A method and system for automatically extracting blood input function based on whole-body pet image

Abstract

The application provides a blood input function automatic extraction method and system based on whole-body PET images, and the method comprises the following steps: S1, acquiring a maximum intensity projection image of the upper body of a patient in a sagittal plane; S2, positioning a typical cardiovascular image layer based on the maximum intensity projection image of the upper body in the sagittal plane to obtain a plurality of typical cardiovascular layer images; S3, positioning a descending aorta based on K-means clustering for the plurality of typical cardiovascular layer images to obtain a descending aorta region selection result; and S4, extracting a blood input function based on the descending aorta region selection result. The application realizes automatic extraction of blood signals in the descending aorta, i.e. blood input functions, effectively improves the efficiency of parameter analysis and imaging, and helps to realize more efficient and accurate dynamic PET pharmacokinetic analysis.

Description

TECHNICAL FIELD

[0001] The embodiment of the application relates to the field of medical imaging technology, and particularly relates to a blood input function automatic extraction method and system based on whole-body PET imaging. BACKGROUND

[0002] Extraction of blood input function (BIF) is an important part of drug research, clinical diagnosis and precision treatment. It not only helps doctors and researchers accurately describe the in-vivo distribution and metabolism of drugs, but also provides key data for personalized treatment, disease diagnosis and image quantitative analysis. By accurately extracting BIF, a more suitable treatment plan can be designed according to the pharmacokinetic characteristics of individuals, especially in some treatments that require precise dose control such as cancer treatment and antibiotic treatment.

[0003] In PET imaging (Positron Emission Tomography), BIF reflects the dynamic process of drugs or radioactive markers entering the bloodstream. If BIF can be accurately extracted, more accurate data on the in-vivo distribution of drugs or markers can be obtained in PET imaging, thereby improving the accuracy of imaging analysis. Currently, there is no technical report on fully automatic blood input function extraction based on whole-body PET imaging data. The current method for obtaining blood input function in the clinic is based on manual picking of blood vessels on the image by an analyst or dynamic arterial blood sampling. The related prior art includes an end-to-end parameter map generation method based on artificial intelligence and a population-based blood input function (PBIF).

[0004] The problems of the pharmacokinetic parameter imaging method based on artificial intelligence (deep learning network) are: (1) a large amount of training data is required; (2) when the training data comes from the same center or the same medical imaging instrument, the algorithm will have generalization and stability problems; (3) the algorithm has poor interpretability, and the calculation result is difficult to obtain the trust of clinicians.

[0005] The introduction of PBIF in the population-based blood input function mainly solves the problem of too long scanning time, and cannot help to realize fully automatic blood input function BIF extraction. The acquisition of part of the blood input function BIF still needs to be manually performed by the clinical analyst in the image, and therefore the efficiency of parameter analysis and imaging cannot be effectively improved. SUMMARY

[0006] Therefore, the embodiment of the present application provides a blood input function automatic extraction method and system based on whole-body PET images. 18 fluoro-fluoroglucose 18 (F-FDG) whole-body dynamic PET images, which helps to realize more efficient and accurate dynamic PET pharmacokinetic analysis.

[0007] According to a first aspect of the embodiment of the present application, a blood input function automatic extraction method based on whole-body PET images is provided, which includes the following steps: S1, obtaining an upper body sagittal maximum intensity projection image of a patient; S2, performing cardiovascular typical image layer positioning based on the upper body sagittal maximum intensity projection image to obtain a plurality of cardiovascular typical layer images; S3, performing descending aorta positioning based on K-means clustering for the plurality of cardiovascular typical layer images to obtain a descending aorta region selection result; and S4, extracting a blood input function based on the descending aorta region selection result.

