Method and apparatus for determining arterial input functions for perfusion imaging

Abstract

The application relates to a method and device for determining an arterial input function for perfusion imaging, a computer device and a storage medium. The method comprises: acquiring three-dimensional image data corresponding to four-dimensional perfusion data scanned by a medical imaging device for a target organ; detecting whether a target arterial blood vessel exists in a slice image of the three-dimensional image data in a target direction, wherein the target direction is a transverse cross-sectional axial direction of a scanning channel of the medical imaging device; if the target arterial blood vessel exists in the slice image of the three-dimensional image data in the target direction, segmenting a target arterial blood vessel image from the slice image; and determining an arterial input function required by the target organ during perfusion parameter calculation according to a first pixel cluster obtained by clustering pixels in the target arterial blood vessel image. The method can improve the accuracy of the arterial input function for perfusion imaging.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method, apparatus, computer device, and storage medium for determining an arterial input function for perfusion imaging. Background Technology

[0002] With the development of medical imaging technology, perfusion imaging technology has emerged. Perfusion imaging can be used to quantitatively analyze tissue hemodynamics and capillary permeability, which is of great significance for clarifying the blood supply to lesions. Determining the arterial input function is a crucial step in perfusion imaging; the accuracy of the arterial input function directly determines the accuracy of perfusion parameter calculations.

[0003] In traditional techniques, the method for determining the arterial input function is to first calculate the time-varying curve of the gray value of each voxel in the three-dimensional image data, and then select the most suitable arterial input function point through clustering, thereby determining the arterial input function based on the arterial input function point.

[0004] However, traditional techniques select arterial input function points from the entire 3D image data to determine the arterial input function, which involves a lot of interference data and results in low accuracy of the arterial input function. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, computer device, and storage medium for determining the arterial input function for perfusion imaging, which can improve the accuracy of the arterial input function used for perfusion imaging, in order to address the above-mentioned technical problems.

[0006] A method for determining an arterial input function for perfusion imaging, the method comprising:

[0007] Acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by medical imaging equipment;

[0008] Detect whether a target artery exists in the slice image of the three-dimensional image data in the target direction, wherein the target direction is the axial direction of the cross section of the scanning channel of the medical imaging device;

[0009] If the target artery exists in the slice image of the three-dimensional image data in the target direction, then the target artery image is segmented from the slice image;

[0010] Based on the first pixel cluster obtained by pixel clustering in the target arterial image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0011] In one embodiment, determining the arterial input function required for perfusion parameter calculation of the target organ based on a first pixel cluster obtained by pixel clustering in the target arterial image includes:

[0012] Based on the perfusion parameters associated with pixels in the target artery image, a first perfusion characteristic value corresponding to the pixels in the target artery image is determined;

[0013] Based on the first perfusion characteristic value, the first pixel cluster is obtained by clustering the pixels in the target artery image, and the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0014] In one embodiment, determining the arterial input function required for perfusion parameter calculation of the target organ by clustering pixels in the target arterial image based on the first perfusion characteristic value includes:

[0015] Based on the first perfusion characteristic value, the pixels in the target arterial image are clustered to obtain multiple first pixel clusters;

[0016] Select the first pixel cluster that satisfies the first preset pixel cluster condition from the plurality of first pixel clusters as the first artery input function point;

[0017] Based on the first arterial input function point, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0018] In one embodiment, selecting a first pixel cluster that satisfies a first preset pixel cluster condition from the plurality of first pixel clusters as the first artery input function point includes:

[0019] Based on the first infusion characteristic value, calculate the sum of the first infusion characteristic values ​​of the pixels contained in each of the plurality of first pixel clusters;

[0020] The average value of the first injection characteristic corresponding to each first pixel cluster is calculated based on the sum of the first injection characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster;

[0021] The first pixel cluster with the highest average value of the first perfusion characteristics is selected from the plurality of first pixel clusters as the first arterial input function point.

[0022] In one embodiment, the target organ is the brain; the method further includes:

[0023] If the target artery is not present in the slice image of the three-dimensional image data in the target direction, then the brain blood vessel image is segmented from the slice image;

[0024] Select the anterior half of the brain blood vessels from the aforementioned brain blood vessel images;

[0025] Based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0026] In one embodiment, determining the arterial input function required for perfusion parameter calculation of the target organ based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image includes:

[0027] The second perfusion characteristic value corresponding to the pixel in the anterior cerebral vascular image is calculated based on the perfusion parameters associated with the pixel in the anterior cerebral vascular image.

[0028] Based on the second perfusion characteristic value, the second pixel cluster is obtained by clustering the pixels in the anterior cerebral vascular image, and the arterial input function required for the target organ to calculate the perfusion parameters is determined.

[0029] In one embodiment, acquiring the three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by the medical imaging device includes:

[0030] The average value of the four-dimensional perfusion data of the target organ scan by medical imaging equipment is calculated in the scanning time dimension to obtain three-dimensional image data.

[0031] An apparatus for determining an arterial input function for perfusion imaging, the apparatus comprising:

[0032] The image acquisition module is used to acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by the medical imaging equipment.

