reconstruction device for implementing the aforementioned medical image reconstruction method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more medical image reconstruction device embodiments provided below can be found in the limitations of the medical image reconstruction method described above and will not be further elaborated here.

[0158] In one embodiment, Figure 13 As shown, a medical image reconstruction method is provided, which is described by taking the method applied to a computed tomography (CT) device as an example, and includes the following steps:

[0159] S701: Obtain scanning data of a target object scanned by a CT device.

[0160] The target object may be any part of the patient's body, such as the abdomen, chest, legs, hands, etc.

[0161] In this embodiment, when medical image reconstruction is required based on the scanning data obtained by the CT device, the data acquisition module in the CT device can first scan the target object to obtain the scanning data corresponding to the target object, and then the processor in the CT device obtains the scanning data corresponding to the target object from the data acquisition module.

[0162] Optionally, the data acquisition module in the CT device can scan the target object in advance, and after obtaining the scanning data corresponding to the target object, save the scanning data corresponding to the target object. Then, the processor in the CT device can obtain the scanning data corresponding to the target object from the data acquisition module at any time when medical image reconstruction is required.

[0163] Optionally, after the data acquisition module in the CT device scans the target object and obtains the scan data corresponding to the target object, the processor in the CT device may directly obtain the scan data corresponding to the target object from the data acquisition module.

[0164] S702: Determine a corresponding image reconstruction strategy according to the type of reconstruction task.

[0165] Among them, the types of reconstruction tasks include perfusion reconstruction type or artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism of starting multiple processes corresponding to multiple subtasks to execute in parallel to perform image reconstruction.

[0166] In this embodiment, a user (e.g., a medical professional scanning a target object) can select a reconstruction task type on the CT device. After receiving the reconstruction task type, the CT device determines an image reconstruction strategy corresponding to the reconstruction task type. It should be noted that there is a one-to-one correspondence between the reconstruction task type and the image reconstruction strategy.

[0167] Optionally, the image reconstruction strategy corresponding to the infusion reconstruction type may be to segment the scan data to obtain multiple scan data segments, configure corresponding subtasks for each scan data segment, and then start a mechanism for parallel execution of multiple processes corresponding to the multiple subtasks to perform image reconstruction; the image reconstruction strategy corresponding to the artificial intelligence reconstruction type may be to segment the scan task to obtain multiple scan task segments, and start multiple processes corresponding to the multiple scan task segments according to the scan data, and perform image reconstruction by a mechanism for parallel execution.

[0168] S703: Perform image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0169] In this embodiment, after the image reconstruction strategy and the scan data are determined as described above, image reconstruction can be performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0170] For example, when the reconstruction task is of the perfusion reconstruction type, the target object needs to be scanned multiple times, and there is a problem of a large amount of scan data. The scan data can be segmented to obtain multiple scan data segments, and corresponding subtasks can be configured for each scan data segment. Then, a mechanism of parallel execution of multiple processes corresponding to multiple subtasks is started to perform image reconstruction to obtain a reconstructed image of the target object.

[0171] Another example is that when the type of reconstruction task is an artificial intelligence reconstruction type, multiple iterative reconstructions and deep learning are required to achieve the reconstruction of medical images, and each reconstruction includes forward projection and back projection. There is a problem of large task volume. The scanning task can be segmented to obtain multiple scanning task segments, and multiple processes corresponding to the multiple scanning task segments are started according to the scanning data. The image is reconstructed by a parallel execution mechanism to obtain a reconstructed image of the target object.

[0172] The medical image reconstruction method provided in the embodiment of the present application realizes the reconstruction task of the medical image by starting multiple processes corresponding to multiple subtasks in the reconstruction task for parallel execution. Compared with the existing method of realizing the reconstruction of medical images by executing serial processes, the present application greatly improves the reconstruction efficiency of medical images. In addition, the method takes into account the types of different reconstruction tasks, determines the corresponding image reconstruction strategies, and realizes the reconstruction of medical images based on the image reconstruction strategies corresponding to the reconstruction tasks. Compared with the existing method of realizing the reconstruction of medical images by only providing one reconstruction strategy, the present application greatly improves the reconstruction flexibility of medical images.

