Reconstruction method, device and equipment of ECT image and storage medium

Abstract

This application relates to the field of medical image processing technology, and provides a method, apparatus, device, and storage medium for reconstructing ECT images. This allows for more thorough capture of subtle changes in the concentration distribution of radioactive tracers at high temporal resolution, helping to reduce image noise levels and improve image quality. This application obtains a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device; the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence; using the estimated second ECT image sequence as a priori image, a target second ECT image sequence is reconstructed; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

Description

Technical Field

[0001] This application relates to the field of medical image processing technology, and in particular to a method, apparatus, computer device, storage medium, and computer program product for reconstructing ECT images. Background Technology

[0002] Emission Computed Tomography (ECT) is a technique that uses radiation emitted by a radioactive tracer within the body of the subject to create an image. The resulting image is called an ECT image.

[0003] When performing iterative reconstruction of ECT images, if we try to further improve the temporal resolution and shorten the duration of a single frame, the amount of projection data participating in the iterative reconstruction will be further reduced. If we continue to rely on the prior information of ECT images with longer single-frame durations, then at high temporal resolution, subtle changes in the concentration distribution of radioactive tracers may not be fully captured. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer device, storage medium, and computer program product for reconstructing ECT images to address the aforementioned technical problems.

[0005] This application provides a method for reconstructing ECT images, the method comprising:

[0006] A first ECT image sequence is obtained; the first ECT image sequence is obtained by scanning with an ECT scanning device; the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence;

[0007] Using the estimated second ECT image sequence as a priori images, a target second ECT image sequence is reconstructed; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

[0008] In one embodiment, the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence, including:

[0009] The first ECT image sequence is divided into multiple image groups; several first ECT images in the same image group are consecutive first ECT images.

[0010] The multiple image groups are respectively input into the intermediate frame prediction network to obtain the intermediate frame subsequence corresponding to each image group;

[0011] The intermediate frame sequence is obtained based on the intermediate frame sub-sequence corresponding to each of the image groups;

[0012] The intermediate frame sequence is inserted into the first ECT image sequence to obtain the estimated second ECT image sequence.

[0013] In one embodiment, obtaining a pre-built intermediate frame prediction network includes:

[0014] In the ECT sample image sequence, several ECT sample images are selected at intervals to obtain the input image sequence;

[0015] The training target is obtained from the ECT sample images located between the selected ECT sample images;

[0016] Based on the input image sequence and the training objective, an intermediate frame prediction network is obtained.

[0017] In one embodiment, reconstructing the target second ECT image sequence using the estimated second ECT image sequence as a priori images includes:

[0018] Using the estimated second ECT image sequence as the prior image, the first spatial kernel matrix is ​​obtained;

[0019] The first time kernel matrix is ​​obtained based on the second projection data; the single frame duration corresponding to the second projection data matches the temporal resolution of the second ECT image sequence.

[0020] Based on the first spatial kernel matrix and the first temporal kernel matrix, the target second ECT image sequence is reconstructed.

[0021] In one embodiment, reconstructing the target second ECT image sequence using the estimated second ECT image sequence as a priori images includes:

[0022] Using the estimated second ECT image sequence as the prior image, the prior image is used as the initial image for iteration;

[0023] The iterative initial image is input into the reconstruction algorithm to obtain the target second ECT image sequence.

[0024] In one embodiment, the method further includes:

[0025] Acquire a third ECT image sequence; the temporal resolution of the third ECT image sequence is lower than that of the first ECT image sequence.

[0026] Based on the third ECT image sequence and the first projection data, the first ECT image sequence is reconstructed; the single frame duration corresponding to the first projection data matches the temporal resolution of the first ECT image sequence.

[0027] In one embodiment, reconstructing the first ECT image sequence based on the third ECT image sequence and the first projection data includes:

[0028] The second spatial kernel matrix is ​​obtained based on the third ECT image sequence;

[0029] Based on the first projection data, the second time kernel matrix is ​​obtained;

[0030] The first ECT image sequence is reconstructed based on the second spatial kernel matrix and the second temporal kernel matrix.

