Image generation method, device, equipment and computer program product

Abstract

The embodiment of the invention provides an image generation method, device and equipment and a computer program product. The image generation method comprises the steps that an artery spin labeling ASL image, reference physiological parameters and experiment physiological parameters are obtained, the reference physiological parameters and the ASL image correspond to the same subject in a standard gravity environment, and the experiment physiological parameters correspond to a subject in a microgravity experiment environment; determining an image generation model used for generating an ASL image in the microgravity experiment environment; and processing the ASL image, the reference physiological parameters and the experimental physiological parameters by using an image generation model, and determining a target ASL image of the main body in the microgravity experimental environment. In the embodiment, the ASL image in the microgravity experiment environment is obtained based on the ASL image in the standard gravity environment, the reference physiological parameters and the experiment physiological parameters, and convenience and reliability of ASL image obtaining in the microgravity environment are effectively guaranteed.

Description

Technical Field

[0001] This invention relates to the field of image technology, and in particular to an image generation method, apparatus, device, and computer program product. Background Technology

[0002] Cerebral perfusion reflects the blood supply to the brain and is an important physiological indicator of the health and function of the brain's neurovascular system. Because the brain has a high demand for oxygen and nutrients, its normal function is highly dependent on a continuous and adequate blood flow. During spaceflight, changes in gravity vectors and posture cause fluid shifts within the body. This headward distribution of fluids leads to blood pooling in the brain, which can potentially increase intracranial pressure (ICP), thereby altering cerebral perfusion. Most reported neurological symptoms and cognitive impairments experienced by astronauts in space are due to changes in cerebral perfusion.

[0003] Currently, arterial spin labeling (ASL) perfusion magnetic resonance imaging (MRI) is a common non-invasive, non-radioactive technique for measuring cerebral blood flow perfusion in quantitative units (ml / 100g / min). However, limitations in the size, transport requirements, and logistical support of MRI equipment prevent its installation on space stations. Therefore, there is an urgent need for a technical solution capable of acquiring ASL images under microgravity conditions to enable the assessment of cerebral blood flow perfusion based on the acquired ASL images. Summary of the Invention

[0004] This invention provides an image generation method, apparatus, device, and computer program product that can acquire ASL images under microgravity experimental conditions to detect the brain health status of a subject based on the acquired ASL images.

[0005] In a first aspect, embodiments of the present invention provide an image generation method, comprising:

[0006] Arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters were acquired, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment.

[0007] Determine the image generation model for generating ASL images in a microgravity experimental environment;

[0008] The ASL image, baseline physiological parameters, and experimental physiological parameters are processed using the image generation model to determine the target ASL image of the subject in a microgravity experimental environment.

[0009] In a second aspect, embodiments of the present invention provide an image generation apparatus, comprising:

[0010] The first acquisition module is used to acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment.

[0011] The first determining module is used to determine the image generation model for generating ASL images in a microgravity experimental environment;

[0012] The first processing module is used to process the ASL image, baseline physiological parameters, and experimental physiological parameters using the image generation model to determine the target ASL image of the subject located in a microgravity experimental environment.

[0013] Thirdly, embodiments of the present invention provide an electronic device, including: a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions, when executed by the processor, implement the image generation method in the first aspect described above.

[0014] Fourthly, embodiments of the present invention provide a computer storage medium for storing a computer program that, when executed by a computer, implements the image generation method described in the first aspect above.

[0015] Fifthly, embodiments of the present invention provide a computer program product, comprising: a computer-readable storage medium storing computer instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the image generation method described in the first aspect above.

[0016] The image generation method, apparatus, device, and computer program product provided in this invention acquires an arterial spin marker (ASL) image, baseline physiological parameters, and experimental physiological parameters of the same subject. Then, an image generation model is determined to generate an ASL image in a microgravity experimental environment. This model is used to process the ASL image, baseline physiological parameters, and experimental physiological parameters to determine the target ASL image of the subject in the microgravity experimental environment. Since there is a correlation between microgravity experimental environments and microgravity environments, the obtained target ASL image in the microgravity experimental environment can be considered as the ASL image of the subject in the microgravity environment. This eliminates the need for specific equipment deployment in the microgravity environment to acquire ASL images, effectively ensuring the convenience and reliability of ASL image acquisition in microgravity. Furthermore, the obtained ASL image can be used to detect the brain health status of the subject, further improving the practicality of the method and facilitating its market promotion and application. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram illustrating the principle of an image generation method provided in an embodiment of the present invention;

[0019] Figure 2 This is a schematic flowchart of an image generation method provided in an embodiment of the present invention;

[0020] Figure 3 This is a schematic diagram illustrating the process of determining a target ASL image using an image generation model, as provided in an embodiment of the present invention.

[0021] Figure 4 A schematic diagram illustrating the principle of a generative adversarial network provided in an embodiment of the present invention;

[0022] Figure 5 This is a schematic diagram illustrating the process of optimizing an image generation model according to an embodiment of the present invention;

[0023] Figure 6 A schematic diagram illustrating another image generation model optimization process provided in an embodiment of the present invention;

[0024] Figure 7This is a schematic diagram illustrating a process for obtaining ground truth edge maps and target edge maps using an edge detector, provided as an application embodiment of the present invention.

[0025] Figure 8 This is a flowchart illustrating the process of determining the brain health status of each brain region after determining the target ASL image of the subject in a microgravity experimental environment, as provided in an embodiment of the present invention.

[0026] Figure 9 This is a schematic diagram of another process for determining the brain health status of each brain region after determining the target ASL image of the subject in a microgravity experimental environment, as provided in an embodiment of the present invention.

[0027] Figure 10 This is a schematic diagram of the structure of an image generation device provided in an embodiment of the present invention;

[0028] Figure 11 To and Figure 10 A schematic diagram of the electronic device corresponding to the image generation apparatus provided in the embodiment is shown. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise; “multiple” generally includes at least two, but does not exclude the inclusion of at least one.

[0031] It should be understood that the term "and / or" used in this document is merely a description of the relationship between associated devices, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " in this document generally indicates that the preceding and following associated devices are in an "or" relationship.

[0032] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0033] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes that element.

[0034] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0035] Terminology Explanation:

[0036] Arterial Spin Labeling (ASL) is a non-invasive imaging technique for measuring arterial blood flow that uses arterial blood as an endogenous contrast agent and employs radiofrequency pulses to magnetically label the incoming blood.

[0037] Microgravity environment: Microgravity refers to an acceleration caused by gravity or other external forces not exceeding 10 eJ / m². -5 ~10e -4 g, where g represents the gravitational acceleration under normal gravity conditions. Microgravity is used to represent a low level of gravity; therefore, a microgravity environment refers to an environment under microgravity conditions, such as the space environment.

[0038] Standard gravity environment: refers to the environment under normal gravity conditions when on the ground.

[0039] Microgravity experimental environment: refers to an experimental environment that simulates microgravity under gravity conditions, specifically designed and constructed for conducting specific scientific research or technology verification.

[0040] Cerebral perfusion: hemodynamics at the level of brain tissue microcirculation.

[0041] To facilitate understanding of the specific implementation process and effects of the image generation method, apparatus, device, and computer program product in this embodiment, the relevant technologies are briefly described below:

[0042] During spaceflight, changes in gravity vectors and attitude cause fluid shifts within the body. This redistribution of fluid in the head can lead to blood pooling in the brain, potentially increasing intracranial pressure and altering cerebral perfusion. Most reported neurological symptoms in astronauts during space travel are due to these changes. While cerebral perfusion can be measured using various neuroimaging techniques, most require exogenous radioactive tracers or contrast agents that can cause side effects. Magnetic Resonance Imaging (MRI) is the only non-invasive, non-radioactive technique that measures whole-brain perfusion in quantitative units (ml / 100g / min). Because it is easy to perform and does not involve the injection of contrast agents (such as gadolinium) or ionizing radiation, it is widely used in cerebral perfusion measurements and is well-suited for testing in sensitive populations with limited access to injectable substances, such as astronauts. However, due to limitations in the size, payload, and logistics of MRI equipment, it is impossible to install it inside the space station. Consequently, it is impossible to obtain ASL images of astronauts using MRI equipment in the microgravity environment of the space station, and thus, it is impossible to obtain information on changes in brain perfusion of astronauts.