[0173] In an exemplary embodiment, Figure 14 As shown, a medical image reconstruction device is provided, comprising: an acquisition module 10, a determination module 11 and a reconstruction module 12, wherein:

[0174] The acquisition module 10 is used to acquire scan data of a target object scanned by a scanning device.

[0175] Determination module 11 is used to determine the corresponding image reconstruction strategy according to the type of reconstruction task; the type of reconstruction task includes perfusion reconstruction type or artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for starting multiple processes corresponding to multiple subtasks to execute in parallel to perform image reconstruction.

[0176] The reconstruction module 12 is configured to perform image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0177] In an exemplary embodiment, the type of the reconstruction task is a perfusion reconstruction type, and the reconstruction module 12 includes: a first splitting unit, a first allocation unit, and a first reconstruction unit, wherein:

[0178] a first splitting unit, specifically configured to split the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, divide the reconstruction task into a plurality of first sub-tasks, and associate the plurality of first sub-tasks with the plurality of first sub-scan data;

[0179] A first allocation unit is specifically configured to allocate first idle threads to a plurality of first subtasks; each first subtask includes a plurality of first sub-node tasks connected in series;

[0180] The first reconstruction unit is specifically configured to call the idle threads corresponding to the first subtasks to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0181] In an exemplary embodiment, the reconstruction module 12 further includes: a segmentation unit, specifically configured to segment the scan data according to the processing task type corresponding to the irrigation reconstruction type, to obtain intermediate segment data corresponding to each processing task type;

[0182] The first splitting unit is specifically configured to split each intermediate segment of data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0183] In an exemplary embodiment, the above-mentioned reconstruction unit is further used to allocate corresponding hardware resources to each idle thread according to the priority of the idle thread corresponding to the first sub-scan data of each segment; the hardware resources include CPU resources and / or GPU resources; and the hardware resources corresponding to each idle thread are called to perform image reconstruction operations to obtain a reconstructed image of the target object.

[0184] In an exemplary embodiment, the reconstruction task is of an artificial intelligence reconstruction type, and the reconstruction task includes a longitudinal task and a transverse task. The reconstruction module further includes: a second splitting unit, a third splitting unit, a second allocation unit, and a second reconstruction unit, wherein:

[0185] a second splitting unit, specifically configured to split the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, divide the horizontal task into multiple second sub-tasks, and associate the multiple second sub-tasks with the multiple second sub-scan data;

[0186] a third splitting unit, specifically configured to split the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, divide the longitudinal task into multiple second sub-node tasks connected in series, and associate the multiple second sub-node tasks with the multiple third sub-scan data;

[0187] A second allocation unit, specifically configured to allocate second idle threads to the plurality of second subtasks;

[0188] The second reconstruction unit is specifically configured to call the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0189] In an exemplary embodiment, the above-mentioned second reconstruction unit is further used to determine the target second subtask according to the execution order of multiple second subtasks; call the idle thread corresponding to the target second subtask to perform the first image reconstruction operation, and generate pre-processing intermediate data and post-processing intermediate data in the process of executing the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back projection processing, and beam rearrangement processing; call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, and associated scanning data to obtain a reconstructed image of the target object.

[0190] In an exemplary embodiment, the above-mentioned second reconstruction unit is further used to schedule multiple graphics card resources to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks contained in the second sub-task during the execution of the second idle thread corresponding to a single second sub-task, so as to obtain a reconstructed image of the target object.

[0191] In an exemplary embodiment, Figure 15 As shown, a medical image reconstruction device is provided, which is applied to a computerized tomography (CT) device, including: an acquisition module 20, a determination module 21 and a reconstruction module 22, wherein:

[0192] An acquisition module 20 is used to acquire scan data of a target object scanned by a CT device;

[0193] A determination module 21 is configured to determine a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; and the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction.