[0033] The blood vessel detection module is used to detect whether a target artery exists in the slice image of the three-dimensional image data in the target direction, wherein the target direction is the axial direction of the cross section of the scanning channel of the medical imaging device;

[0034] An image segmentation module is used to segment out a target artery image from the slice image if a target artery exists in the slice image in the target direction of the three-dimensional image data.

[0035] The function determination module is used to determine the arterial input function required by the target organ for calculating perfusion parameters based on the first pixel cluster obtained by pixel clustering in the target arterial image.

[0036] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0037] Acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by medical imaging equipment;

[0038] Detect whether a target artery exists in the slice image of the three-dimensional image data in the target direction, wherein the target direction is the axial direction of the cross section of the scanning channel of the medical imaging device;

[0039] If the target artery exists in the slice image of the three-dimensional image data in the target direction, then the target artery image is segmented from the slice image;

[0040] Based on the first pixel cluster obtained by pixel clustering in the target arterial image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0041] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0042] Acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by medical imaging equipment;

[0043] Detect whether a target artery exists in the slice image of the three-dimensional image data in the target direction, wherein the target direction is the axial direction of the cross section of the scanning channel of the medical imaging device;

[0044] If the target artery exists in the slice image of the three-dimensional image data in the target direction, then the target artery image is segmented from the slice image;

[0045] Based on the first pixel cluster obtained by pixel clustering in the target arterial image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0046] The aforementioned method, apparatus, computer device, and storage medium for determining the arterial input function for perfusion imaging first detect whether a target artery exists in a slice image of the three-dimensional image data corresponding to the four-dimensional perfusion data in the target direction. If it exists, the target artery image is segmented from the slice image, and then the arterial input function is determined based on the target artery image. This application obtains the target artery image through target detection and image segmentation, reducing interference information in the image. Therefore, determining the arterial input function using this target artery image helps improve the accuracy of the arterial input function. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating a method for determining an arterial input function for perfusion imaging in one embodiment;

[0048] Figure 2 This is a schematic diagram illustrating the detection of a target artery in a slice image of three-dimensional image data in the target direction, as shown in one embodiment.

[0049] Figure 3 This is a schematic diagram illustrating the segmentation of a target arterial blood vessel image from a sheet image in one embodiment;

[0050] Figure 4 This is a schematic diagram illustrating the determination of an artery input function based on pixels in a target artery image in one embodiment;

[0051] Figure 5 This is a flowchart illustrating the process of determining an arterial input function from an anterior cerebral vascular image in one embodiment.

[0052] Figure 6 This is a schematic diagram illustrating the segmentation of brain blood vessels from a sheet image in one embodiment;

[0053] Figure 7 This is a schematic diagram illustrating the determination of an arterial input function based on pixels in an anterior cerebral vascular image in one embodiment;

[0054] Figure 8 This is a schematic diagram illustrating how, in one embodiment, the average value of four-dimensional perfusion data from a medical imaging device scanning a target organ is calculated along the scanning time dimension to obtain three-dimensional image data.

[0055] Figure 9 This is a flowchart illustrating a method for determining an arterial input function for perfusion imaging in another embodiment;

[0056] Figure 10 This is a structural block diagram of a device for determining an arterial input function for perfusion imaging in one embodiment;

[0057] Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] In one embodiment, such as Figure 1As shown, a method for determining the arterial input function for perfusion imaging is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0060] Step S102: Obtain the three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by the medical imaging equipment.

[0061] Medical imaging equipment refers to various instruments that use different media as information carriers to reproduce the internal structures of a living organism as images. The image information corresponds to the actual structure of the organism in both space and time. Medical imaging equipment includes computed tomography (CT) and magnetic resonance imaging (MR), among others.

[0062] Alternatively, the target organ can be the brain, heart, lungs, etc.

[0063] Among them, four-dimensional perfusion data is obtained by medical imaging equipment through imaging the perfusion process of blood entering the tissues of the target organ.

[0064] Specifically, the terminal acquires three-dimensional image data corresponding to the four-dimensional perfusion data scanned by the medical imaging equipment targeting the organ. Optionally, when the medical imaging equipment scans the perfusion process of the target organ, it acquires scan data in the x-axis, y-axis, and z-axis directions, and performs imaging based on these scan data to obtain three-dimensional image data. Optionally, the z-axis direction is the axial direction of the cross-section of the scanning channel of the medical imaging equipment; the horizontal direction perpendicular to the z-axis direction is called the x-axis direction; and the directions perpendicular to both the x-axis and z-axis directions are called the y-axis directions. The four-dimensional perfusion data includes the three-dimensional image data and the corresponding scan time.

[0065] Step S104: Detect whether the target artery exists in the slice image of the three-dimensional image data in the target direction.

[0066] The target direction refers to the axial direction of the cross-section of the scanning channel of the medical imaging equipment. Taking the Z-axis as an example, the Z-axis contains multiple slices. A slice image refers to a two-dimensional image scanned in the XOY plane; therefore, slice images of multiple slices constitute three-dimensional image data.

[0067] Specifically, please refer to the following: Figure 2The terminal performs target artery vessel detection on the slice image of the 3D image data in the target direction, and determines whether the target artery vessel exists in the slice image. Optionally, the terminal inputs the slice image of the 3D image data in the target direction into a trained target detection model, and uses the trained target detection model to detect the target artery vessel.