[0043] Currently, the method for acquiring Arterial Spin Labeling (ASL) images under microgravity is based on deep learning-based medical image synthesis. This involves learning a nonlinear mapping to transform the image from the source domain X into the target domain Y. In other words, by establishing a mapping relationship between baseline ASL images and ASL images under microgravity, a deep generative model is used to generate ASL images under microgravity, and the brain health status is then detected based on the obtained ASL images.

[0044] However, the spatial resolution of ASL images is typically much lower than the cortical thickness (average 2.5 × 2.5 × 2.5 mm). 3 A typical ASL resolution is 4×4×4mm. 3 Therefore, the aforementioned deep generative models often struggle to fully capture the blurred tissue type boundaries in ASL images, exacerbating the partial volume effect and reducing the quantitative accuracy of ASL perfusion, thus affecting the detection results of brain health status. The partial volume effect refers to the fact that in MRI, each voxel represents a volume element; if a voxel contains multiple different tissues (e.g., gray matter, white matter, and cerebrospinal fluid), its signal will be a mixture of these tissue signals.

[0045] To address the aforementioned technical problems, this embodiment provides an image generation method, apparatus, device, and computer program product, as detailed in the appendix. Figure 1As shown, the image generation method is executed by an image generation device 200, which is communicatively connected to an information acquisition device 100. The image generation device 200 can be implemented as a server or a cloud server, etc. When the image generation device 200 is implemented as a cloud server, the image generation method can be executed in the cloud. Several computing nodes (cloud servers) can be deployed in the cloud, each with computing, storage, and other processing resources. In the cloud, multiple computing nodes can be organized to provide a certain service; of course, a single computing node can also provide one or more services. The cloud can provide this service by providing a service interface, which users can call to use the corresponding service. Service interfaces include Software Development Kits (SDKs), Application Programming Interfaces (APIs), etc.

[0046] The image generation device 200 refers to a device capable of providing image generation operations in a network virtual environment, typically a device that utilizes a network for information planning and image generation. Physically, the image generation device 200 can be any device capable of providing computing services, responding to image generation requests, and performing image generation operations. This device can be integrated into pre-built equipment, such as a cluster server, a regular server, a cloud server, a cloud host, or a virtual data center. The image generation device 200 mainly consists of a processor, hard disk, memory, and system bus, similar to a general computer architecture.

[0047] The information acquisition device 100 can refer to an electronic device capable of performing information acquisition operations and obtaining arterial spin labeling (ASL) images and physiological parameters. Specifically, the information acquisition device 100 can be implemented as an MRI imaging device, a physiological parameter acquisition device, etc. Furthermore, the basic structure of the information acquisition device 100 may include at least one processor. The number of processors depends on the configuration and type of the information acquisition device 100. The information acquisition device 100 may also include a memory, which can be volatile, such as Random Access Memory (RAM), or non-volatile, such as Read-Only Memory (ROM), flash memory, etc., or both types may be included simultaneously. The memory typically stores an operating system (OS), one or more application programs, and may also store program data. In addition to the processing unit and memory, the information acquisition device 100 also includes some basic configurations, such as chips, I / O buses, display components, and some peripheral devices. Optionally, some peripheral devices may include, for example, a keyboard, mouse, input pen, printer, etc. Other peripheral devices are well known in the art and will not be described in detail here.

[0048] In this embodiment described above, the information acquisition device 100 and the image generation device 200 are connected via a network, which can be a wireless or wired network connection. If the information acquisition device 100 and the image generation device 200 are connected via a communication connection, the mobile network standard can be any one of 2G (GSM), 2.5G (GPRS), 3G (WCDMA, TD-SCDMA, CDMA2000, UTMS), 4G (LTE), 4G+ (LTE+), WiMax, 5G, 6G, etc.

[0049] The information acquisition device 100 is used to acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters. The information acquisition device 100 may include multiple sub-acquisition devices for acquiring different data. The baseline physiological parameters and the ASL image correspond to the same subject in a standard gravity environment, while the experimental physiological parameters correspond to the same subject in a microgravity experimental environment. The subject can be a human body, an animal, or a pre-defined experimental subject, etc. It is important to note that the information acquisition device 100 can perform information acquisition operations in real time or at preset time points. In this case, the acquired parameters are not only related to time but also to the type of environment at that time. For example, when the information acquisition device 100 performs information acquisition operations at a preset time point, if the subject is in a standard gravity environment at that time point, the information acquisition device 100 can acquire the baseline physiological parameters and the ASL image after acquiring information at the preset time point; if the subject is in a microgravity experimental environment at that preset time point, the information acquisition device 100 can acquire the experimental physiological parameters after acquiring information at the preset time point. After acquiring the ASL image, baseline physiological parameters, and experimental physiological parameters, in order to acquire the target ASL image in the microgravity experimental environment, the acquired ASL image, baseline physiological parameters, and experimental physiological parameters can be sent to the image generation device 200.

[0050] The image generation device 200 is communicatively connected to the information acquisition device 100. It is used to acquire arterial spin-labeled ASL images, baseline physiological parameters, and experimental physiological parameters. Then, it determines an image generation model, which is used to generate ASL images under microgravity experimental conditions. The image generation model is used to process the arterial spin-labeled ASL images, baseline physiological parameters, and experimental physiological parameters, thereby determining the target ASL image of the subject located in the microgravity experimental environment. Because there is a correlation between microgravity experimental environments and microgravity environments, the target ASL image obtained in the microgravity experimental environment can be regarded as the ASL image of the subject in the microgravity environment. In this way, ASL images in the microgravity environment can be obtained without deploying specific equipment in the microgravity environment, thus effectively ensuring the convenience and reliability of ASL image acquisition in the microgravity environment. Furthermore, the brain health status of the subject can be detected based on the obtained ASL images. At the same time, the image generation model used to generate the target ASL image can be flexibly optimized or updated, which helps to improve the quality and effect of the target ASL image generated by the image generation model, thereby effectively improving the accuracy of ASL image acquisition in the microgravity environment and further enhancing the practicality of the scheme.

[0051] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Where there is no conflict between the embodiments, the following embodiments and features can be combined with each other. Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0052] Figure 2 This is a flowchart illustrating an image generation method provided in an embodiment of the present invention; see attached diagram. Figure 2 As shown, this embodiment provides an image generation method. The execution subject of this method can be an image generation device 200, which can be implemented as software or a combination of software and hardware. Based on the image generation device 200 described above, the image generation method in this embodiment includes:

[0053] Step S201: Acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment.

[0054] Step S202: Determine the image generation model for generating ASL images in a microgravity experimental environment.

[0055] Step S203: Use an image generation model to process the ASL image, baseline physiological parameters, and experimental physiological parameters to determine the target ASL image of the subject located in the microgravity experimental environment.

[0056] The specific implementation process and effects of each of the above steps are explained in detail below:

[0057] Step S201: Acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment.

[0058] Arterial spin labeling (ASL) is a non-invasive, non-radioactive technique for measuring whole-brain perfusion in quantitative units (ml / 100g / min). Arterial spin labeling (ASL) images are generated using this technique. Baseline physiological parameters refer to the physiological parameters of a subject collected under standard gravity conditions. Specifically, these parameters may include at least one of the following: heart rate, blood pressure, blood oxygen saturation, body temperature, and respiratory rate. It is understood that different physiological parameters may be obtained for different applications. Standard gravity environment refers to the environment in which the subject is located under terrestrial conditions. Corresponding to baseline physiological parameters are experimental physiological parameters, which are the physiological parameters of a subject collected under microgravity experimental conditions. Microgravity experimental environment refers to simulating the microgravity state in space (usually 10⁻⁶ g / min) under terrestrial conditions. -2 g to 10 -6 The experimental environment is on the order of g (g). The subject refers to the object from which the image is generated in this embodiment. There can be one or more subjects. When there are multiple subjects, each subject corresponds to a different combination of ASL images, baseline physiological parameters, and experimental physiological parameters.