[0008] In an implementation manner, the step S1 includes: defining an upper body part of a PET whole-body image of the patient as a three-dimensional image within a range of 1 meter from the top of the head; performing axial intercepting setting according to a layer thickness parameter of the PET whole-body image; selecting a plurality of layer images starting from the top of the head as the upper body part in the PET whole-body image, and intercepting each frame of the entire dynamic image sequence in the PET whole-body image according to the set axial direction to obtain an upper body dynamic image sequence; obtaining a difference image of an early frame and a late frame in the upper body dynamic image sequence; and performing maximum intensity projection of the sagittal plane on the difference image to obtain the upper body sagittal maximum intensity projection image of the patient.

[0009] In another implementation manner, the difference image is obtained by the following formula:

[0010] X dif = X early -X late

[0011] X dif represents the difference image, X early represents the early frame, and X late represents the late frame.

[0012] In another implementation manner, the step S2 includes: performing amplitude truncation of the upper body sagittal maximum intensity projection image based on a threshold parameter ξ mip , and MIP amplitude > ξ mipThe effective value is defined as follows: the number of effective voxels in each row of the upper body sagittal maximum intensity projection image is calculated in the image layer direction to obtain an effective voxel count curve; the effective voxel count curve is smoothed based on mean filtering to obtain a smoothed count curve, wherein the kernel size of the smoothing operation is an odd number; the maximum point of the smoothed count curve is approximately corresponding to the image slice with the largest area in the liver in the PET whole body image, which is defined as the typical layer of the liver, and the image layer region above the typical layer of the liver is defined as the axial field of interest AFOI; the bottom 40% of the image layer of the axial field of interest AFOI is defined as the chest region of the patient, and the middle several layers of the chest region are defined as the typical image layers of the cardiovascular system to obtain several typical image layers of the cardiovascular system.

[0013] In another implementation, the horizontal axis of the effective voxel count curve is the layer index of the PET whole body image or the row index of the upper body sagittal maximum intensity projection image, and the vertical axis is the number of effective voxels in each row of the effective upper body sagittal maximum intensity projection image.

[0014] In another implementation, step S3 includes: performing voxel K-means clustering on the n s typical image layers of the cardiovascular system one by one, including: using a set threshold value ξ v to determine the image voxels of the typical image layers of the cardiovascular system, and determining the voxel points with an activity concentration > ξ v in the early frame as blood vessel voxels to form a blood vessel voxel Boolean value image Mask hv ; spatially intercepting the left half of Mask hv , the left half of Mask hv contains the left half of the target descending aorta region, forming a voxel range for inputting K-means clustering; and dividing the voxels in the formed voxel range for inputting K-means clustering into n k classes through a K-means clustering algorithm; after the K-means clustering of the single-layer typical image layers of the cardiovascular system is completed, the cluster belonging to the descending aorta in the current layer is determined as the cluster of the descending aorta in the bottom of the typical image layer of the cardiovascular system, and then the voxels of all descending aortas in the n s typical image layers of the cardiovascular system constitute the descending aorta region selection result.

[0015] In another implementation, the method further includes: defining the number of voxels in a single voxel class as n v , and then n k can be determined by the following formula:

[0016]

[0017] wherein N is the total number of voxels for voxel clustering, and n vThe preset influence value K-means clustering finds the accuracy of descending aortic region.

[0018] According to a second aspect of the embodiment of the present application, an automatic blood input function extraction system based on whole-body PET images is provided, comprising: an acquisition module configured to acquire a maximum intensity projection image of an upper body of a patient in a sagittal plane; a first positioning module configured to position a typical cardiovascular image layer based on the maximum intensity projection image of the upper body in the sagittal plane to obtain a plurality of typical cardiovascular layer images; a second positioning module configured to position a descending aorta based on K-means clustering for the plurality of typical cardiovascular layer images to obtain a descending aorta region selection result; and an extraction module configured to extract the blood input function based on the descending aorta region selection result.

[0019] According to a third aspect of the embodiment of the present application, an electronic device is provided, comprising a processor and a memory storing a program. The program comprises instructions which, when executed by the processor, cause the processor to perform the steps of the method according to the first aspect.