[0068] Step S106: If the target artery exists in the slice image of the three-dimensional image data in the target direction, then the target artery image is segmented from the slice image.

[0069] The target artery can be the anterior cerebral artery, middle cerebral artery, or posterior cerebral artery, etc.

[0070] Specifically, please refer to the following: Figure 3 If the target artery exists in the slice image of the 3D image data in the target direction, the terminal segments the target artery image from the slice image. Optionally, the terminal inputs the slice image into a trained target segmentation model, and obtains the target artery image through the trained target segmentation model.

[0071] More specifically, if the target artery exists in multiple slice images in the target direction of the three-dimensional image data, the probability of each slice image containing the target artery is obtained by using a trained target detection model, and the slice image with the highest probability is selected and segmented to obtain the target artery image.

[0072] Step S108: Based on the first pixel cluster obtained by pixel clustering in the target arterial image, determine the arterial input function required for perfusion parameter calculation of the target organ.

[0073] Among them, a pixel cluster refers to a cluster composed of pixels.

[0074] Among them, the arterial input function (AIF) is used to characterize the concentration change curve of contrast agent in arteries.

[0075] Specifically, the terminal clusters the pixels in the target arterial image to obtain a first pixel cluster. Then, it determines the arterial input function points based on the first pixel cluster. Finally, it determines the arterial input function required for perfusion parameter calculation of the target organ based on the arterial input function points. Optionally, the terminal performs k-means clustering on the pixels in the target arterial image to obtain the first pixel cluster.

[0076] In the aforementioned method for determining the arterial input function for perfusion imaging, the presence of a target artery in the slice image of the three-dimensional image data corresponding to the four-dimensional perfusion data in the target direction is first detected. If present, the target artery image is segmented from the slice image, and then the arterial input function is determined based on this target artery image. This application obtains the target artery image through target detection and image segmentation, reducing interference information in the image. Therefore, determining the arterial input function using this target artery image is beneficial for improving the accuracy of the arterial input function.

[0077] In one embodiment, one possible implementation of step S108, "determining the arterial input function required for perfusion parameter calculation of the target organ based on the first pixel cluster obtained by pixel clustering in the target arterial image," is as follows: Based on the above embodiment, step S108 can be specifically implemented through the following steps:

[0078] Step S1082: Determine the first perfusion characteristic value corresponding to the pixel in the target artery image based on the perfusion parameters associated with the pixel in the target artery image;

[0079] Step S1084: Based on the first perfusion characteristic value, cluster the pixels in the target arterial image to obtain the first pixel cluster, and determine the arterial input function required by the target organ when calculating the perfusion parameters.

[0080] Perfusion parameters refer to data generated during the perfusion process of blood entering the tissues of the target organ. Perfusion parameters include one or more of the following: time to peak concentration, contrast agent concentration, or average transit time of the contrast agent. When imaging the perfusion process of a target organ based on multiple scan times, a correlation can be established between pixels in the three-dimensional image data obtained at each scan time and the perfusion parameters generated during the perfusion process at the corresponding location within the target organ at that scan time.

[0081] Specifically, please refer to Figure 4The terminal acquires perfusion parameters associated with pixels in the target arterial image and calculates the first perfusion characteristic value corresponding to each pixel in the target arterial image. Optionally, the terminal acquires the peak time and its weight, contrast agent concentration and its weight, and the average transit time of the contrast agent and its weight associated with each pixel in the target arterial image. It calculates the first product of the peak time and its weight, the second product of the contrast agent concentration and its weight, and the third product of the average transit time of the contrast agent and its weight. The first, second, and third products are then added together to obtain the first perfusion characteristic value corresponding to each pixel in the target arterial image. Optionally, an early peak time, high contrast agent concentration, and short average transit time of the contrast agent are characteristics of a good arterial input function. Therefore, the weights of different types of perfusion parameters can be set based on these good characteristics. Next, the terminal clusters the pixels in the target arterial image according to the first perfusion characteristic value to obtain multiple first pixel clusters. From these multiple first pixel clusters, the terminal selects the first pixel clusters that meet the first preset pixel cluster conditions as the first arterial input function points. Based on the first arterial input function points, the terminal determines the arterial input function required by the target organ for perfusion parameter calculation. Optionally, the first preset pixel cluster condition can be a condition such as the highest average value of the first injection characteristic or the second highest average value of the first injection characteristic.

[0082] In this embodiment, the perfusion parameters associated with the pixels in the target arterial image are used to score each pixel. Since the perfusion parameters can characterize the characteristics of the perfusion process, the first perfusion characteristic value of the pixel can accurately measure the perfusion characteristics of each position in the target organ corresponding to each pixel, thus further improving the accuracy of the arterial input function.

[0083] In one embodiment, one possible implementation of the above step "selecting a first pixel cluster that satisfies a first preset pixel cluster condition from a plurality of first pixel clusters as a first arterial input function point" is as follows.