[0059] To facilitate understanding the generation process of the target ASL image, a brief description of the microgravity experiment is provided below: The microgravity experimental environment can be simulated using a head-down position bed rest experiment. In this environment, at least one subject participating in the experiment follows the same schedule, such as waking up at 6:30 AM and turning off the lights at 11:00 PM. During the bed rest period, strict bed rest is required, and all activities, including eating, using the toilet, and showering, are performed in a head-down position. The bed rest period can be, for example, 12 hours, 1 day, 3 days, or 7 days. The head-down angle can be between -2 degrees and -12 degrees, typically -6 degrees is chosen for subsequent experiments. For at least one subject participating in the experiment, the bed rest posture can be supine, lateral, or prone, but at least one shoulder must always be in contact with the bed. Before bed rest, an MRI scanner can be used to collect ASL data of at least one subject's head, and a multi-parameter monitor can be used to collect physiological parameters of at least one subject to obtain baseline physiological parameters and ASL images of the subject in a standard gravity environment. Similarly, after bed rest ends, a multi-parameter monitor can be used to collect physiological parameters of at least one subject to obtain experimental physiological parameters of the subject in a microgravity experimental environment.

[0060] Since standard gravity and microgravity experimental environments are different, while baseline physiological parameters and ASL images correspond to those in the standard gravity environment, and experimental physiological parameters correspond to those in the microgravity environment, the baseline physiological parameters, ASL images, and experimental physiological parameters can be obtained by data acquisition operations at different times using preset ASL devices and physiological detection devices. These preset ASL devices and physiological detection devices can be integrated into the image generation device; that is, the image generation device in this embodiment has data acquisition capabilities. Therefore, acquiring arterial spin labeling ASL images, baseline physiological parameters, and experimental physiological parameters can include: acquiring ASL images of the subject under standard gravity conditions using the ASL device; detecting the subject under standard gravity conditions using the physiological detection device to obtain baseline physiological parameters of the same subject under standard gravity conditions; and detecting the subject under microgravity conditions using the physiological detection device to obtain experimental physiological parameters of the same subject. This effectively ensures the accuracy and reliability of acquiring ASL images, baseline physiological parameters, and experimental physiological parameters.

[0061] In other instances, ASL images, baseline physiological parameters, and experimental physiological parameters can be obtained not only through data acquisition operations based on ASL devices and physiological detection devices, but also through user transmission operations. In this case, acquiring arterial spin labeling ASL images, baseline physiological parameters, and experimental physiological parameters may include: displaying a data upload interface or data upload interface; the user transmitting ASL images, baseline physiological parameters, and experimental physiological parameters to the image generation device through the data upload interface or data upload interface. The ASL images, baseline physiological parameters, and experimental physiological parameters transmitted by the user belong to the same entity, and the ASL images and baseline physiological parameters are pre-acquired under standard gravity conditions, while the experimental physiological parameters are pre-acquired under microgravity experimental conditions. This also ensures the stability and reliability of acquiring ASL images, baseline physiological parameters, and experimental physiological parameters.

[0062] In other instances, ASL images, baseline physiological parameters, and experimental physiological parameters can be obtained not only through data acquisition operations based on ASL devices and physiological detection devices, but also through access operations to preset devices. In this case, acquiring arterial spin labeling ASL images, baseline physiological parameters, and experimental physiological parameters may include: determining a preset device that is communicatively connected to the image generation device, wherein the preset device stores ASL images, baseline physiological parameters, and experimental physiological parameters; and performing active or passive access operations to the preset device so that the image generation device can stably acquire ASL images, baseline physiological parameters, and experimental physiological parameters stored in the preset device.

[0063] It should be noted that, to ensure the effectiveness of the experimental physiological parameters obtained, multiple data collections can be performed at different times during the microgravity experimental environment. Based on these multiple data collections, multiple target ASL images of the same subject in the microgravity experimental environment can be generated, thereby improving the accuracy of the determined target ASL images. For example, the same subject may be in the microgravity experimental environment for different durations, such as 12 hours, 1 day, 3 days, and 7 days. Experimental physiological parameters can then be collected for each of these durations. Based on the experimental physiological parameters, ASL images, and baseline physiological parameters at different times, target ASL images corresponding to each experimental time period can be determined. This allows for the assessment of the brain health status of the same subject using the target ASL images corresponding to each experimental time period, thus improving the accuracy and reliability of brain health status assessment.

[0064] Step S202: Determine the image generation model for generating ASL images in a microgravity experimental environment.

[0065] The image generation model can be a pre-trained network model for generating ASL images in a microgravity experimental environment. In some instances, the image generation model can be implemented as a generative adversarial network model, for example, it can be implemented as a generative adversarial network constructed using the Swin-Transformer architecture. In this case, the image generation model can include: an input layer, an input encoding layer, a window attention layer, a feedforward layer, and a normalization layer.

[0066] In other instances, the image generation model is obtained by training a network model. To improve the inference accuracy of the image generation model, the network model can be trained first. Specifically, ASL image samples and physiological parameter samples of the same subject under standard gravity conditions, as well as ASL experimental image samples and experimental physiological parameter samples of the same subject under microgravity conditions, can be obtained first. Based on the ASL image samples, physiological parameter samples, experimental physiological parameter samples, and ASL experimental image samples, the model is trained to obtain the aforementioned image generation model. To improve the training quality and effectiveness of the image generation model to a certain extent, the model can be trained multiple times on the aforementioned data of multiple different subjects, or multiple times on the aforementioned data of the same subject at different times.

[0067] After training the network model multiple times, the training performance of different image generation models is compared (determined by inference accuracy and / or precision metrics). The best-performing image generation model is then selected as the model for generating ASL images in a microgravity experimental environment. "Best-performing" can mean that the image generation model trained on data from a specific subject produces the clearest ASL images in a microgravity experimental environment, or that the ASL images generated by the model trained on data from a specific subject have the highest similarity to experimental ASL images of that subject in a microgravity experimental environment.

[0068] Furthermore, during the training process of the image generation model, to further ensure the training quality, the 3D ASL image can be split into multiple 2D sagittal slice images. For example, the initial resolution of the ASL image is 61*73*61 (mm). 3 If so, it can be divided into 61 pieces of 64*64 (mm) each. 2 The ASL image is first divided into 61 two-dimensional slices along the depth direction, each original slice being 73×61 mm in size. 2 Then, by applying zero padding to the height and width dimensions, the size of each slice is made uniformly 64×64 (mm). 2 Then, the slices are normalized so that the pixels of each slice are in the range of [-1, 1]. The model can then be trained based on the processed slice images. This allows the image generation model to learn more image details from the slice images, which helps to ensure the accuracy of image generation.

[0069] Step S203: Use an image generation model to process the ASL image, baseline physiological parameters, and experimental physiological parameters to determine the target ASL image of the subject located in the microgravity experimental environment.

[0070] After obtaining the ASL image, baseline physiological parameters, and experimental physiological parameters, an image generation model can be used to process the ASL image, baseline physiological parameters, and experimental physiological parameters. Specifically, the ASL image, baseline physiological parameters, and experimental physiological parameters can be input into the image generation model for processing, thereby determining the target ASL image output by the image generation model, which is located in the microgravity experimental environment. This effectively achieves flexible and reliable generation of target ASL images in the microgravity experimental environment.

[0071] Furthermore, for ASL images, slight head movements during the scanning process, such as breathing or other head position movements, can cause motion artifacts in the imaging results. Therefore, in order to reduce artifacts caused by head movements and ensure the accuracy of the generated target ASL image, preprocessing operations can be performed on the ASL image before processing the ASL image, baseline physiological parameters, and experimental physiological parameters using the image generation model. Specifically, preprocessing of the ASL image can be performed using preset algorithms or application software. Processing methods include, but are not limited to, head movement correction, spatial smoothing, and skull dissection, which can obtain more accurate ASL images.

[0072] Specifically, after acquiring ASL images, rigid body registration can be performed using motion correction tools (e.g., mcflirt). For example, the Iterative Closest Point (ICP) algorithm can be used to perform head motion correction. This algorithm adjusts the image by comparing consecutive image frames to correct head movements during scanning, reducing motion artifacts and ensuring temporal consistency and spatial accuracy of the image sequence. Alternatively, the image generation model can use spatial filtering to smooth out noise that occurs during ASL image scanning. This operation can improve the signal-to-noise ratio (SNR), making perfusion signals more prominent and easier to analyze. Alternatively, the image generation model can also perform skull dissection to remove non-brain elements from the ASL images.

[0073] Furthermore, since the position, shape, and size of the head and brain may vary among each subject during the scan, these individual differences will result in different spatial positions and coordinates for the scanned ASL images. To facilitate data processing and analysis and eliminate the influence of inter-individual differences, a unified coordinate system operation can be performed on the ASL images to register each subject's ASL image to the same space. Specifically, after the subject's ASL image undergoes affine transformations, rigid body transformations, etc., it is aligned with a standard brain space template. This ensures that the same brain regions in the ASL images of different subjects are located at the same position in the same space. The standard brain template used in performing the above registration operation can be, for example, the MNI152 standard brain template.