[0031] This application provides an apparatus for reconstructing ECT images, the apparatus comprising:

[0032] An image acquisition module is used to acquire a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device.

[0033] The intermediate frame prediction module is used to input the first ECT image sequence into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence;

[0034] The reconstruction processing module is used to reconstruct the target second ECT image sequence using the estimated second ECT image sequence as a priori images.

[0035] This application provides a computer device, including a memory and a processor, wherein the memory stores a computer program and the processor executes the above-described method.

[0036] This application provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor using the methods described above.

[0037] This application provides a computer program product having a computer program stored thereon, the computer program being executed by a processor using the above-described method.

[0038] The aforementioned ECT image reconstruction method, apparatus, computer equipment, storage medium, and computer program product estimate that the second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence; this application obtains the first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device; the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain the estimated second ECT image sequence; the estimated second ECT image sequence can more fully reflect the subtle changes in the concentration distribution of radioactive tracers at high temporal resolution, providing better prior guidance for reconstructing the target second ECT image sequence, helping to reduce the noise level of the target second ECT image sequence and improve image quality. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating a method for reconstructing ECT images in one embodiment;

[0041] Figure 2 This is a schematic diagram illustrating the training of an intermediate frame prediction network in one embodiment;

[0042] Figure 3 This is a schematic diagram of the process for obtaining an estimated second ECT image sequence in one embodiment;

[0043] Figure 4 This is a schematic diagram of the process for obtaining a target second ECT image sequence in one embodiment;

[0044] Figure 5 This is a schematic diagram of the process for obtaining the target second ECT image sequence in another embodiment;

[0045] Figure 6 This is a structural block diagram of an ECT image reconstruction device in one embodiment;

[0046] Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0047] 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.

[0048] The ECT image reconstruction method provided in this application involves ECT scanning equipment, including PET scanning equipment and SPECT scanning equipment. This method can be used for PET image reconstruction as well as SPECT image reconstruction. When used for PET image reconstruction, the ECT image includes the PET image; when used for SPECT image reconstruction, the ECT image includes the SPECT image. PET stands for Positron Emission Tomography. SPECT stands for Single-Photon Emission Computed Tomography.

[0049] The ECT image reconstruction method provided in this application can be performed by computer equipment, including... Figure 1 The steps shown are as follows:

[0050] Step S101: Obtain the first ECT image sequence.

[0051] The first ECT image sequence was obtained by scanning with an ECT scanning device.

[0052] The estimated second ECT image sequence and the target second ECT image sequence mentioned in this application belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

[0053] The first ECT image and the second ECT image are relative concepts. A single-frame ECT image is obtained by reconstructing a certain projection data, which has a corresponding single-frame duration. When two different single-frame durations exist, and one single-frame duration is longer than the other, the longer single-frame duration is called the first single-frame duration, and the shorter single-frame duration is called the second single-frame duration. In some scenarios, the first single-frame duration is on the order of a fraction of a second (e.g., 0.1 seconds), and the second single-frame duration is on the order of a fraction of a second (e.g., 0.025 seconds).

[0054] If the ECT image reconstructed from the projection data of the first single frame duration is called the first ECT image, then the ECT image reconstructed from the projection data of the second single frame duration is called the second ECT image. For example, if the first single frame duration is 0.1 seconds and the second single frame duration is 0.025 seconds, then the first ECT image is obtained by reconstructing the projection data of 0.1 seconds, and the second ECT image is obtained by reconstructing the projection data of 0.025 seconds. The temporal resolution of the first ECT image sequence is 0.1 seconds / frame, and the temporal resolution of the second ECT image sequence is 0.025 seconds / frame. The temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

[0055] After obtaining raw data for a set duration (e.g., one hour) through ECT scanning, the raw data can be processed to obtain projected data for the set duration. The projected data for the target time period is then determined from the projected data for the set duration. The time period of interest within the set duration can be used as the target time period; for example, if the 40th to 42nd second of the set duration is the period of interest, then the 40th to 42nd second will be used as the target time period.