[0084] Based on the above embodiments, this step can be specifically implemented through the following steps:

[0085] Step S108a: Based on the first infusion characteristic value, calculate the sum of the first infusion characteristic values ​​of the pixels contained in each first pixel cluster in the plurality of first pixel clusters;

[0086] Step S108b: Calculate the average value of the first injection characteristic corresponding to each first pixel cluster based on the sum of the first injection characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster;

[0087] Step S108c: Select the first pixel cluster with the highest average value of the first perfusion characteristics from multiple first pixel clusters as the first artery input function point.

[0088] Specifically, the terminal calculates the sum of the first perfusion characteristic values ​​of the pixels contained in each of the multiple first pixel clusters based on the first perfusion characteristic value (obtained by summing the first perfusion characteristic values ​​of all pixels in the first pixel cluster). Then, based on the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster, the terminal calculates the average first perfusion characteristic value corresponding to each first pixel cluster. Afterward, the terminal selects the first pixel cluster with the highest average first perfusion characteristic value from the multiple first pixel clusters as the first artery input function point.

[0089] In this embodiment, the first pixel cluster with the highest average value of the first perfusion characteristics is used to determine the arterial input function, which helps to improve the accuracy of the arterial input function.

[0090] In one embodiment, such as Figure 5 As shown, the method also includes the following steps:

[0091] Step S112: If the target artery is not present in the slice image of the three-dimensional image data in the target direction, then the brain blood vessel image is segmented from the slice image.

[0092] Step S114: Select the anterior half of the cerebral blood vessels from the cerebral blood vessel images;

[0093] Step S116: Based on the second pixel cluster obtained by pixel clustering in the first half of the cerebral vascular image, determine the arterial input function required for the target organ when calculating perfusion parameters.

[0094] The anterior hemisphere can be understood as the area in the front of the brain, while the posterior hemisphere is the area in the back of the brain.

[0095] Specifically, please refer to the following: Figure 6If the target artery is not present in the slice image of the 3D image data in the target direction, the terminal segments the cerebral blood vessel image from all or part of the slice image. Optionally, the terminal uses an adaptive threshold segmentation method to segment the cerebral blood vessel image from the slice image. Optionally, the terminal uses a trained support vector machine to segment the cerebral blood vessel image from the slice image. Next, considering that arteries are abundant in the anterior hemisphere and veins are abundant in the posterior hemisphere, the terminal crops the anterior hemisphere blood vessel image from the cerebral blood vessel image to filter out as many arteries as possible. Optionally, the terminal crops the slice image with preset cropping parameters, identifying the cropped image containing the anterior half or frontal lobe of the brain as the anterior hemisphere blood vessel image, and the other cropped portion as the posterior hemisphere blood vessel image. In most cases, the image including the frontal lobe can be identified as the anterior hemisphere blood vessel image. Optionally, the cropping parameters can be a 50 / 50 crop or a cropping with a preset cropping ratio, such as a 1 / 3:2 / 3 cropping ratio. Finally, the terminal clusters the pixels in the anterior cerebral vascular image to obtain a second pixel cluster. Then, it determines the arterial input function points based on the second pixel cluster, and finally determines the arterial input function required by the target organ for perfusion parameter calculation. Optionally, the terminal clusters the pixels in the anterior cerebral vascular image based on the perfusion parameters associated with the pixels to obtain the second pixel cluster.

[0096] In this embodiment, the use of anterior brain images containing a large number of arterial vessels is beneficial to improving the accuracy of the arterial input function.

[0097] In one embodiment, one possible implementation of step S116, "determining the arterial input function required for perfusion parameter calculation of the target organ based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image," is as follows: Based on the above embodiment, step S116 can be specifically implemented through the following steps:

[0098] Step S1162: Calculate the second perfusion characteristic value corresponding to the pixel in the anterior cerebral vascular image based on the perfusion parameters associated with the pixel in the anterior cerebral vascular image.

[0099] Step S1164: Based on the second perfusion characteristic value, cluster the pixels in the first half of the cerebral vascular image to obtain the second pixel cluster, and determine the arterial input function required by the target organ when calculating the perfusion parameters.

[0100] Specifically, please refer to Figure 7The terminal acquires perfusion parameters associated with pixels in the anterior cerebral vascular image and determines the second perfusion characteristic value corresponding to the pixels in the anterior cerebral vascular image. Optionally, the terminal acquires the peak time and its weight, contrast agent concentration and its weight, and the average transit time of the contrast agent and its weight associated with the pixels in the anterior cerebral vascular image. It calculates the fourth product of the peak time and its weight, the fifth product of the contrast agent concentration and its weight, and the sixth product of the average transit time of the contrast agent and its weight. The fourth, fifth, and sixth products are added together to obtain the second perfusion characteristic value corresponding to the pixels in the anterior cerebral vascular image. Next, the terminal clusters the pixels in the anterior cerebral vascular image according to the second perfusion characteristic value to obtain multiple second pixel clusters. From the multiple second pixel clusters, the second pixel cluster that meets the second preset pixel cluster condition is selected as the second artery input function point. Optionally, the second preset pixel cluster condition can be the highest average value of the second perfusion characteristic, the second highest average value of the second perfusion characteristic, etc. Finally, the terminal clusters the pixels in the first half of the cerebral vascular image based on the second perfusion characteristic value to determine the arterial input function required by the target organ when calculating the perfusion parameters.