[0074] Similarly, after obtaining the baseline physiological parameters and experimental physiological parameters, they can be preprocessed, such as converting their formats, generating corresponding encoding information based on them, or directly encoding them. This helps improve the accuracy and reliability of the image generation operation.

[0075] The image generation method provided in this embodiment acquires arterial spin marker (ASL) images, baseline physiological parameters, and experimental physiological parameters of the same subject. Then, it determines an image generation model for generating ASL images in a microgravity experimental environment. The image generation model is used to process the ASL images, baseline physiological parameters, and experimental physiological parameters to determine the target ASL image of the subject in the microgravity experimental environment. Since there is a correlation between microgravity experimental environments and microgravity environments, the obtained target ASL image in the microgravity experimental environment can be considered as the ASL image of the subject in the microgravity environment. This eliminates the need for specific equipment deployment in the microgravity environment to acquire ASL images, effectively ensuring the convenience and reliability of ASL image acquisition in microgravity. Furthermore, the obtained ASL images can be used to detect the brain health status of the subject, further improving the practicality of the method and facilitating its market promotion and application.

[0076] Figure 3 This is a schematic diagram illustrating the process of determining a target ASL image using an image generation model, provided in an embodiment of the present invention; based on the above embodiment, refer to the appendix. Figure 3 As shown, this embodiment provides a scheme for processing ASL images, baseline physiological parameters, and experimental physiological parameters using an image generation model to determine the target ASL image of the subject located in a microgravity experimental environment. Specifically, the method in this embodiment includes:

[0077] Step S301: Encode the baseline physiological parameters and experimental physiological parameters to obtain parameter characterization information.

[0078] Since the baseline physiological parameters and experimental physiological parameters are often implemented as text-type data, and ASL images are implemented as image-type data, that is, ASL images, baseline physiological parameters, and experimental physiological parameters are data of different modalities, in order to further ensure the accuracy and reliability of the image generation operation, after obtaining the baseline physiological parameters and experimental physiological parameters, parameter encoding operation can be performed to obtain parameter representation information corresponding to the baseline physiological parameters and experimental physiological parameters. This parameter representation information is used to represent the information contained in the baseline physiological parameters and experimental physiological parameters, and the parameter representation information can be implemented as vector representation information.

[0079] In some instances, after obtaining baseline and experimental physiological parameters, these parameters can be concatenated to obtain concatenated physiological parameters. Then, parameter encoding is performed on these concatenated parameters to obtain encoded parameter representation information. Alternatively, in other instances, after obtaining baseline and experimental physiological parameters, they can be separately encoded to obtain baseline and experimental parameter representations. Then, a concatenation operation is performed based on these representations to obtain encoded parameter representation information. This effectively ensures the accuracy and reliability of determining the parameter representation information.

[0080] Step S302: Input the parameter representation information and the ASL image into the image generation model for processing to obtain the target ASL image output by the image generation model.

[0081] After obtaining the parameter representation information, the encoded parameter representation information and the ASL image can be input into the image generation model to obtain the target ASL image output by the image generation model. Specifically, since the image generation model can be implemented as a generative adversarial network, then, as... Figure 4 As shown, the image generation model may include: an input layer 41, an input encoding layer 42, a window attention layer 43, a normalization layer 44, a feedforward layer 45, and a normalization layer 46. Therefore, the process by which the image generation model processes the ASL image, baseline physiological parameters, and experimental physiological parameters specifically includes: first, the ASL image is segmented in the input layer 41 to obtain multiple ASL image patches and their corresponding positional encoding information; then, the input encoding layer 42 is used to encode the multiple ASL image patches to obtain graph representation information corresponding to each ASL image patch; and based on the positional encoding information, the graph representation information and the aforementioned parameter representation information are stitched together to obtain stitched representation information; finally, the window attention layer 43, normalization layer 44, feedforward layer 45, and normalization layer 46 are used sequentially to process the stitched representation information to obtain the target ASL image.

[0082] In this embodiment, parameter encoding is performed on two physiological parameters to obtain parameter representation information, thereby enabling the image generation model to process the parameter representation information and ASL image. This effectively ensures that the input data in the image generation model is in a suitable format, which in turn helps to improve the processing quality and efficiency of the image generation model.

[0083] Figure 5 This is a schematic diagram of an image generation model optimization process provided by an embodiment of the present invention. Based on the above embodiment, refer to the appendix. Figure 5As shown, for the image generation model, in order to improve and ensure the effect and quality of the image generation operation, after determining the target ASL image of the subject located in the microgravity experimental environment, the image generation model can be optimized. Specifically, the method in this embodiment may include:

[0084] Step S501: Obtain the true ASL image corresponding to the experimental physiological parameters. The true ASL image corresponds to the subject located in the microgravity experimental environment.

[0085] The true ASL image refers to the ASL image corresponding to the experimental physiological parameters of the subject under microgravity experimental conditions. This true ASL image can be acquired by using an ASL device or related equipment to collect data from the subject. Alternatively, the true ASL image can be a pre-collected ASL image stored in a preset area. In this case, the true ASL image corresponding to the experimental physiological parameters can be obtained by accessing the preset area.

[0086] Step S502: Optimize the image generation model based on the ground truth ASL image and the target ASL image to obtain the optimized image generation model.

[0087] Since the target ASL image is generated based on an image generation model, and the ground truth ASL image is the actual ASL image acquired through an ASL device, comparing the similarity between the ground truth ASL image and the target ASL image can determine whether the image generation model needs optimization. In some instances, optimizing the image generation model based on the ground truth ASL image and the target ASL image to obtain an optimized image generation model may include: determining the image similarity between the ground truth ASL image and the target ASL image; and optimizing the image generation model based on the ground truth ASL image when the image similarity is less than or equal to a preset threshold, thus obtaining the optimized image generation model.

[0088] Specifically, when the image generation model is implemented as a generative adversarial network (GAN), it can include a generator and a discriminator. The generator generates the target ASL image, and the discriminator determines the image similarity between the ground truth ASL image and the target ASL image. After determining the image similarity between the ground truth ASL image and the target ASL image, the image similarity is compared with a pre-set threshold. If the image similarity is greater than the threshold, it indicates that the image similarity between the ground truth ASL image and the target ASL image is high, meaning that the target ASL image generated by the image generation model meets the user's requirements. In this case, no optimization or update of the image generation model is needed.

[0089] Correspondingly, when the image similarity is less than or equal to a preset threshold, meaning the similarity between the ground truth ASL image and the target ASL image is not high, it indicates that the target ASL image generated by the image generation model is not performing well. In this case, it is necessary to optimize the image generation model based on the ground truth ASL image to obtain an optimized image generation model. Since the image generation model includes a generator and a discriminator, the optimization operation can be implemented in the following three ways: 1) Optimize the hyperparameters of the generator in the image generation model based on the ground truth ASL image to obtain an optimized image generation model; 2) Optimize the hyperparameters of the discriminator in the image generation model based on the ground truth ASL image to obtain an optimized image generation model; 3) Optimize the hyperparameters of both the generator and the discriminator in the image generation model based on the ground truth ASL image to obtain an optimized image generation model. This achieves the optimization and update operation of the image generation model to obtain an image generation model with better performance in generating target ASL images, i.e., an optimized image generation model.

[0090] In this embodiment, by acquiring the ground truth ASL image and optimizing the image generation model based on the ground truth ASL image and the target ASL image, the optimization operation of the image generation model is effectively realized, and the image generation effect of the image generation model is further improved.

[0091] Figure 6 This is a schematic diagram illustrating another image generation model optimization process provided by an embodiment of the present invention. Based on the above embodiments, refer to the appendix. Figure 6 As shown, for image generation models, the optimization and update operations can be implemented not only directly based on the similarity between the ground truth ASL image and the target ASL image, but also based on the constructed edge loss function. The edge loss function is determined based on both the ground truth ASL image and the target ASL image. To improve and ensure the image generation effect and quality of the image generation model, this embodiment optimizes the image generation model based on the ground truth ASL image and the target ASL image to obtain the optimized image generation model, which may include:

[0092] Step S601: Obtain the ground truth edge map corresponding to the ground truth ASL image and the target edge map corresponding to the target ASL image.