[0020] According to a fourth aspect of the embodiment of the present application, a computer storage medium is provided, which stores a computer program. The program, when executed by a processor, implements the method according to the first aspect.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] (1) The method of the present application does not need to rely on the construction of a large number of training data sets;

[0023] (2) The method of the present application is based on the basic law of human body structure to stably position the descending aorta, and has high interpretability;

[0024] (3) The method of the present application can extract the blood input function based on a single data, and does not cause errors due to the introduction of external data;

[0025] (4) The method of the present application effectively improves the efficiency of parameter analysis and imaging, and helps to realize more efficient and accurate dynamic PET pharmacokinetic analysis. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0027] Figure 1 Flow chart of the method of the present application for automatic extraction of blood input function based on whole-body PET images;

[0028] Figure 2 Fig. 1 is a schematic diagram of a process for acquiring a maximum intensity projection image of a sagittal plane of an upper body;

[0029] Figure 3 Fig. 2 is a schematic diagram of a process for positioning a typical image layer of a cardiovascular system;

[0030] Figure 4 Fig. 3 is a schematic diagram of a process from a typical layer image of a cardiovascular system to a positioning result of a descending aorta and a blood input signal extraction result;

[0031] Figure 5 Fig. 4 is a schematic diagram of a descending aorta region acquired from an early frame image using the method of the present application;

[0032] Figure 6 Fig. 5 is a schematic diagram of a Ki / Vt parameter imaging result based on an automatically acquired blood input function BIF;

[0033] Figure 7 Fig. 6 is a structural block diagram of a blood input function automatic extraction system based on whole-body PET images according to the present application;

[0034] Figure 8 Fig. 7 is a structural schematic diagram of an electronic device according to the present application. DETAILED DESCRIPTION

[0035] In order to make the technical features, objectives and effects of the embodiments of the present application clearer, the specific implementation manners of the embodiments of the present application will be described with reference to the drawings.

[0036] In this document, "exemplarily" means "serving as an example, an instance or an illustration", and any illustration, implementation manner described as "exemplarily" in this document should not be interpreted as a more preferred or more advantageous technical solution.

[0037] In order to make those skilled in the art better understand the technical solutions in the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings of the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art should belong to the protection scope of the present application.

[0038] The specific implementation of the embodiments of the present application will be further described below with reference to the drawings of the embodiments of the present application.

[0039] Referring to Figure 1 The blood input function automatic extraction method based on whole-body PET images provided by the present application is a method based on 18The full-automatic BIF acquisition method of F-FDG whole-body dynamic PET image helps to realize more efficient and accurate dynamic PET pharmacokinetic analysis. The implementation of the method is based on the following points:

[0040] (1) Basic law of human body scale;

[0041] (2) 18 Basic characteristics of F-FDG transport and metabolism time in human body;

[0042] (3) K-means simple data clustering theory.

[0043] The method of the present application mainly comprises the following steps:

[0044] Step S1, acquiring an upper body sagittal maximum intensity projection image of a patient;

[0045] Step S2, positioning a cardiovascular typical image layer based on the upper body sagittal maximum intensity projection image to obtain a plurality of cardiovascular typical layer images;

[0046] Step S3, performing descending aorta positioning based on K-means clustering for the plurality of cardiovascular typical layer images to obtain descending aorta region selection results;

[0047] Step S4, extracting a blood input function based on the descending aorta region selection results.

[0048] It should be understood that the upper body sagittal maximum intensity projection image is a MIP image. The whole-body PET image refers to a three-dimensional image covering the entire region of the human body from head to foot, which is obtained by scanning through the positron emission tomography (PET) technology. It is usually used to observe the metabolic activity or detect the distribution and characteristics of lesions (such as tumors, infections, inflammation, etc.) in the whole body. The descending aorta blood signal, i.e. the blood input function, is abbreviated as BIF in English, and the full name is Blood Input Function.

[0049] The implementation effect of the method of the present application is to realize full-automatic descending aorta blood signal (i.e. BIF) extraction based on whole-body dynamic PET image data obtained by using the uEXPLORER PET / CT imaging system of Union Imaging Company, for subsequent pharmacokinetic analysis and imaging.