[0101] In this embodiment, the perfusion parameters associated with pixels in the anterior cerebral vascular image are used to score each pixel. Since the perfusion parameters can characterize the characteristics of the perfusion process, the second perfusion characteristic value of the pixel can accurately measure the perfusion characteristics of each position in the target organ corresponding to each pixel, thus further improving the accuracy of the arterial input function.

[0102] In one embodiment, one possible implementation of step S102, "acquiring three-dimensional image data corresponding to the four-dimensional perfusion data of a medical imaging device scanning a target organ," is as follows: Based on the above embodiment, step S102 can be specifically implemented through the following steps:

[0103] Step S1022: Calculate the average value of the four-dimensional perfusion data of the target organ scan by the medical imaging equipment in the scanning time dimension to obtain three-dimensional image data.

[0104] Alternatively, in step S1024, the maximum value is calculated in the scanning time dimension of the four-dimensional perfusion data obtained by the medical imaging equipment scanning the target organ to obtain three-dimensional image data.

[0105] Specifically, the terminal acquires multiple sets of four-dimensional (4D) perfusion data from medical imaging equipment scanning the target organ. This 4D perfusion data includes three-dimensional image data and the corresponding scan time. Then, the terminal preprocesses the 4D perfusion data along the scan time dimension to obtain the three-dimensional image data. Optionally, please refer to [link to relevant documentation]. Figure 8The terminal calculates the average value of the four-dimensional perfusion data of the target organ scanned by the medical imaging equipment in the scanning time dimension to obtain three-dimensional image data. Optionally, the terminal calculates the maximum value of the four-dimensional perfusion data of the target organ scanned by the medical imaging equipment in the scanning time dimension to obtain three-dimensional image data.

[0106] More specifically, the three-dimensional image data is obtained by averaging the four-dimensional perfusion data of a target organ scan using medical imaging equipment along the scanning time dimension, and can be calculated using the following formula:

[0107]

[0108] Where im(i,t) represents four-dimensional perfusion data, i∈I, i represents the pixel position in the three-dimensional image data, t∈T, t represents the number of scans, and t max This indicates the number of scan times.

[0109] Optionally, after obtaining the three-dimensional image data, color markers are added to the blood vessels in the three-dimensional image data to obtain three-dimensional image data in which the blood vessels are highlighted. By averaging the four-dimensional perfusion data over time, the obtained three-dimensional image data has less noise information compared to the four-dimensional perfusion data, which is beneficial to improving the accuracy of the arterial input function.

[0110] In this embodiment, by preprocessing the four-dimensional perfusion data in the time dimension to obtain three-dimensional image data, noise information in the three-dimensional image data can be reduced, the adverse effects of interference information in the three-dimensional image data on the determination of the arterial input function can be reduced, and the accuracy of the arterial input function can be improved.

[0111] The following describes an embodiment of this application using a specific application scenario. See details below. Figure 9 As shown, the method includes the following steps:

[0112] Step S201: Calculate the average value of the four-dimensional perfusion data of the target organ scan by the medical imaging equipment in the scanning time dimension to obtain three-dimensional image data.

[0113] Step S202: Input the slice image of the three-dimensional image data in the target direction into the trained target detection model, and use the target detection model to detect whether the middle cerebral artery exists in the slice image of the three-dimensional image data in the target direction;

[0114] Step S203: If the middle cerebral artery is present in the slice image of the three-dimensional image data in the target direction, the slice image is input into the trained target segmentation model, and the middle cerebral artery image is obtained by segmentation through the target segmentation model.

[0115] Step S204: Determine the arterial input function required for brain perfusion parameter calculation based on the pixel clusters obtained by pixel clustering in the middle cerebral artery image;

[0116] Step S205: If the middle cerebral artery is not present in the slice image of the three-dimensional image data in the target direction, then the cerebral blood vessel image is segmented from the slice image using an adaptive threshold segmentation method.

[0117] Step S206: Select the anterior cerebral vascular image from the cerebral vascular image, and determine the arterial input function required for the target organ to calculate perfusion parameters based on the pixel clusters obtained by pixel clustering in the anterior cerebral vascular image.

[0118] Optionally, the object detection model includes a deep learning classification network model. Taking the MobileNetV2 deep learning classification network model as an example, the terminal uses the MobileNetV2 model to classify and predict the layered images of the 3D image data in the target direction, obtaining the probability that each layered image contains the middle cerebral artery. The closer the probability is to 1, the greater the likelihood that it contains the middle cerebral artery; the closer the probability is to 0, the smaller the likelihood that it contains the middle cerebral artery. The training and validation of the MobileNetV2 model are based on manually labeled four-dimensional perfusion data of the brain. In the gold standard, 0 represents not containing the middle cerebral artery, and 1 represents containing the middle cerebral artery. The layered image size input to the MobileNetV2 model is 512*512, and the gray values ​​of the layered images are normalized to the range of 0 and 1 using a fixed window width and window level method. The training of the MobileNetV2 model uses cross-entropy loss and the Adam optimizer. The probability p of the layered image in the target direction containing the middle cerebral artery is obtained. i Then, select the slice image i with the highest probability. max and its probability like Less than the probability threshold p threshold The terminal determines that the slice image does not contain the middle cerebral artery; if Greater than or equal to the probability threshold p threshold The terminal determined that the slice image contained the middle cerebral artery.