[0093] The ground truth edge map refers to the image obtained by edge extraction of the ground truth ASL image. The edges mentioned above are located between different attribute regions in the ground truth ASL image. Similarly, the target edge map refers to the image obtained by edge extraction of the target ASL image. In this embodiment, it can be understood as the brain tissue boundary. The brain tissue boundary can be, for example, the gray-white matter boundary or the cerebrospinal fluid region.

[0094] In some instances, the ground truth edge map and the target edge map can be obtained through a pre-trained edge extraction model. In this case, obtaining the ground truth edge map corresponding to the ground truth ASL image and the target edge map corresponding to the target ASL image can include: determining the pre-trained edge extraction model; inputting the ground truth ASL image and the target ASL image into the edge extraction model to perform edge extraction operations, thereby obtaining the ground truth edge map and the target edge map. This effectively ensures the accuracy and reliability of obtaining the ground truth edge map and the target edge map.

[0095] In other instances, ground truth edge maps and target edge maps can be obtained not only through pre-trained edge extraction models, but also through edge detectors that communicate with image generation models, as shown in the appendix. Figure 7 As shown, at this time, obtaining the ground truth edge map corresponding to the ground truth ASL image and the target edge map corresponding to the target ASL image may include: determining an edge detector for extracting the edges of the ASL image; inputting the target ASL image and the ground truth ASL image into the edge detector for edge extraction operation to obtain the target edge map and the ground truth edge map, which effectively ensures the accuracy and reliability of obtaining the ground truth edge map and the target edge map.

[0096] Step S602: Determine the edge loss function based on the ground truth edge map and the target edge map.

[0097] Since the ground truth edge map corresponds to the ground truth ASL image and the target edge map corresponds to the target ASL image, after obtaining the ground truth edge map and the target edge map, the ground truth edge map and the target edge map can be analyzed and processed to determine the edge loss function corresponding to the ground truth ASL image and the target ASL image. This edge loss function can represent the degree of similarity between the ground truth edge map and the target edge map.

[0098] In some instances, since the target ASL image is generated based on the generator in the image generation model, to ensure that the edge map of the target ASL image generated by the image generation model is as similar as possible to or matches the edge map of the ground truth ASL image, an edge loss function corresponding to the generator can be constructed based on the ground truth edge map and the target edge map. This allows edge information to constrain the generation result of the image generation model. Specifically, the edge loss function corresponding to the generator in the image generation model can be implemented as follows:

[0099]

[0100] Where G represents the generator, Let H represent the edge loss function corresponding to the generator, H represent the height of the ground truth ASL image and the target ASL image, W represent the width of the ground truth ASL image and the target ASL image, and Y represent the ground truth ASL image. Let S(Y) represent the target ASL image, and S(Y) represent the ground truth edge map. Let (i,j) represent the target edge map, and let (i,j) represent the position coordinates of pixels in the ground truth ASL image and the target ASL image.

[0101] Step S603: Optimize the image generation model using the edge loss function to obtain the optimized image generation model.

[0102] After constructing the edge loss function, the hyperparameters of the image generation model can be optimized and updated using the edge loss function, thereby realizing the optimization and update operation of the image generation model. Specifically, since the edge loss function corresponds to the generator in the image generation model, after constructing the edge loss function, the hyperparameters of the generator in the image generation model can be optimized and updated using the edge loss function, thereby obtaining an optimized image generation model.

[0103] In other instances, to improve the image generation quality of an image generation model, the scheme of optimizing the image generation model using an edge loss function to obtain an optimized image generation model may also include: determining a first loss function corresponding to the discriminator and a second loss function corresponding to the generator; determining a model loss function based on the first loss function, the second loss function, and the edge loss function; and optimizing the hyperparameters of the generator and the discriminator using the model loss function to obtain an optimized image generation model.

[0104] The first loss function, corresponding to the discriminator in the image generation model, can be constructed as follows:

[0105]

[0106] The second loss function corresponding to the generator in the image generation model can be constructed as follows:

[0107]

[0108] Where D represents the discriminator, Let X represent the first loss function corresponding to the discriminator, X represent the ASL image of the subject under standard gravity conditions, and Y represent the ground truth ASL image. Represents the target ASL image, E X,YLet represent the expected value of the ASL image and the ground truth ASL image of the subject under standard gravity conditions, and let D(X,Y) represent the output of the discriminator network when the inputs are the ASL image and the ground truth ASL image. E represents the second loss function corresponding to the generator. X This represents the expected value of an ASL image of a subject under standard gravity conditions. This represents the output of the discriminator network when the inputs are the true ASL image and the target ASL image.

[0109] After constructing the first loss function, the second loss function, and the marginal loss function, the model loss function can be determined based on these functions. In some instances, the sum of the first loss function, the second loss function, and the marginal loss function can be directly used as the model loss function. Alternatively, the model loss function can be determined based on the weight parameters corresponding to the marginal loss function. In this case, determining the model loss function based on the first loss function, the second loss function, and the marginal loss function can include: determining the weight parameters corresponding to the marginal loss function, where the weight parameters are trainable hyperparameters; determining the product value between the weight parameters and the marginal loss function; and using the first loss function, the second loss function, and the sum of their product values ​​as the model loss function.

[0110] Specifically, after obtaining the edge loss function, the corresponding weight parameters can be determined based on it. Then, the edge loss function is multiplied by its corresponding weight parameters, and the product is added to the first and second loss functions to obtain the final model loss function. Taking the edge loss function corresponding to the generator, the first loss function corresponding to the discriminator, and the second loss function corresponding to the generator as an example, the resulting model loss function can be expressed as:

[0111]

[0112] Where G represents the generator and D represents the discriminator. This represents the model loss function corresponding to the image generation model. This represents the first loss function corresponding to the discriminator. This represents the second loss function corresponding to the generator, where α represents the weight parameters corresponding to the edge loss function. This represents the edge loss function corresponding to the generator.

[0113] After determining the model loss function, the hyperparameters of the generator and / or the discriminator can be optimized and updated using the model loss function, thereby realizing the optimization and update operation of the image generation model and obtaining the optimized image generation model.

[0114] In this embodiment, a model loss function is obtained by constructing a first loss function, a second loss function, and an edge loss function. The image generation model is then optimized and trained based on the model loss function, which effectively realizes the optimization operation of the image generation model and further improves the image generation effect of the image generation model.

[0115] Figure 8 This is a flowchart illustrating the process of determining the brain health status of various brain regions after determining the target ASL image of the subject in a microgravity experimental environment, as provided in this embodiment of the invention. Based on the above embodiments, refer to the appendix. Figure 8 As shown, after determining the target ASL image of the subject in a microgravity experimental environment, the obtained target ASL can be used to identify brain health status. In this embodiment, the method may include:

[0116] Step S801: Divide the target ASL image into brain regions to obtain at least one brain region ASL image corresponding to the target ASL image.

[0117] Brain regions refer to different areas of the brain divided according to their different functions, such as the frontal lobe, temporal lobe, parietal lobe, or occipital lobe. Brain perfusion is an important physiological indicator of cerebral neurovascular health and brain function. Brain perfusion information (i.e., cerebral blood flow (CBF)) reflects the amount of blood flow to the brain. Therefore, after determining the target ASL image of the subject in a microgravity experimental environment, brain regions can be divided into the target ASL image. Specifically, a preset brain anatomy template or image segmentation model can be used to segment the target ASL image. The brain anatomy template can be an Automated Anatomical Labeling Atlas (AAL), a Harvard-Oxford Atlas, etc., thereby obtaining brain region ASL images corresponding to the target ASL image.

[0118] Generally, the brain region ASL image is at least a part of the target ASL image, and the number of brain region ASL images can be one or more. When there is only one brain region ASL image corresponding to the target ASL image, the brain perfusion information corresponding to that brain region is determined. When there are multiple brain region ASL images corresponding to the target ASL image, the brain perfusion information corresponding to each of the multiple brain regions indicated by the multiple brain region ASL images is determined separately. Therefore, based on the obtained brain perfusion information of different brain regions, the brain health status of each brain region can be determined.

[0119] Step S802: Based on ASL images of at least one brain region, determine the brain perfusion information corresponding to each of the at least one brain region in the subject.

[0120] After obtaining ASL images of at least one brain region, the brain region ASL images can be directly analyzed to determine the brain perfusion information corresponding to each of the at least one brain region in the subject. In a specific implementation scenario, the brain perfusion information can be implemented as cerebral blood flow.