[0050] Optionally, the step S1 comprises: defining the upper body part of the PET whole-body image of the patient as a three-dimensional image within a range of 1 meter from the top of the head; setting axial intercept according to the layer thickness parameter of the PET whole-body image reconstruction; selecting several layer images starting from the top of the head in the PET whole-body image as the upper body part, intercepting each frame of the whole dynamic image sequence in the PET whole-body image according to the set axial direction to obtain the upper body dynamic image sequence; obtaining the difference image of the early frame and the late frame in the upper body dynamic image sequence; obtaining the sagittal maximum intensity projection image of the upper body of the patient by performing sagittal maximum intensity projection on the difference image.

[0051] Optionally, the difference image is obtained by the following formula:

[0052] X dif = X early - X late

[0053] wherein X dif represents the difference image, X early represents the early frame, and X late represents the late frame.

[0054] Optionally, the step S2 comprises: performing amplitude truncation on the sagittal maximum intensity projection image of the upper body based on a threshold parameter ξ mip , and MIP amplitude > ξ mip is defined as an effective value; calculating the number of effective voxels of each row of the sagittal maximum intensity projection image of the upper body in the image layer direction to obtain an effective voxel count curve; performing smoothing processing on the effective voxel count curve based on mean filtering to obtain a smoothed count curve, wherein the kernel size of the smoothing operation is an odd number; approximating the maximum value point of the smoothed count curve to the image slice in the PET whole-body image with the largest area in the liver, i.e., defining it as a typical layer of the liver, and defining the image layer region above the typical layer of the liver as an axial region of interest AFOI; defining the bottom 40% of the image layer of the axial region of interest AFOI as the chest region of the patient, and defining the middle several layers of the chest region as the typical image layers of the cardiovascular system to obtain several typical layer images of the cardiovascular system.

[0055] Optionally, the horizontal axis of the effective voxel count curve is the layer index of the PET whole-body image or the row index of the sagittal maximum intensity projection image of the upper body, and the vertical axis is the number of effective voxels of each row of the effective sagittal maximum intensity projection image of the upper body.

[0056] Optionally, the step S3 comprises: performing voxel K-means clustering on the n s typical layer images of the cardiovascular system of the early frame, comprising: using a set threshold ξ v to determine the image voxels of the typical layer images of the cardiovascular system, and the activity concentration in the early frame > ξv The voxel point is determined as a blood vessel voxel to form a blood vessel voxel Boolean value image Mask hv ; the Mask is spatially intercepted hv Left half, Mask hv The left half contains the left half of the target descending aorta region, forming a voxel range for inputting K-means clustering; the voxels in the formed voxel range for inputting K-means clustering are divided into n k classes by a K-means clustering algorithm; after the K-means clustering of the single-layer cardiovascular typical layer image is completed, the cluster at the bottom of the cardiovascular typical layer image space is determined as the cluster to which the descending aorta belongs in the layer, and n s The voxels of all descending aortas in the n

[0057] Optionally, the method further comprises: defining the number of voxels in a single voxel class as n v , then n k can be determined by the following formula:

[0058]

[0059] Wherein, N is the total number of voxels for voxel clustering, n v The preset value of n

[0060] Specifically, the scheme of the present application is further described according to the following examples:

[0061] The implementation process of the present application can be roughly divided into the following steps:

[0062] (1) Upper body sagittal maximum intensity projection (MIP) image acquisition

[0063] The upper body sagittal maximum intensity projection (MIP) image acquisition process is shown in Figure 2 In this method, the upper body part of the PET whole body image of the patient is defined as a three-dimensional image within a range of 1 meter from the top of the head. The axial intercept of the image needs to refer to the image reconstruction layer thickness parameter, for example, the common reconstruction layer thickness of the PET image is 2.886 millimeters, so the upper body part image defined in this method contains about 350 image slices of the whole body image (about 673 layers) from the top slice. The axial intercept is performed on each frame of the entire dynamic image sequence to obtain an upper body dynamic image sequence.