[0194] The reconstruction module 22 is configured to perform image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

[0195] Each module in the medical image reconstruction device described above may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0196] In an exemplary embodiment, a scanning device is provided. The scanning device may be a server, and its internal structure may be as shown in FIG. Figure 16As shown. The scanning device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the scanning device is used to provide computing and control capabilities. The memory of the scanning device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the scanning device is used to store scan data. The input / output interface of the scanning device is used to exchange information between the processor and an external device. The communication interface of the scanning device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for reconstructing medical images is implemented.

[0197] Those skilled in the art will understand that Figure 16 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0198] In an exemplary embodiment, a scanning device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0199] Obtaining scan data of a target object scanned by a scanning device;

[0200] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0201] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0202] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0203] Splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, dividing the reconstruction task into a plurality of first sub-tasks, and associating the plurality of first sub-tasks with the plurality of first sub-scan data;

[0204] Allocating first idle threads to a plurality of first subtasks; each first subtask includes a plurality of first subnode tasks connected in series;

[0205] The idle threads corresponding to the first subtasks are called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0206] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0207] According to the processing task type corresponding to the irrigation reconstruction type, the scan data is segmented to obtain the intermediate segment data corresponding to each processing task type;

[0208] The scan data is split according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data, including:

[0209] According to the number of scan circles corresponding to the scan data and the idle resources on the scanning device, each intermediate segment of data is split to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0210] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0211] Allocate corresponding hardware resources to each idle thread according to the priority of the idle thread corresponding to the first sub-scan data of each segment; the hardware resources include CPU resources and / or GPU resources;

[0212] The hardware resources corresponding to each idle thread are called to perform image reconstruction operations to obtain a reconstructed image of the target object.

[0213] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0214] Splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data;

[0215] Splitting the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, dividing the longitudinal task into multiple second sub-node tasks connected in series, and associating the multiple second sub-node tasks with the multiple third sub-scan data;

[0216] Allocating a second idle thread to the plurality of second subtasks;

[0217] The second idle threads corresponding to the second subtasks are called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0218] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0219] Determining a target second subtask according to the execution order of the plurality of second subtasks;

[0220] Invoking the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing;

[0221] Call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the associated scanning data, to obtain a reconstructed image of the target object.

[0222] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0223] During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

[0224] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0225] Obtaining scan data of a target object scanned by a CT device;

[0226] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0227] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0228] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0229] Obtaining scan data of a target object scanned by a scanning device;

[0230] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0231] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0232] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0233] Splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, dividing the reconstruction task into a plurality of first sub-tasks, and associating the plurality of first sub-tasks with the plurality of first sub-scan data;

[0234] Allocating first idle threads to a plurality of first subtasks; each first subtask includes a plurality of first subnode tasks connected in series;

[0235] The idle threads corresponding to the first subtasks are called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0236] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0237] According to the processing task type corresponding to the irrigation reconstruction type, the scan data is segmented to obtain the intermediate segment data corresponding to each processing task type;

[0238] The scan data is split according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data, including:

[0239] According to the number of scan circles corresponding to the scan data and the idle resources on the scanning device, each intermediate segment of data is split to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0240] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0241] Allocate corresponding hardware resources to each idle thread according to the priority of the idle thread corresponding to the first sub-scan data of each segment; the hardware resources include CPU resources and / or GPU resources;

[0242] The hardware resources corresponding to each idle thread are called to perform image reconstruction operations to obtain a reconstructed image of the target object.

[0243] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0244] Splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data;

[0245] Splitting the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, dividing the longitudinal task into multiple second sub-node tasks connected in series, and associating the multiple second sub-node tasks with the multiple third sub-scan data;

[0246] Allocating a second idle thread to the plurality of second subtasks;

[0247] The second idle threads corresponding to the second subtasks are called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0248] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0249] Determining a target second subtask according to the execution order of the plurality of second subtasks;

[0250] Invoking the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing;

[0251] Call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the associated scanning data, to obtain a reconstructed image of the target object.