[0121] Step S803: Determine the brain health status of each brain region based on brain perfusion information.

[0122] After obtaining brain perfusion information, this information can be analyzed and processed to determine the brain health status of each brain region. This health status can include healthy, unhealthy, and so on. In some cases, brain health status can be determined using a brain health detection model. In this case, determining the brain health status of each brain region based on brain perfusion information can include: identifying a pre-trained brain health detection model; inputting the brain perfusion information into the model to perform brain health status detection operations; and obtaining the brain health status of the brain region corresponding to the brain perfusion information. This effectively ensures the accuracy and reliability of determining the brain health status.

[0123] In other instances, brain health status can be determined not only based on brain health detection models but also based on the standard brain perfusion range corresponding to each brain region. In this case, determining the brain health status of each brain region based on brain perfusion information can include: determining the standard brain perfusion range corresponding to at least one brain region in the subject, where the standard brain perfusion range is used to identify the brain perfusion range in which the subject's brain region is in a healthy state; if the brain perfusion information is within the standard brain perfusion range, the corresponding brain region is determined to be in a healthy state; if the brain perfusion information is outside the standard brain perfusion range, the corresponding brain region is determined to be in a non-healthy state.

[0124] The standard brain perfusion range can be a range used to identify the healthy state of a specific brain region in a given subject; that is, different subjects may have different standard brain perfusion ranges. For example, different subjects may have different daily physical indicators, daily physiological parameters, and past medical history, thus corresponding to different standard brain perfusion ranges. Specifically, for the subject in this example, when determining the standard brain perfusion ranges corresponding to different brain regions of the subject, the standard brain perfusion ranges corresponding to different brain regions of the subject can be obtained in a targeted manner based on the subject's daily physical indicators (such as age, height, or weight), daily physiological parameters (such as heart rate, blood pressure, blood oxygen, body temperature, or respiratory rate), and past medical history (such as whether there is a history of brain disease or family history of brain disease).

[0125] In some instances, it is also possible to find other subjects with the same or similar daily physical indicators, daily physiological parameters, and past medical history as the subject in this instance. The brain perfusion range of different brain regions of different subjects, including the subject, can be determined. For each brain region, the average brain perfusion range of each brain region can be obtained based on the brain perfusion range of each subject. The average brain perfusion range of different brain regions can be used as the standard brain perfusion range of different brain regions of the subject.

[0126] For the subject, if the brain perfusion information corresponding to at least one brain region of the subject is within the standard brain perfusion range of each brain region, the brain health status of the corresponding brain region is determined to be healthy; conversely, if the brain perfusion information corresponding to at least one brain region of the subject is outside the standard brain perfusion range of each brain region, the brain health status of the corresponding brain region is determined to be unhealthy.

[0127] In this embodiment, the brain health status can be determined solely by the brain perfusion information corresponding to the target ASL image, thus improving the efficiency of determining brain health status.

[0128] Figure 9 This invention provides another flowchart illustrating the process of determining the brain health status of various brain regions after determining the target ASL image of the subject in a microgravity experimental environment, based on the above embodiments and with reference to the appendix. Figure 9 As shown, after determining the target ASL image of the subject in a microgravity experimental environment, brain health status can be identified not only directly based on the target ASL image, but also by combining the ASL image with the brain health status identification. In this case, the method in this embodiment may include:

[0129] Step S901: Based on the ASL image and the target ASL image, determine the brain perfusion change information of each brain region in the subject.

[0130] Among them, brain perfusion change information represents the changes in brain perfusion information of each brain region in the ASL image and the changes in brain perfusion information of the corresponding brain region in the target ASL image.

[0131] In some instances, the brain perfusion change information of a subject can be determined by direct comparison. In this case, the brain perfusion information of each brain region corresponding to the ASL image and the brain perfusion information of each brain region corresponding to the target ASL image can be determined separately. Then, the brain perfusion information of the ASL images corresponding to the same brain regions of the same subject and the brain perfusion information of the target ASL image can be compared to obtain the brain perfusion change information of each brain region of the subject.

[0132] In other instances, determining brain perfusion change information for each brain region in the subject based on the ASL image and the target ASL image may also include: stitching the ASL image and the target ASL image together to obtain a stitched ASL image; and determining brain perfusion change information for each brain region in the subject based on the stitched ASL image, wherein the brain perfusion change information includes at least one of the following: brain perfusion change magnitude and brain perfusion change rate.

[0133] The process of stitching together the ASL image and the target ASL image to obtain the stitched ASL image may include: dividing the ASL image and the target ASL image into brain regions respectively to obtain at least one reference brain region ASL image corresponding to the ASL image and at least one target brain region ASL image corresponding to the target ASL image; and stitching together the at least one reference brain region ASL image and the at least one target brain region ASL image to obtain the stitched ASL image.

[0134] Specifically, after determining the target ASL image of the subject in the microgravity experimental environment, brain regions are divided into both the target and reference ASL images to obtain a baseline ASL image corresponding to the target ASL image, and a target ASL image corresponding to the target ASL image. The number of reference and target ASL images can be one or multiple. If only one baseline and target ASL image exists, they are stitched together to obtain a stitched ASL image. If multiple baseline and target ASL images exist, stitching is performed separately on the corresponding baseline and target ASL images for the same brain region to obtain stitched ASL images for each brain region.

[0135] In some other instances, brain perfusion change information of the subject can be determined based on a brain perfusion determination model. In this case, based on stitched ASL images, the brain perfusion change information of each brain region in the subject is determined. The brain perfusion change information includes at least one of the following: brain perfusion change amplitude and brain perfusion change rate. This can be achieved by: determining a brain perfusion determination model; inputting the stitched ASL images into the brain perfusion determination model for processing, and obtaining the brain perfusion change information of each brain region output by the model. The brain perfusion change information includes at least one of the following: brain perfusion change amplitude and brain perfusion change rate.

[0136] Step S902: Based on the brain perfusion change information, determine the brain health status of each brain region.

[0137] After obtaining information on changes in brain perfusion, this information can be analyzed and processed to determine the brain health status of each brain region. In some instances, determining the brain health status of each brain region based on brain perfusion change information may include: determining the brain perfusion change threshold corresponding to each brain region for analyzing and processing the brain perfusion change information; determining the brain health status of the corresponding brain region as healthy when the brain perfusion change information is less than or equal to the brain perfusion change threshold; and determining the brain health status of the corresponding brain region as unhealthy when the brain perfusion change information is greater than the brain perfusion change threshold.

[0138] The brain perfusion change threshold is a numerical value used to identify whether a certain brain region of the subject is in a healthy state. Different brain regions of the same subject may correspond to different brain perfusion change thresholds.

[0139] Specifically, for the subject in this example, when determining the brain perfusion change thresholds for different brain regions, the thresholds can be determined based on the functional roles of different brain regions and by referring to the subject's daily physical indicators, daily physiological parameters, and past medical history. Typically, the brain perfusion change threshold can be set to 20%, but depending on actual needs, it can be set to a value between 15% and 20%. For this subject, if the brain perfusion change information for each brain region is less than or equal to the brain perfusion change threshold, it indicates that the degree of brain perfusion change in that brain region is within the normal range, and the brain health status of the corresponding brain region can be determined as healthy. Conversely, if the brain perfusion change information for each brain region is greater than the brain perfusion change threshold, it indicates that the degree of brain perfusion change in that brain region is no longer within the normal range, and the brain health status of the corresponding brain region is determined as unhealthy.

[0140] In some instances, it is also possible to find other subjects with the same or similar daily physical indicators, daily physiological parameters, and past medical history as the subject in this instance. Brain perfusion change information of different brain regions of different subjects, including the subject, can be determined. For each brain region, the average brain perfusion change value of each brain region can be obtained based on the brain perfusion change information of each subject. The average brain perfusion change value of different brain regions can be used as the brain perfusion change threshold of different brain regions of the subject.

[0141] In some other instances, brain health status can be determined using a brain health detection model. In this case, determining the brain health status of each brain region based on brain perfusion change information can include: determining a pre-trained brain health detection model; inputting brain perfusion change information into the brain health detection model to perform brain health status detection operations, and obtaining the brain health status of the brain region corresponding to the brain perfusion change information. This effectively ensures the accuracy and reliability of determining brain health status.