[0064] The difference between the early and late image frames in the dynamic sequence is used to emphasize human vascular tissue. In the dynamic PET image sequence using 18 F-FDG, the early frame (X earlyThe time point is approximately 200 seconds after drug injection, in the later frame (X). late This can be defined as the last frame of the dynamic sequence, approximately 3600 seconds after drug injection. The difference image (X) between the early and later frames... dif The answer can be obtained by direct subtraction:

[0065] X dif =X early -X late

[0066] Among them, X dif Represents the difference image, X early Indicates an early frame, X late This indicates a later frame.

[0067] For X dif Performing maximum intensity projection in the sagittal plane yields the subsequent required upper body sagittal maximum intensity projection (MIP) image.

[0068] (2) Localization of typical cardiovascular imaging layers

[0069] The typical image layer localization process of cardiovascular disease is as follows: Figure 3 As shown. First, the MIP image obtained in the above steps, i.e., the maximum intensity projection image of the upper body in the sagittal plane, is subjected to a threshold parameter ξ. mip Amplitude truncation, MIP amplitude > ξ mip This is defined as the effective value. Then, the effective voxels (MIP>ξ) of each row of the MIP image are calculated in the image layer direction. mip ) quantity, to obtain the effective voxel counting curve.

[0070] It should be noted that the horizontal axis of the effective voxel count curve is the layer label of the original PET image, i.e., the whole-body PET image, or the row label of the upper body sagittal maximum intensity projection image, and the vertical axis is the number of effective voxels in each row of the effective upper body sagittal maximum intensity projection image.

[0071] Subsequently, the effective voxel count curve is smoothed using mean filtering, and the kernel size for the smoothing operation must be odd. The maximum point of the smoothed count curve approximately corresponds to the image slice with the largest area of ​​the liver in the patient's whole-body PET image, which is defined as the typical liver slice. The image layer region above the typical liver slice is defined as the region of interest (AFOI) in this method. The bottom 40% of the image layer of the AFOI covers most of the patient's thoracic cavity region, and the middle layers of this thoracic cavity region are defined as the typical cardiovascular image layers of the patient's image. The number of typical cardiovascular layers is a variable parameter n. s This can be empirically set to 5. n in the early frames of the dynamic avatar s A typical cardiovascular layer image will be used for subsequent K-means clustering-based localization of the descending aorta.

[0072] (3) K-means clustering based descending aorta localization and blood signal extraction

[0073] For a certain example of early frame image n s cardiovascular typical layer images, the present application performs voxel K-means clustering for each image. The process from the cardiovascular typical layer image slice to the descending aorta localization result and the blood input signal extraction result is shown in Figure 4 .

[0074] First, the image voxels are judged using a set threshold ξ v . The voxel points with activity concentration > ξ v in the early frame are determined as blood vessel voxels, forming a blood vessel voxel Boolean value image Mask hv .

[0075] Further, the left half of Mask hv is intercepted in space (in medical image slices, the left and right directions are opposite to vision, for example Figure 4 ), that is, the left half of the target descending aorta region, forming the voxel range for inputting K-means clustering. The above selected voxels are divided into n k classes by the K-means clustering algorithm, n k is a preset value, which can be determined by defining the approximate voxel number contained in a single voxel class. For example, if the approximate voxel number of a single voxel class is defined as n v , then n k can be determined by the following formula:

[0076]

[0077] where N is the total number of voxels for voxel clustering, n v The preset value of n s to some extent affects the accuracy of K-means clustering in finding the descending aorta region.

[0078] After K-means clustering of a single layer image is completed, the cluster at the bottom of the image space can be determined as the descending aorta cluster in this layer. As mentioned earlier, an image of a patient contains n s cardiovascular typical layer images, then all the descending aorta voxels in n s images constitute the descending aorta region selection result, which is used for blood input function extraction.