[0252] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0253] During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

[0254] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0255] Obtaining scan data of a target object scanned by a CT device;

[0256] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0257] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0258] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0259] Obtaining scan data of a target object scanned by a scanning device;

[0260] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0261] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0262] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0263] Splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, dividing the reconstruction task into a plurality of first sub-tasks, and associating the plurality of first sub-tasks with the plurality of first sub-scan data;

[0264] Allocating first idle threads to a plurality of first subtasks; each first subtask includes a plurality of first subnode tasks connected in series;

[0265] The idle threads corresponding to the first subtasks are called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

[0266] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0267] According to the processing task type corresponding to the irrigation reconstruction type, the scan data is segmented to obtain the intermediate segment data corresponding to each processing task type;

[0268] The scan data is split according to the number of scan circles corresponding to the scan data and the idle resources on the scanning device to obtain multiple segments of first sub-scan data, including:

[0269] According to the number of scan circles corresponding to the scan data and the idle resources on the scanning device, each intermediate segment of data is split to obtain multiple segments of first sub-scan data corresponding to each processing task type.

[0270] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0271] Allocate corresponding hardware resources to each idle thread according to the priority of the idle thread corresponding to the first sub-scan data of each segment; the hardware resources include CPU resources and / or GPU resources;

[0272] The hardware resources corresponding to each idle thread are called to perform image reconstruction operations to obtain a reconstructed image of the target object.

[0273] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0274] Splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data;

[0275] Splitting the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, dividing the longitudinal task into multiple second sub-node tasks connected in series, and associating the multiple second sub-node tasks with the multiple third sub-scan data;

[0276] Allocating a second idle thread to a plurality of second subtasks; each second subtask includes a plurality of second sub-node tasks connected in series;

[0277] The second idle threads corresponding to the second subtasks are called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

[0278] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0279] Determining a target second subtask according to the execution order of the plurality of second subtasks;

[0280] Invoking the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing;

[0281] Call the idle threads corresponding to other second subtasks in the multiple second subtasks to obtain pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, as well as the associated scanning data, to obtain a reconstructed image of the target object.

[0282] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0283] During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

[0284] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0285] Obtaining scan data of a target object scanned by a CT device;

[0286] Determining a corresponding image reconstruction strategy based on the type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating multiple processes corresponding to the multiple subtasks to execute in parallel to perform image reconstruction;

[0287] Image reconstruction is performed according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

[0288] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0289] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0290] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A medical image reconstruction method, characterized in that: The method comprises: Obtaining scan data of a target object scanned by a scanning device; Determining a corresponding image reconstruction strategy based on a type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating parallel execution of multiple processes corresponding to the multiple subtasks to perform image reconstruction; Image reconstruction is performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

2. The method according to claim 1, characterized in that The type of the reconstruction task is the perfusion reconstruction type, and performing image reconstruction according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object includes: Splitting the scan data according to the number of scan revolutions corresponding to the scan data and idle resources on the scanning device to obtain a plurality of segments of first sub-scan data, dividing the reconstruction task into a plurality of first sub-tasks, and associating the plurality of first sub-tasks with the plurality of first sub-scan data; Allocating a first idle thread to the plurality of first subtasks; each of the first subtasks includes a plurality of first sub-node tasks connected in series; The first idle threads corresponding to the first subtasks are called to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object.

3. The method according to claim 2, characterized in that The method further comprises: Segmenting the scan data according to the processing task type corresponding to the irrigation reconstruction type to obtain intermediate segment data corresponding to each processing task type; The step of splitting the scan data according to the number of scan turns corresponding to the scan data and the idle resources on the scanning device to obtain a plurality of segments of first sub-scan data includes: According to the number of scan circles corresponding to the scan data and the idle resources on the scanning device, each intermediate segment of data is split to obtain multiple segments of first sub-scan data corresponding to each processing task type.