[0056] In the following example, the temporal resolution of the first ECT image sequence is 0.1 seconds / frame, the temporal resolution of the second ECT image sequence is 0.025 seconds / frame, and the target time period is from the 40th to the 42nd second of the set duration.

[0057] After obtaining the projection data from second 40 to second 42, this data can be segmented at 0.1-second intervals to obtain 20 0.1-second projection data segments. These 20 0.1-second projection data segments are then reconstructed to obtain 20 first ECT images. These first ECT images have their own timestamps; for example, the timestamp of the first first ECT image corresponds to second 40 to second 40.1, and the timestamp of the second first ECT image corresponds to second 40.1 to second 40.2. The first ECT image sequence can be formed according to the temporal order of their timestamps.

[0058] Step S102: Input the first ECT image sequence into the pre-constructed intermediate frame prediction network to obtain the estimated second ECT image sequence.

[0059] As in the previous example, the temporal resolution of the first ECT image sequence is 0.1 seconds / frame, and the temporal resolution of the second ECT image sequence is 0.025 seconds / frame. The target time period is from the 40th to the 42nd second of the set duration. In this example, since the temporal resolution of the second ECT image sequence is 0.025 seconds / frame, the number of frames of the second ECT image within the 40th to the 42nd second period is 80.

[0060] The first ECT image sequence from 40 seconds to 42 seconds includes 20 first ECT images.

[0061] In some embodiments, the first ECT image sequence can be directly input into a pre-constructed intermediate frame prediction network. Based on the output of the intermediate frame prediction network, the first ECT image sequence is subjected to frame interpolation to obtain 80 second ECT images between the 40th and 42nd seconds.

[0062] In other possible embodiments, the first ECT image sequence can be first divided into blocks to obtain multiple first ECT image block sequences. All first ECT image block sequences are traversed, and for each first ECT image block sequence, the difference value within adjacent image block groups is calculated. If the difference value of a certain first ECT image block sequence is higher than a set threshold, the first ECT image block sequence is input into a pre-constructed intermediate frame prediction network. Based on the output of the intermediate frame prediction network, frame interpolation is performed on the first ECT image block sequence to obtain an interpolated first ECT image block sequence. If the difference value of the first ECT image block sequence is not higher than the threshold, it indicates that the tracer distribution has not changed between adjacent time frames within the first ECT image block, and the first ECT image block can be copied to obtain a copied image block. The copied image block is then inserted between adjacent image blocks in the first ECT image block sequence to obtain an interpolated first ECT image block sequence. Next, the interpolated first ECT image block sequences obtained according to the above operation process are arranged and integrated in sequence to obtain 80 second ECT images from the 40th to the 42nd second.

[0063] The 80 obtained second ECT images are estimated images, and their accuracy may not yet meet requirements. Therefore, these 80 second ECT images can be called estimated second ECT images. Each of the 80 estimated second ECT images has its own timestamp. Based on the chronological order of their timestamps, the 80 estimated second ECT images can form an estimated second ECT image sequence. In the estimated second ECT image sequence, the timestamp of the first estimated second ECT image corresponds to seconds 40 to 40.025, and the timestamp of the last estimated second ECT image corresponds to seconds 41.975 to 42.

[0064] Step S103: Using the estimated second ECT image sequence as the prior image, the target second ECT image sequence is reconstructed.