[0079] Referring to Figure 5 the schematic diagram of the descending aorta region obtained from the early frame image using the method of the present application, Figure 6For the Ki / Vt parameter imaging result schematic diagram based on the automatically acquired blood input function BIF (compared with the standard calculation result), the method has been verified based on the clinical whole body dynamic PET image data set, the experimental result is good, and the effectiveness and feasibility of the method are proved.

[0080] The method is used for full-automatic descending aorta blood input function extraction of whole body dynamic PET image sequence.

[0081] (1) The implementation of the method does not need to rely on the construction of a large number of training data sets;

[0082] (2) The method is based on the stable positioning of the descending aorta based on the basic law of human body structure, and has high interpretability;

[0083] (3) The method can extract the blood input function based on a single data, and does not cause errors due to the introduction of external data;

[0084] (4) The method effectively improves the efficiency of parameter analysis and imaging, and helps to realize more efficient and accurate dynamic PET pharmacokinetic analysis.

[0085] Referring to Figure 7 , the embodiment of the application also provides a blood input function automatic extraction system 700 based on whole body PET image, comprising:

[0086] The acquisition module 710 is used for acquiring the maximum intensity projection image of the upper body of the patient in the sagittal plane;

[0087] The first positioning module 720 is used for positioning the typical cardiovascular image layer based on the maximum intensity projection image of the upper body in the sagittal plane, and obtaining a plurality of typical cardiovascular layer images;

[0088] The second positioning module 730 is used for positioning the descending aorta based on K-means clustering for the plurality of typical cardiovascular layer images, and obtaining a descending aorta region selection result;

[0089] The extraction module 740 is used for extracting the blood input function based on the descending aorta region selection result.

[0090] It should be understood that the blood input function automatic extraction system based on whole body PET image of the embodiment is used for implementing the corresponding method in the plurality of method embodiments, and has the beneficial effects of the corresponding method embodiments.

[0091] As another example, referring to Figure 8, an electronic device 800 is provided, and a structural block diagram of the electronic device 800 which can be a server or a client of the present application will now be described, which is an example of a hardware device that can be applied to aspects of the present application. The electronic device is intended to represent a wide variety of digital electronic computer devices such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computer devices. The electronic device can also represent a wide variety of mobile devices such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0092] The electronic device 800 can include a processor 802, a communications interface 804, a memory 806, and a communications bus 808.

[0093] The processor 802, the communications interface 804, and the memory 806 complete communications with each other through the communications bus 808. The communications interface 804 is configured to communicate with other electronic devices or servers.

[0094] The processor 802 is configured to execute the program 810, and in particular, can execute related steps in the above method embodiments.

[0095] In particular, the program 810 can include program code including computer operation instructions.

[0096] The processor 802 can be a processor CPU, or an application specific integrated circuit ASIC, or one or more integrated circuits configured to implement embodiments of the present application. The one or more processors included in the smart device can be the same type of processor, such as one or more CPUs; or can be different types of processors, such as one or more CPUs and one or more ASICs.

[0097] The memory 806 is configured to store the program 810. The memory 806 can include a high-speed RAM memory, and can also include a non-volatile memory such as at least one disk memory.

[0098] The program 810, when executed by the processor 802, is configured to enable the electronic device to perform the blood input function automatic extraction method based on whole-body PET images, including: step S1, acquiring an upper body sagittal maximum intensity projection image of a patient; step S2, performing typical image layer positioning of a cardiovascular system based on the upper body sagittal maximum intensity projection image, to obtain a plurality of typical cardiovascular layer images; step S3, performing descending aorta positioning based on K-means clustering for the plurality of typical cardiovascular layer images, to obtain a descending aorta region selection result; and step S4, extracting the blood input function based on the descending aorta region selection result.

[0099] In addition, the specific implementation of each step in the program 810 can refer to the corresponding description in the corresponding steps and units in the above-mentioned method embodiments, and will not be described here. It can be clearly understood by those skilled in the art that, for the convenience and brevity of description, the specific working process of the above-mentioned device and module can refer to the corresponding process description in the above-mentioned method embodiments, and will not be described here.