[0142] In some instances, after determining the brain health status of each brain region, the method further includes: generating health guidance information corresponding to the brain health status; and / or, generating recommendations on whether the subject is suitable for performing a space mission.

[0143] Among them, health guidance information refers to guidance information corresponding to different brain regions of the subject, and there is a corresponding relationship between health guidance information and the brain health status of each brain region.

[0144] Specifically, for brain regions in a healthy state, health guidance information can include preventative measures or training methods. Preventative measures could include: ensuring a scientifically balanced diet, maintaining regular sleep patterns, keeping a normal weight, quitting smoking and alcohol, and preventing diseases such as hypertension, diabetes, and hyperlipidemia. Training methods could include: cognitive, motor, language, or other training, strengthening physical exercise, and training to adapt to microgravity experimental environments. For brain regions in an unhealthy state, health guidance information can include improvement measures or treatment measures. For example, improvement measures or treatment measures could include: taking appropriate medications and using medical devices to improve or treat the corresponding brain regions.

[0145] In some instances, after determining the brain health status of each brain region of a subject, a comprehensive assessment can be conducted based on this status to determine whether the subject is suitable for a space mission. Specifically, a threshold for normal brain regions can be pre-set based on the total number of brain regions in the subject. If the number of healthy brain regions in the subject is greater than or equal to the normal brain region threshold, a recommendation that the subject is suitable for a space mission is generated. Conversely, if the number of healthy brain regions in the subject is less than the normal brain region threshold, a recommendation that the subject is unsuitable for a space mission is generated.

[0146] In another example, weights can be generated for the brain regions segmented from the subject. The different weights of each brain region indicate their varying importance in a microgravity experimental environment. A positive correlation exists between the weight of a brain region and its corresponding importance; that is, the higher the weight of a brain region, the more important it is in a microgravity experimental environment. In this case, a normal score threshold can be pre-set. Then, based on the weights of each brain region and the corresponding brain health status, a brain region score is generated for each region. This score is then used to generate a brain fraction for the subject, which is compared to the pre-set normal score threshold. If the subject's brain fraction is greater than or equal to the normal score threshold, a recommendation is generated that the subject is suitable for a space mission. Conversely, if the subject's brain fraction is less than the normal score threshold, a recommendation is generated that the subject is unsuitable for a space mission.

[0147] In this embodiment, brain perfusion change information is determined by ASL images and target ASL images. Then, brain health status can be determined based on the brain perfusion change information, which improves the accuracy of determining the subject's brain health status. This effectively improves the accuracy of generating health guidance information and / or suggestion information for the subject, further enhancing the practicality of the method and facilitating its promotion and application in the market.

[0148] Figure 10 This is a schematic diagram of the structure of an image generation device provided in an embodiment of the present invention; see attached diagram. Figure 10 As shown, this embodiment provides an image generation apparatus, which may include:

[0149] The first acquisition module 11 is used to acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment.

[0150] The first determining module 12 is used to determine an image generation model for generating ASL images in a microgravity experimental environment;

[0151] The first processing module 13 is used to process the ASL image, baseline physiological parameters and experimental physiological parameters using an image generation model to determine the target ASL image of the subject located in a microgravity experimental environment.

[0152] In some instances, after acquiring the arterial spin-labeled ASL image, baseline physiological parameters, and experimental physiological parameters, and determining the image generation model, the first processing module 13 is further configured to: encode the baseline physiological parameters and experimental physiological parameters to obtain parameter characterization information; input the parameter characterization information and the ASL image into the image generation model for processing to obtain the target ASL image output by the image generation model.

[0153] In some instances, after determining the target ASL image, the first acquisition module 11 and the first determination module 12 in this embodiment are used to perform the following steps:

[0154] The first acquisition module 11 is used to acquire the true ASL image corresponding to the experimental physiological parameters. The true ASL image corresponds to the subject located in the microgravity experimental environment.

[0155] The first determining module 12 is used to optimize the image generation model based on the ground ASL image and the target ASL image to obtain the optimized image generation model.

[0156] In some instances, after determining the target ASL image, the first determining module 12 in this embodiment performs the following steps: determining the image similarity between the ground truth ASL image and the target ASL image; if the image similarity is less than or equal to a preset threshold, optimizing the image generation model based on the ground truth ASL image to obtain the optimized image generation model.

[0157] In some instances, after determining the target ASL image, the first acquisition module 11 and the first determination module 12 in this embodiment are used to perform the following steps:

[0158] The first acquisition module 11 is used to acquire the ground truth edge map corresponding to the ground truth ASL image and the target edge map corresponding to the target ASL image;

[0159] The first determining module 12 is used to determine the edge loss function based on the ground truth edge map and the target edge map; and to optimize the image generation model using the edge loss function to obtain the optimized image generation model.

[0160] In some instances, the image generation model includes a generator for generating a target ASL image and a discriminator for determining image similarity. After determining the target ASL image, the first determining module 12 in this embodiment performs the following steps: determining a first loss function corresponding to the discriminator and a second loss function corresponding to the generator; determining a model loss function based on the first loss function, the second loss function, and the edge loss function; and optimizing the hyperparameters of the generator and the discriminator using the model loss function to obtain an optimized image generation model.

[0161] In some instances, after determining the target ASL image, the first determining module 12 in this embodiment performs the following steps: determining the weight parameters corresponding to the edge loss function, wherein the weight parameters are trainable hyperparameters; determining the product value between the weight parameters and the edge loss function; and determining the first loss function, the second loss function, and the sum of the product values ​​as the model loss function.

[0162] In some instances, after determining the target ASL image, the first acquisition module 11 and the first determination module 12 in this embodiment are used to perform the following steps:

[0163] The first acquisition module 11 is used to divide the target ASL image into brain regions and obtain at least one brain region ASL image corresponding to the target ASL image.

[0164] The first determining module 12 is used to determine the brain perfusion information corresponding to at least one brain region in the subject based on ASL images of at least one brain region; and to determine the brain health status of each brain region based on the brain perfusion information.

[0165] In some instances, after determining the target ASL image, the first determining module 12 in this embodiment performs the following steps: determining the standard brain perfusion range corresponding to at least one brain region in the subject, wherein the standard brain perfusion range is used to identify the brain perfusion range in which the brain region of the subject is in a healthy state; if the brain perfusion information is within the standard brain perfusion range, determining the brain health state of the corresponding brain region as healthy; if the brain perfusion information is outside the standard brain perfusion range, determining the brain health state of the corresponding brain region as unhealthy.

[0166] In some instances, after determining the target ASL image, the first determining module 12 in this embodiment is used to perform the following steps: based on the ASL image and the target ASL image, determine the brain perfusion change information of each brain region in the subject; based on the brain perfusion change information, determine the brain health status of each brain region.

[0167] In some instances, after determining the target ASL image, the first acquisition module 11 and the first determination module 12 in this embodiment are used to perform the following steps:

[0168] The first acquisition module 11 is used to stitch the ASL image and the target ASL image together to obtain a stitched ASL image;

[0169] The first determining module 12 is used to determine brain perfusion change information of each brain region in the subject based on the stitched ASL image. The brain perfusion change information includes at least one of the following: brain perfusion change amplitude and brain perfusion change rate.

[0170] In some instances, after determining the target ASL image, the first acquisition module 11 in this embodiment is used to perform the following steps: dividing the ASL image and the target ASL image into brain regions respectively to obtain at least one reference brain region ASL image corresponding to the ASL image and at least one target brain region ASL image corresponding to the target ASL image; stitching together at least one reference brain region ASL image and at least one target brain region ASL image to obtain a stitched ASL image.

[0171] In some instances, after determining the target ASL image, the first determining module 12 in this embodiment performs the following steps: determining the brain perfusion change threshold corresponding to each brain region for analyzing and processing brain perfusion change information; determining the brain health status of the corresponding brain region as healthy when the brain perfusion change information is less than or equal to the brain perfusion change threshold; and determining the brain health status of the corresponding brain region as unhealthy when the brain perfusion change information is greater than the brain perfusion change threshold.

[0172] In some instances, after determining the target ASL image, the first processing module 13 in this embodiment is used to perform the following steps: generating health guidance information corresponding to brain health status; and / or generating suggestion information on whether the subject is suitable for performing a space mission.