[0065] Using the previous example, the estimated second ECT image sequence from 40 seconds to 42 seconds can be obtained in the manner described above. This estimated second ECT image sequence can more fully reflect the subtle changes in the concentration distribution of the radioactive tracer at high temporal resolution. Therefore, the estimated second ECT image sequence can be used as a priori images for reconstruction, resulting in 80 second ECT images from 40 seconds to 42 seconds. The accuracy of these second ECT images is significantly improved compared to the estimated second ECT images; therefore, these second ECT images can be called target second ECT images. Based on the temporal order of the timestamps of the 80 target second ECT images, a target second ECT image sequence can be formed. In the target second ECT image sequence, the timestamp of the first target second ECT image corresponds to 40 seconds to 40.025 seconds, and the timestamp of the last target second ECT image corresponds to 41.975 seconds to 42 seconds. The target second ECT image sequence corresponds to 40 seconds to 42 seconds.

[0066] In the above-mentioned ECT image reconstruction method, the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence; this application obtains the first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device; the first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain the estimated second ECT image sequence; the estimated second ECT image sequence can more fully reflect the subtle changes in the concentration distribution of radioactive tracers at high temporal resolution, providing better prior guidance for reconstructing the target second ECT image sequence, helping to reduce the noise level of the target second ECT image sequence and improve image quality.

[0067] In one embodiment, inputting a first ECT image sequence into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence includes: dividing the first ECT image sequence into multiple image groups; several first ECT images in the same image group are consecutive first ECT images; inputting the multiple image groups into the intermediate frame prediction network respectively to obtain intermediate frame sub-sequences corresponding to each image group; obtaining intermediate frame sequences based on the intermediate frame sub-sequences corresponding to each image group; and inserting intermediate frame sequences into the first ECT image sequence to obtain the estimated second ECT image sequence.

[0068] During training, the intermediate frame prediction network determines the number of intermediate frames to be predicted for a group of images by obtaining the difference between the number of first ECT images and the number of second ECT images in the target time period, dividing this difference by the number of first ECT images in the target time period, and using the result as the number of intermediate frames to be predicted. The number of intermediate frames to be predicted for a group of images is also the number of images included in the intermediate frame subsequence corresponding to a single group of images.

[0069] Using the previous example, the temporal resolution of the first ECT image sequence is 0.1 seconds / frame, the temporal resolution of the second ECT image sequence is 0.025 seconds / frame, and the target time period is from the 40th to the 42nd second of the set duration.

[0070] The first ECT image sequence from 40 seconds to 42 seconds consists of 20 first ECT images, which can be denoted as {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20}. The temporal resolution of the second ECT image sequence is 0.025 seconds / frame; therefore, the number of frames in the second ECT images from 40 seconds to 42 seconds is 80. The difference between the number of first ECT images and the number of second ECT images from 40 seconds to 42 seconds is determined to be 60. Dividing 60 by the number of first ECT images from 40 seconds to 42 seconds (i.e., 20) yields a division result of 3. Therefore, the intermediate frame prediction network needs to predict 3 intermediate frames for one image group.

[0071] The group of images used as input data for the intermediate frame prediction network includes at least two images.

[0072] Example 1: If the image group contains 4 images, the following processing can be performed:

[0073] Acquire projection data from the 39.9th to the 40th second within a set time period, and reconstruct the first ECT image based on the projection data from the 39.9th to the 40th second; the timestamp of the first ECT image corresponds to the 39.9th to the 40th second. If the first ECT image is not within the target time period from the 40th to the 42nd second, it can be recorded as I_0.

[0074] Acquire projection data from the 42nd to the 42nd second within a set time period, and reconstruct the first ECT image based on the projection data from the 42nd to the 42nd second. The timestamp of the first ECT image corresponds to the 42nd to the 42nd second. The first ECT image is not within the target time period from the 40th to the 42nd second, and can be denoted as I_21.

[0075] Acquire projection data from 42.1 to 42.2 seconds within a set time period, and reconstruct the first ECT image based on the projection data from 42.1 to 42.2 seconds; the timestamp of the first ECT image corresponds to 42.1 to 42.2 seconds. The first ECT image is not within the target time period from 40 to 42 seconds, and can be denoted as I_22.