[0100] The exemplary embodiments of the present application also provide a computer storage medium storing a computer program, wherein the computer program is executed by a processor to implement the method of the embodiments of the present application, and can refer to the corresponding process description in the above-mentioned method embodiments, and will not be described here.

[0101] The above-mentioned method according to the embodiments of the present application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium such as a CD ROM, a RAM, a floppy disk, a hard disk or an optical disk, or downloaded through a network and originally stored in a remote recording medium or a non-transitory machine readable medium and then stored in a local recording medium, so that the method described herein can be processed by such software on a recording medium using a general computer, a special processor or programmable or special hardware such as an ASIC or an FPGA. It can be understood that the computer, the processor, the microprocessor controller or the programmable hardware includes a storage component (for example, RAM, ROM, flash memory, etc.) that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, the processor or the hardware, the method described herein is implemented. In addition, when a general computer accesses the code for implementing the method shown herein, the execution of the code will convert the general computer into a special computer for executing the method shown herein.

[0102] To this end, particular embodiments of the present application have been described. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve desirable results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain implementations, multitasking and parallel processing can be advantageous.

[0103] It should be understood that, although the present specification describes various embodiments, each of which contains only one independent technical solution, the specification is merely for clarity and those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that those skilled in the art can understand.

[0104] Finally, it should be noted that: the above implementation is only for illustrating the embodiments of the present application, and not for limiting the embodiments of the present application, the ordinary skilled in the art related to the technology, without departing from the spirit and scope of the embodiments of the present application, can also make various changes and modifications, therefore all equivalent technical solutions also belong to the scope of the embodiments of the present application, the patent protection scope of the embodiments of the present application should be defined by the claims.

Claims

1. A method for automatic extraction of blood input function based on whole-body PET images, characterized in that, The method comprises the following steps: S1, obtaining a maximum intensity projection image of an upper body of a patient in a sagittal plane; Step S2: Based on the upper body sagittal maximum intensity projection image, perform cardiovascular typical image layer localization to obtain several cardiovascular typical layer images, including: performing threshold-based analysis on the upper body sagittal maximum intensity projection image MIP. Amplitude truncation, MIP amplitude This is defined as the effective value; the effective voxel count curve is obtained by calculating the number of effective voxels in each row of the maximum intensity projection image (MIP) of the upper body sagittal plane in the image layer direction; the effective voxel count curve is smoothed by mean filtering to obtain a smoothed count curve, where the kernel size of the smoothing operation is odd; the maximum point of the smoothed count curve is approximately corresponding to the image slice with the largest area of ​​the liver in the whole-body PET image, which is defined as the typical liver layer; the image layer area above the typical liver layer is defined as the region of interest (AFOI); the bottom 40% of the image layer of the region of interest (AFOI) is defined as the patient's thoracic cavity region; and several layers in the middle of the thoracic cavity region are defined as typical cardiovascular image layers to obtain several typical cardiovascular image layers. Step S3: Perform K-means clustering-based descending aorta localization on several typical cardiovascular layer images to obtain descending aorta region selection results, including: for early frames... Image-by-image voxel K-means clustering was performed on typical cardiovascular slice images, including: using a set threshold. Determine the image voxels of typical cardiovascular layer images and the activity concentration in early frames. The voxel points were identified as vascular voxels, and a vascular voxel Boolean image was generated. ; to intercept in space Left half The left half encompasses the left half of the target descending aorta region, forming a voxel range for input K-means clustering; the voxels within this range are then divided using the K-means clustering algorithm. After K-means clustering of a single-layer typical cardiovascular image is completed, the cluster at the bottom of the typical cardiovascular image space is determined to be the cluster to which the descending aorta belongs in this layer. The voxels of all descending aortas in a typical cardiovascular layer image constitute the selection result of the descending aorta region; S4, extracting a blood input function based on the selected result of the descending aorta region.