[0119] Optionally, the target segmentation model includes a deep learning segmentation network model. Taking a 2D UNet-based deep learning segmentation network model as an example, the terminal uses the 2D UNet-based model to segment the middle cerebral artery in the slice image, obtaining the middle cerebral artery image. The training and validation of the 2D UNet-based model are based on manually labeled four-dimensional perfusion data of the brain. In the gold standard, 0 represents the background and 1 represents the middle cerebral artery. The slice image size input to the 2D UNet-based model is 512*512, and the grayscale values ​​of the slice image are normalized to the range of 0 and 1 using a fixed window width and window level method. The training of the 2D UNet-based model uses cross-entropy loss and the Adam optimizer.

[0120] In this embodiment, when the data of the middle cerebral artery is present in the three-dimensional image data of the brain, the most suitable arterial input function can be accurately determined on the middle cerebral artery. When the data of the middle cerebral artery is not present in the three-dimensional image data of the brain, the most suitable arterial input function can be determined on the blood vessels of the anterior brain, which helps to improve the reliability of the determination of the arterial input function.

[0121] It should be understood that, although Figure 1-9 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order in which these steps are executed, and they can be performed in other orders. Furthermore, Figure 1-9 At least some of the steps in the process may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the steps or stages in other steps.

[0122] In one embodiment, such as Figure 10 As shown, an apparatus for determining an arterial input function for perfusion imaging is provided, comprising: an image acquisition module 302, a blood vessel detection module 304, an image segmentation module 306, and a function determination module 308, wherein:

[0123] The image acquisition module 302 is used to acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the medical imaging equipment scanning the target organ;

[0124] The blood vessel detection module 304 is used to detect whether a target artery exists in a slice image of three-dimensional image data in a target direction, wherein the target direction is the axial direction of the cross section of the scanning channel of the medical imaging equipment.

[0125] The image segmentation module 306 is used to segment the target artery image from the slice image if the three-dimensional image data contains a target artery in the slice image in the target direction.

[0126] The function determination module 308 is used to determine the arterial input function required for perfusion parameter calculation of the target organ based on the first pixel cluster obtained by pixel clustering in the target arterial image.

[0127] In the aforementioned apparatus for determining the arterial input function for perfusion imaging, the presence of a target artery in a slice image of the three-dimensional image data corresponding to the four-dimensional perfusion data in the target direction is first detected. If present, the target artery image is segmented from the slice image, and then the arterial input function is determined based on this target artery image. This application obtains the target artery image through target detection and image segmentation, reducing interference information in the image. Therefore, determining the arterial input function using this target artery image helps improve the accuracy of the arterial input function.

[0128] In one embodiment, the function determination module 308 is specifically used to determine a first perfusion characteristic value corresponding to a pixel in the target arterial image based on perfusion parameters associated with the pixel in the target arterial image; and to determine a first pixel cluster obtained by clustering the pixels in the target arterial image based on the first perfusion characteristic value, thereby determining the arterial input function required by the target organ when calculating the perfusion parameters.

[0129] In one embodiment, the function determination module 308 is specifically used to cluster pixels in the target arterial image according to the first perfusion characteristic value to obtain multiple first pixel clusters; select the first pixel cluster that meets the first preset pixel cluster condition from the multiple first pixel clusters as the first arterial input function point; and determine the arterial input function required by the target organ when calculating perfusion parameters according to the first arterial input function point.

[0130] In one embodiment, the function determining module 308 is specifically configured to: calculate the sum of the first perfusion characteristic values ​​of the pixels contained in each of the plurality of first pixel clusters based on the first perfusion characteristic value; calculate the average first perfusion characteristic value corresponding to each first pixel cluster based on the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster; and select the first pixel cluster with the highest average first perfusion characteristic value from the plurality of first pixel clusters as the first artery input function point.

[0131] In one embodiment, it further includes: an image selection module, wherein:

[0132] The image segmentation module 306 is also used to segment the brain blood vessel image from the slice image if the target artery does not exist in the slice image in the target direction of the three-dimensional image data.

[0133] This image selection module is used to select images of the anterior half of the brain's blood vessels from brain vascular images;

[0134] The function 308 is also used to determine the arterial input function required for perfusion parameter calculation of the target organ based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image.

[0135] In one embodiment, the function determination module 308 is specifically used to calculate a second perfusion characteristic value corresponding to a pixel in the anterior cerebral vascular image based on perfusion parameters associated with the pixel in the anterior cerebral vascular image; and to determine the arterial input function required by the target organ when calculating the perfusion parameters by clustering the pixels in the anterior cerebral vascular image based on the second perfusion characteristic value to obtain a second pixel cluster.

[0136] In one embodiment, the image acquisition module 302 is specifically used to calculate the average value of the four-dimensional perfusion data of the medical imaging device scanning the target organ in the scanning time dimension to obtain three-dimensional image data.