[0173] Figure 10 The device shown can perform Figures 1-9 For the methods shown in the embodiments, the parts not described in detail in this embodiment can be referred to the following: Figures 1-7 The relevant descriptions of the illustrated embodiments are provided below. For the execution process and technical effects of this technical solution, please refer to [link / reference]. Figures 1-9 The descriptions in the illustrated embodiments will not be repeated here.

[0174] In one possible design, Figure 10 The structure of the image generation device shown can be implemented as an electronic device, which can be a controller, personal computer, server, or other similar devices. Figure 11 As shown, the electronic device may include a first processor 21 and a first memory 22. The first memory 22 is used to store data executed by the corresponding electronic device. Figures 1-9 In the illustrated embodiment, the program for training a speech recognition model includes a first processor 21 configured to execute a program stored in a first memory 22.

[0175] The program includes one or more computer instructions, wherein when executed by the first processor 21, the one or more computer instructions can perform the following steps: acquiring arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment; determining an image generation model for generating ASL images in a microgravity experimental environment; and processing the ASL images, baseline physiological parameters, and experimental physiological parameters using the image generation model to determine the target ASL image of the subject located in the microgravity experimental environment.

[0176] Furthermore, the first processor 21 is also used to perform the aforementioned Figures 1-9 All or part of the steps in the illustrated embodiments.

[0177] The structure of the electronic device may also include a first communication interface 23 for communication between the electronic device and other devices or communication networks.

[0178] In addition, embodiments of the present invention provide a computer storage medium for storing computer software instructions used by an electronic device, which includes instructions for executing the above-described... Figures 1-9 The procedure involved in the image generation method in the illustrated embodiment.

[0179] Furthermore, embodiments of the present invention provide a computer program product, comprising: a computer-readable storage medium storing computer instructions, which, when executed by one or more processors, cause one or more processors to perform the aforementioned... Figures 1-9 The steps in the image generation method shown in the embodiment of the method.

[0180] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in the embodiments of this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0181] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the objectives of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any inventive effort.

[0182] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of a necessary general-purpose hardware platform, or by a combination of hardware and software. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a computer product. The present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0183] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0184] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the functions specified in one or more boxes. In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory. Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0185] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0186] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An image generation method, characterized in that, include: Arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters were acquired, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment. Determine an image generation model for generating ASL images in the microgravity experimental environment; The ASL image, baseline physiological parameters, and experimental physiological parameters are processed using the image generation model to determine the target ASL image of the subject located in the microgravity experimental environment.

2. The method according to claim 1, characterized in that, The image generation model is used to process the ASL image, baseline physiological parameters, and experimental physiological parameters to determine the target ASL image of the subject in a microgravity experimental environment, including: The baseline physiological parameters and the experimental physiological parameters are encoded to obtain parameter characterization information; The parameter representation information and the ASL image are input into the image generation model for processing to obtain the target ASL image output by the image generation model.

3. The method according to claim 1, characterized in that, After determining the target ASL image of the subject located in a microgravity experimental environment, the method further includes: Obtain a true ASL image corresponding to the experimental physiological parameters, wherein the true ASL image corresponds to the subject located in the microgravity experimental environment; The image generation model is optimized based on the ground truth ASL image and the target ASL image to obtain an optimized image generation model.

4. The method according to claim 3, characterized in that, The image generation model is optimized based on the ground truth ASL image and the target ASL image to obtain an optimized image generation model, including: Determine the image similarity between the ground truth ASL image and the target ASL image; If the image similarity is less than or equal to a preset threshold, the image generation model is optimized based on the ground truth ASL image to obtain an optimized image generation model.

5. The method according to claim 3, characterized in that, The image generation model is optimized based on the ground truth ASL image and the target ASL image to obtain an optimized image generation model, including: Obtain the ground truth edge map corresponding to the ground truth ASL image and the target edge map corresponding to the target ASL image; Based on the ground truth edge map and the target edge map, determine the edge loss function; The image generation model is optimized using the edge loss function to obtain an optimized image generation model.

6. The method according to claim 5, characterized in that, The image generation model includes a generator for generating the target ASL image and a discriminator for determining the image similarity. The image generation model is optimized using the edge loss function to obtain an optimized image generation model, including: Determine a first loss function corresponding to the discriminator and a second loss function corresponding to the generator; The model loss function is determined based on the first loss function, the second loss function, and the edge loss function; The hyperparameters of the generator and the discriminator are optimized using the model loss function to obtain an optimized image generation model.

7. The method according to claim 6, characterized in that, The model loss function is determined based on the first loss function, the second loss function, and the edge loss function, including: Determine the weight parameters corresponding to the edge loss function, wherein the weight parameters are trainable hyperparameters; Determine the product value between the weight parameters and the edge loss function; The first loss function, the second loss function, and the sum of the product values ​​are determined as the model loss function.

8. The method according to any one of claims 1-7, characterized in that, After determining the target ASL image of the subject located in a microgravity experimental environment, the method further includes: The target ASL image is divided into brain regions to obtain at least one brain region ASL image corresponding to the target ASL image; Based on the ASL images of the at least one brain region, determine the brain perfusion information corresponding to each of the at least one brain region in the subject; Based on the brain perfusion information, the brain health status of each brain region is determined.

9. The method according to claim 8, characterized in that, Based on the brain perfusion information, the brain health status of each brain region is determined, including: Determine the standard brain perfusion range corresponding to each of at least one brain region in the subject, the standard brain perfusion range being used to identify the brain perfusion range of the brain region of the subject in a healthy state; When the brain perfusion information is within the standard brain perfusion range, the brain health status of the corresponding brain region is determined to be healthy. If the brain perfusion information is outside the standard brain perfusion range, the brain health status of the corresponding brain region is determined to be unhealthy.

10. The method according to any one of claims 1-7, characterized in that, After determining the target ASL image of the subject located in a microgravity experimental environment, the method further includes: Based on the ASL image and the target ASL image, brain perfusion change information of each brain region in the subject is determined; Based on the brain perfusion change information, the brain health status of each brain region is determined.

11. The method according to claim 9, characterized in that, Based on the ASL image and the target ASL image, brain perfusion change information for each brain region in the subject is determined, including: The ASL image and the target ASL image are stitched together to obtain a stitched ASL image; Based on the stitched ASL image, brain perfusion change information for each brain region in the subject is determined, and the brain perfusion change information includes at least one of the following: brain perfusion change magnitude and brain perfusion change rate.

12. The method according to claim 11, characterized in that, The ASL image and the target ASL image are stitched together to obtain a stitched ASL image, including: Brain regions are divided into the ASL image and the target ASL image respectively to obtain at least one reference brain region ASL image corresponding to the ASL image and at least one target brain region ASL image corresponding to the target ASL image; The at least one reference brain region ASL image and the at least one target brain region ASL image are stitched together to obtain the stitched ASL image.

13. The method according to claim 10, characterized in that, Based on the brain perfusion change information, the brain health status of each brain region is determined, including: Determine the brain perfusion change thresholds corresponding to each brain region for analyzing and processing the brain perfusion change information; If the brain perfusion change information is less than or equal to the brain perfusion change threshold, the brain health status of the corresponding brain region is determined to be healthy. If the brain perfusion change information is greater than the brain perfusion change threshold, the brain health status of the corresponding brain region is determined to be unhealthy.

14. The method according to claim 13, characterized in that, After determining the brain health status of each brain region, the method further includes: Generate health guidance information corresponding to the brain health state; and / or, Generate recommendation information on whether the subject is suitable for performing space missions.

15. An image generation apparatus, characterized in that, include: The first acquisition module is used to acquire arterial spin labeling (ASL) images, baseline physiological parameters, and experimental physiological parameters, wherein the baseline physiological parameters and the ASL images correspond to the same subject located in a standard gravity environment, and the experimental physiological parameters correspond to the subject located in a microgravity experimental environment. The first determining module is used to determine the image generation model for generating ASL images in a microgravity experimental environment; The first processing module is used to process the ASL image, baseline physiological parameters, and experimental physiological parameters using the image generation model to determine the target ASL image of the subject located in a microgravity experimental environment.

16. An electronic device, characterized in that, include: A memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions, when executed by the processor, implement the method of any one of claims 1-14.

17. A computer storage medium, characterized in that, Used to store a computer program that, when executed by a computer, implements the method of any one of claims 1-14.

18. A computer program product, characterized in that, include: A computer program, when executed by a processor of an electronic device, causes the processor to perform the steps of the method of any one of claims 1-14.

Images