4. The method according to claim 2, characterized in that The calling of the first idle threads corresponding to the first subtasks to perform image reconstruction operations in parallel based on the associated first sub-scan data to obtain a reconstructed image of the target object includes: Allocate corresponding hardware resources to each first idle thread according to the priority of the first idle thread corresponding to each segment of the first sub-scan data; the hardware resources include CPU resources and / or GPU resources; The hardware resources corresponding to each of the first idle threads are called to perform an image reconstruction operation to obtain a reconstructed image of the target object.

5. The method according to claim 1, wherein The type of the reconstruction task is the artificial intelligence reconstruction type, the reconstruction task includes a longitudinal task and a transverse task, and the image reconstruction is performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object, including: Splitting the scan data according to the horizontal task to obtain multiple segments of second sub-scan data, dividing the horizontal task into multiple second sub-tasks, and associating the multiple second sub-tasks with the multiple second sub-scan data; splitting the second scan data associated with each second subtask according to the longitudinal task to obtain multiple segments of third sub-scan data, dividing the longitudinal task into multiple second sub-node tasks connected in series, and associating the multiple second sub-node tasks with the multiple third sub-scan data; and allocating a second idle thread to the multiple second subtasks; The second idle threads corresponding to the second subtasks are called to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object.

6. The method according to claim 5, characterized in that The calling of the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object includes: Determining a target second subtask according to the execution order of the plurality of second subtasks; Invoking the idle thread corresponding to the target second subtask to perform a first image reconstruction operation, and generating pre-processing intermediate data and post-processing intermediate data during the execution of the first image reconstruction operation to obtain a reconstructed image corresponding to the first image; the first image reconstruction operation includes image correction processing, filtering processing, back-projection processing, and beam rearrangement processing; Call the idle threads corresponding to other second subtasks among the multiple second subtasks to obtain the pre-processing intermediate data and / or post-processing intermediate data, and perform image reconstruction operations in parallel based on the pre-processing intermediate data and / or post-processing intermediate data, and the associated scanning data to obtain a reconstructed image of the target object.

7. The method according to claim 5, characterized in that The calling of the second idle threads corresponding to the second subtasks to perform image reconstruction operations in parallel based on the associated second sub-scan data to obtain a reconstructed image of the target object includes: During the execution of the second idle thread corresponding to a single second subtask, multiple graphics card resources are scheduled to perform image reconstruction operations in parallel on the third sub-scan data associated with multiple second sub-node tasks included in the second subtask to obtain a reconstructed image of the target object.

8. A medical image reconstruction method, characterized in that: Applied to a computerized tomography (CT) device, the method comprises: Obtaining scan data of a target object scanned by a CT device; Determining a corresponding image reconstruction strategy based on a type of reconstruction task; the type of reconstruction task includes a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task includes multiple subtasks; the image reconstruction strategy includes a mechanism for initiating parallel execution of multiple processes corresponding to the multiple subtasks to perform image reconstruction; Image reconstruction is performed according to the image reconstruction strategy and the scan data to obtain a reconstructed image of the target object.

9. A medical image reconstruction device, characterized in that: The device comprises: An acquisition module is used to acquire scanning data of a target object scanned by a scanning device; a determination module, configured to determine a corresponding image reconstruction strategy based on a type of reconstruction task; the type of reconstruction task including a perfusion reconstruction type or an artificial intelligence reconstruction type; the reconstruction task including a plurality of subtasks; and the image reconstruction strategy including a mechanism for initiating parallel execution of a plurality of processes corresponding to the plurality of subtasks to perform image reconstruction; A reconstruction module is used to perform image reconstruction according to the image reconstruction strategy and the scanning data to obtain a reconstructed image of the target object.

10. A scanning device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images