2. The method of claim 1, wherein, S1 comprises: defining an upper body part of a PET whole body image of the patient as a three-dimensional image within a range of 1 meter from the top of the head; setting axial intercepting according to a layer thickness parameter of the PET whole body image; selecting a number of layer images from the top of the head as the upper body part in the PET whole body image, and intercepting each frame of the whole dynamic image sequence in the PET whole body image according to the set axial direction to obtain an upper body dynamic image sequence; obtaining a difference image of an early frame and a late frame in the upper body dynamic image sequence; performing maximum intensity projection in a sagittal plane on the difference image to obtain a maximum intensity projection image of the upper body of the patient in a sagittal plane.

3. The method of claim 2, wherein, The difference image is obtained by the following formula: wherein, denotes a difference image, denotes an early frame, denotes a late frame.

4. The method of claim 1, wherein, The horizontal axis of the effective voxel count curve is the layer mark of the PET whole body image or the row mark of the maximum intensity projection image of the upper body in a sagittal plane, and the vertical axis is the number of effective voxels of each row of the effective maximum intensity projection image of the upper body in a sagittal plane.

5. The method of claim 1, wherein, The method further comprises: The number of voxels defining a single voxel class is then is determined by the equation: wherein, is the total number of voxels for voxel clustering, The preset influence of the value K-means clustering finds the accuracy of the descending aortic region.

6. A system for automatic extraction of blood input function based on whole-body PET images, characterized by, The method comprises the following steps: an obtaining module configured to obtain a maximum intensity projection image of an upper body of a patient in a sagittal plane; The first localization module is used to locate typical cardiovascular image layers based on the upper body sagittal maximum intensity projection image (MIP), obtaining several typical cardiovascular image layers, including: performing threshold parameter-based localization on the upper body sagittal maximum intensity projection image (MIP). Amplitude truncation, MIP amplitude This is defined as the effective value; the effective voxel count curve is obtained by calculating the number of effective voxels in each row of the maximum intensity projection image (MIP) of the upper body sagittal plane in the image layer direction; the effective voxel count curve is smoothed by mean filtering to obtain a smoothed count curve, where the kernel size of the smoothing operation is odd; the maximum point of the smoothed count curve is approximately corresponding to the image slice with the largest area of ​​the liver in the whole-body PET image, which is defined as the typical liver layer; the image layer area above the typical liver layer is defined as the region of interest (AFOI); the bottom 40% of the image layer of the region of interest (AFOI) is defined as the patient's thoracic cavity region; and several layers in the middle of the thoracic cavity region are defined as typical cardiovascular image layers to obtain several typical cardiovascular image layers. The second localization module is used to perform K-means clustering-based localization of the descending aorta on several typical cardiovascular slice images, obtaining the descending aorta region selection results, including: for early frames... Image-by-image voxel K-means clustering was performed on typical cardiovascular slice images, including: using a set threshold. Determine the image voxels of typical cardiovascular layer images and the activity concentration in early frames. The voxel points were identified as vascular voxels, and a vascular voxel Boolean image was generated. ; to intercept in space Left half The left half encompasses the left half of the target descending aorta region, forming a voxel range for input K-means clustering; the voxels within this range are then divided using the K-means clustering algorithm. After K-means clustering of a single-layer typical cardiovascular image is completed, the cluster at the bottom of the typical cardiovascular image space is determined to be the cluster to which the descending aorta belongs in this layer. The voxels of all descending aortas in a typical cardiovascular layer image constitute the selection result of the descending aorta region; an extracting module configured to extract a blood input function based on the selected result of the descending aorta region.

7. An electronic device, comprising: The method comprises the following steps: a processor; a memory storing a program; wherein the program comprises instructions which, when executed by the processor, cause the processor to perform the steps of the method according to any one of claims 1-5.

8. A computer storage medium, characterized in that A computer program is stored thereon, which is executed by a processor to implement the method according to any one of claims 1-5.

Images