[0076] Based on the first ECT image sequence {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20} from 40 seconds to 42 seconds, the first ECT image Ⅰ_0 corresponding to 39.9 seconds to 40 seconds, the first ECT image Ⅰ_21 corresponding to 42 seconds to 42.1 seconds, and the first ECT image Ⅰ_22 corresponding to 42.1 seconds to 42.2 seconds, the prediction image sequence {Ⅰ_0,Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20,Ⅰ_21,Ⅰ_22} can be obtained. Taking four adjacent first ECT images in the prediction image sequence as a group, 20 image groups can be obtained, such as {I_0,I_1,I_2,I_3}, {I_1,I_2,I_3,I_4},……,{I_19,I_20,I_21,I_22}.

[0077] For each image group, the image group is input into the intermediate frame prediction network to obtain three intermediate frames corresponding to the image group. The timestamps of these three intermediate frames are located between the two middle first ECT images in the image group. For example, if the image group includes I_0, I_1, I_2 and I_3, after inputting the image group into the intermediate frame prediction network, three intermediate frames can be obtained. The timestamps of these three intermediate frames are located between I_1 and I_2. These three intermediate frames can be denoted as I_1a, I_1b and I_1c. That is, the intermediate frame subsequence corresponding to the image group {I_0, I_1, I_2, I_3} is {I_1a, I_1b, I_1c}.

[0078] In this way, the intermediate frame subsequences corresponding to other image groups can be obtained. For example, the intermediate frame subsequence corresponding to image group {Ⅰ_1,Ⅰ_2,Ⅰ_3,Ⅰ_4} is {Ⅰ_2a,Ⅰ_2b,Ⅰ_2c}, and the intermediate frame subsequence corresponding to image group {Ⅰ_2,Ⅰ_3,Ⅰ_4,Ⅰ_5} is {Ⅰ_3a,Ⅰ_3b,Ⅰ_3c}. The intermediate frame subsequences corresponding to 20 image groups can form the intermediate frame sequence {Ⅰ_1a,Ⅰ_1b,Ⅰ_1c,Ⅰ_2a,Ⅰ_2b,Ⅰ_2c,Ⅰ_3a,Ⅰ_3b,Ⅰ_3c,......,Ⅰ_20a,Ⅰ_20b,Ⅰ_20c}.

[0079] Based on the timestamp of each intermediate frame, the intermediate frame sequence can be inserted into the first ECT image sequence {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20} from 40 seconds to 42 seconds, resulting in {Ⅰ_1,Ⅰ_1a,Ⅰ_1b,Ⅰ_1c,Ⅰ_2,Ⅰ_2a,Ⅰ_2b,Ⅰ_2c,Ⅰ_3,......,Ⅰ_20,Ⅰ_20a,Ⅰ_20b,Ⅰ_20c}. The resulting sequence is then used as the estimated second ECT image sequence from 40 seconds to 42 seconds.

[0080] Example 2: If the image group contains 2 images, the following processing can be performed:

[0081] Acquire projection data from the 42nd to the 42nd second within a set time period, and reconstruct the first ECT image based on the projection data from the 42nd to the 42nd second. The timestamp of the first ECT image corresponds to the 42nd to the 42nd second. The first ECT image is not within the target time period from the 40th to the 42nd second, and can be denoted as I_21.

[0082] Based on the first ECT image sequence {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20} from 40 seconds to 42 seconds and the first ECT image Ⅰ_21 corresponding to 42 seconds to 42.1 seconds, the prediction image sequence {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20,Ⅰ_21} can be obtained. Taking two adjacent first ECT images in the prediction image sequence as a group, 20 image groups can be obtained.

[0083] For each image group, the image group is input into the intermediate frame prediction network to obtain three intermediate frames corresponding to the image group. The timestamps of these three intermediate frames are located between the two first ECT images in the image group. For example, if the image group includes I_1 and I_2, after inputting the image group into the intermediate frame prediction network, three intermediate frames can be obtained. The timestamps of these three intermediate frames are located between I_1 and I_2. These three intermediate frames can be denoted as I_1a, I_1b, and I_1c. That is, the intermediate frame subsequence corresponding to the image group {I_1, I_2} is {I_1a, I_1b, I_1c}.