[0137] Specific limitations regarding the apparatus for determining the arterial input function for perfusion imaging can be found in the limitations of the method for determining the arterial input function for perfusion imaging described above, and will not be repeated here. Each module in the aforementioned apparatus for determining the arterial input function for perfusion imaging can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0138] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 11As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When executed by the processor, the computer program implements a method for determining an arterial input function for perfusion imaging. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0139] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0140] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0141] Acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by medical imaging equipment;

[0142] The detection method aims to determine whether a target artery exists in a slice image of a three-dimensional image data in a target direction, where the target direction is the axial direction of the cross section of the scanning channel of the medical imaging equipment.

[0143] If the target artery exists in the slice image in the target direction of the 3D image data, then the target artery image is segmented from the slice image;

[0144] Based on the first pixel cluster obtained by pixel clustering in the target arterial image, the arterial input function required for perfusion parameter calculation of the target organ is determined.

[0145] In the aforementioned computer device, the presence of a target artery in the slice image of the three-dimensional image data corresponding to the four-dimensional perfusion data in the target direction is first detected. If it exists, the target artery image is segmented from the slice image, and then the arterial input function is determined based on the target artery image. This application obtains the target artery image through target detection and image segmentation, reducing interference information in the image. Therefore, determining the arterial input function through the target artery image helps improve the accuracy of the arterial input function.

[0146] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0147] Based on the perfusion parameters associated with pixels in the target arterial image, a first perfusion characteristic value corresponding to the pixel in the target arterial image is determined; based on the first perfusion characteristic value, a first pixel cluster is obtained by clustering the pixels in the target arterial image, and the arterial input function required by the target organ for perfusion parameter calculation is determined.

[0148] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0149] The pixels in the target arterial image are clustered according to the first perfusion characteristic value to obtain multiple first pixel clusters; the first pixel clusters that meet the first preset pixel cluster conditions are selected from the multiple first pixel clusters as the first arterial input function points; the arterial input function required by the target organ for perfusion parameter calculation is determined according to the first arterial input function points.

[0150] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0151] Based on the first perfusion characteristic value, calculate the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster in the plurality of first pixel clusters; based on the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster, calculate the average first perfusion characteristic value corresponding to each first pixel cluster; select the first pixel cluster with the highest average first perfusion characteristic value from the plurality of first pixel clusters as the first artery input function point.

[0152] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0153] If the target artery is not present in the slice image of the three-dimensional image data in the target direction, the cerebral vascular image is segmented from the slice image; the anterior cerebral vascular image is selected; based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0154] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0155] Based on the perfusion parameters associated with pixels in the anterior cerebral vascular image, a second perfusion characteristic value corresponding to the pixel in the anterior cerebral vascular image is calculated; based on the second perfusion characteristic value, the pixels in the anterior cerebral vascular image are clustered to obtain a second pixel cluster, and the arterial input function required by the target organ in the calculation of perfusion parameters is determined.

[0156] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0157] The average value of the four-dimensional perfusion data of the target organ scan by medical imaging equipment is calculated in the scanning time dimension to obtain three-dimensional image data.

[0158] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0159] Acquire three-dimensional image data corresponding to the four-dimensional perfusion data of the target organ scan by medical imaging equipment;

[0160] The detection method aims to determine whether a target artery exists in a slice image of a three-dimensional image data in a target direction, where the target direction is the axial direction of the cross section of the scanning channel of the medical imaging equipment.

[0161] If the target artery exists in the slice image in the target direction of the 3D image data, then the target artery image is segmented from the slice image;

[0162] Based on the first pixel cluster obtained by pixel clustering in the target arterial image, the arterial input function required for perfusion parameter calculation of the target organ is determined.

[0163] In the aforementioned computer-readable storage medium, the presence of a target artery in a slice image in the target direction is first detected in the three-dimensional image data corresponding to the four-dimensional perfusion data. If present, the target artery image is segmented from the slice image, and then the arterial input function is determined based on this target artery image. This application obtains the target artery image through target detection and image segmentation, reducing interference information in the image. Therefore, determining the arterial input function using this target artery image improves the accuracy of the arterial input function.

[0164] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0165] Based on the perfusion parameters associated with pixels in the target arterial image, a first perfusion characteristic value corresponding to the pixel in the target arterial image is determined; based on the first perfusion characteristic value, a first pixel cluster is obtained by clustering the pixels in the target arterial image, and the arterial input function required by the target organ for perfusion parameter calculation is determined.

[0166] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0167] The pixels in the target arterial image are clustered according to the first perfusion characteristic value to obtain multiple first pixel clusters; the first pixel clusters that meet the first preset pixel cluster conditions are selected from the multiple first pixel clusters as the first arterial input function points; the arterial input function required by the target organ for perfusion parameter calculation is determined according to the first arterial input function points.

[0168] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0169] Based on the first perfusion characteristic value, calculate the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster in the plurality of first pixel clusters; based on the sum of the first perfusion characteristic values ​​of the pixels contained in each first pixel cluster and the number of pixels contained in each first pixel cluster, calculate the average first perfusion characteristic value corresponding to each first pixel cluster; select the first pixel cluster with the highest average first perfusion characteristic value from the plurality of first pixel clusters as the first artery input function point.

[0170] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0171] If the target artery is not present in the slice image of the three-dimensional image data in the target direction, the cerebral vascular image is segmented from the slice image; the anterior cerebral vascular image is selected; based on the second pixel cluster obtained by pixel clustering in the anterior cerebral vascular image, the arterial input function required for the perfusion parameter calculation of the target organ is determined.

[0172] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0173] Based on the perfusion parameters associated with pixels in the anterior cerebral vascular image, a second perfusion characteristic value corresponding to the pixel in the anterior cerebral vascular image is calculated; based on the second perfusion characteristic value, the pixels in the anterior cerebral vascular image are clustered to obtain a second pixel cluster, and the arterial input function required by the target organ in the calculation of perfusion parameters is determined.

[0174] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0175] The average value of the four-dimensional perfusion data of the target organ scan by medical imaging equipment is calculated in the scanning time dimension to obtain three-dimensional image data.

[0176] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0177] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0178] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for determining an arterial input function for perfusion imaging, characterized in that, The method comprises: obtaining three-dimensional image data corresponding to four-dimensional perfusion data scanned by a medical imaging device for a target organ; the target organ is a brain; detecting whether a target arterial vessel exists in a slice image of the three-dimensional image data in a target direction, wherein the target direction is a transverse axial direction of a scanning channel of the medical imaging device; if the target arterial vessel exists in the slice image of the three-dimensional image data in the target direction, segmenting a target arterial vessel image from the slice image; determining an arterial input function required by the target organ in perfusion parameter calculation according to a first pixel cluster obtained by clustering pixels in the target arterial vessel image; if the target arterial vessel does not exist in the slice image of the three-dimensional image data in the target direction, segmenting a cerebral vessel image from the slice image; selecting a pre-cerebral vessel image in the cerebral vessel image; determining the arterial input function required by the target organ in perfusion parameter calculation according to a second pixel cluster obtained by clustering pixels in the pre-cerebral vessel image.

2. The method of claim 1, wherein, The method comprises: determining a first perfusion characteristic value corresponding to a pixel in the target arterial vessel image according to a perfusion parameter associated with the pixel in the target arterial vessel image; determining the arterial input function required by the target organ in perfusion parameter calculation according to a first pixel cluster obtained by clustering the pixels in the target arterial vessel image according to the first perfusion characteristic value.

3. The method of claim 2, wherein, The method comprises: obtaining a plurality of first pixel clusters by clustering the pixels in the target arterial vessel image according to the first perfusion characteristic value; selecting a first pixel cluster satisfying a first preset pixel cluster condition from the plurality of first pixel clusters as a first arterial input function point; determining the arterial input function required by the target organ in perfusion parameter calculation according to the first arterial input function point.

4. The method of claim 3, wherein, The method comprises: calculating a first perfusion characteristic value sum of pixels included in each first pixel cluster in the plurality of first pixel clusters according to the first perfusion characteristic value; calculating a first perfusion characteristic average value corresponding to each first pixel cluster according to the first perfusion characteristic value sum of the pixels included in each first pixel cluster and a number of the pixels included in each first pixel cluster; selecting a first pixel cluster with the highest first perfusion characteristic average value from the plurality of first pixel clusters as the first arterial input function point.

5. The method according to any one of claims 1 to 4, characterized in that, The method comprises: calculating a second perfusion characteristic value corresponding to a pixel in the pre-hemisphere cerebral vessel image according to a perfusion parameter associated with the pixel in the pre-hemisphere cerebral vessel image; determining an arterial input function required by the target organ in perfusion parameter calculation according to a second pixel cluster obtained by clustering pixels in the pre-hemisphere cerebral vessel image according to the second perfusion characteristic value.

6. The method according to any one of claims 1 to 4, characterized in that, The three-dimensional image data corresponding to the four-dimensional perfusion data scanned by the medical imaging device for the target organ includes: calculating an average value of the four-dimensional perfusion data scanned by the medical imaging device for the target organ in the scanning time dimension to obtain the three-dimensional image data.

7. The method according to any one of claims 1 to 4, characterized in that, The cerebral vessel image is segmented from the slice image by using an adaptive threshold segmentation method. The device includes:

8. An apparatus for determining an arterial input function for perfusion imaging, characterized in that An image acquisition module configured to acquire three-dimensional image data corresponding to four-dimensional perfusion data scanned by a medical imaging device for a target organ; the target organ is a brain; A vessel detection module configured to detect whether a target arterial vessel exists in a slice image of the three-dimensional image data in a target direction; the target direction is an axial direction of a transverse plane of a scanning channel of the medical imaging device; An image segmentation module configured to segment a target arterial vessel image from the slice image if the target arterial vessel exists in the slice image of the three-dimensional image data in the target direction; A function determination module configured to determine an arterial input function required by the target organ in perfusion parameter calculation according to a first pixel cluster obtained by clustering pixels in the target arterial vessel image; The image segmentation module is further configured to segment a cerebral vessel image from the slice image if the target arterial vessel does not exist in the slice image of the three-dimensional image data in the target direction; An image selection module configured to select a pre-hemisphere cerebral vessel image from the cerebral vessel image; The function determination module is further configured to determine the arterial input function required by the target organ in perfusion parameter calculation according to a second pixel cluster obtained by clustering pixels in the pre-hemisphere cerebral vessel image. The processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The computer program, when executed by the processor, implements the steps of the method of any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, ​

Images