[0084] In this way, the intermediate frame subsequences corresponding to other image groups can be obtained. For example, the intermediate frame subsequences corresponding to image group {Ⅰ_2,Ⅰ_3} are {Ⅰ_2a,Ⅰ_2b,Ⅰ_2c}, and the intermediate frame subsequences corresponding to image group {Ⅰ_3,Ⅰ_4} are {Ⅰ_3a,Ⅰ_3b,Ⅰ_3c}. The intermediate frame subsequences corresponding to the 20 image groups can form the intermediate frame sequence {Ⅰ_1a,Ⅰ_1b,Ⅰ_1c,Ⅰ_2a,Ⅰ_2b,Ⅰ_2c,Ⅰ_3a,Ⅰ_3b,Ⅰ_3c,......,Ⅰ_20a,Ⅰ_20b,Ⅰ_20c}.

[0085] Based on the timestamp of each intermediate frame, the intermediate frame sequence can be inserted into the first ECT image sequence {Ⅰ_1,Ⅰ_2,Ⅰ_3,......,Ⅰ_20} from 40 seconds to 42 seconds, resulting in {Ⅰ_1,Ⅰ_1a,Ⅰ_1b,Ⅰ_1c,Ⅰ_2,Ⅰ_2a,Ⅰ_2b,Ⅰ_2c,Ⅰ_3,......,Ⅰ_20,Ⅰ_20a,Ⅰ_20b,Ⅰ_20c}. The resulting sequence is then used as the estimated second ECT image sequence from 40 seconds to 42 seconds.

[0086] In this embodiment, the first ECT image sequence is first divided into multiple image groups. Several first ECT images in the same image group are consecutive first ECT images. The more images included in the same image group, the more accurate the prediction. As a result, the estimated second ECT image sequence can better reflect the subtle changes in the concentration distribution of radioactive tracers under high temporal resolution, thereby improving the reconstruction quality.

[0087] In one embodiment, obtaining a pre-built intermediate frame prediction network includes: selecting several ECT sample images at intervals in an ECT sample image sequence to obtain an input image sequence; obtaining a training target based on the ECT sample images located between the selected ECT sample images; and obtaining an intermediate frame prediction network based on the input image sequence and the training target.

[0088] In one of the aforementioned examples, the image group used as input data for the intermediate frame prediction network contains 4 images. The intermediate frame prediction network needs to predict 3 intermediate frames from one image group, and the 3 intermediate frames output by the network are located between the two middle images in the image group. This allows for the construction of corresponding training samples, for example:

[0089] An ECT image sequence with a specific temporal resolution can be obtained, consisting of multiple consecutive ECT images. This ECT image sequence is then used as an ECT sample image sequence. If the ECT sample image sequence includes 50 ECT sample images, these 50 ECT sample images can be denoted as Frame_1, Frame_2, ..., Frame_50.

[0090] From the ECT sample image sequence, Frame_1, Frame_5, Frame_9, and Frame_13 can be selected as the input image sequence, and Frame_6, Frame_7, and Frame_8, located between Frame_5 and Frame_9, can be used as the training targets. (Refer to...) Figure 2 After inputting Frame_1, Frame_5, Frame_9, and Frame_13 into the neural network, the intermediate prediction frames output by the neural network can be obtained, which can be denoted as Frame_6', Frame_7', and Frame_8'. Then, based on the differences between Frame_6' and Frame_6, the differences between Frame_7' and Frame_7, and the differences between Frame_8' and Frame_8, the loss values ​​are obtained.

[0091] Similarly, Frame_2, Frame_6, Frame_10, and Frame_14 can be used as the input image sequence, and Frame_7, Frame_8, and Frame_9, located between Frame_6 and Frame_10, can be used as the training targets. After inputting Frame_2, Frame_6, Frame_10, and Frame_14 into the neural network, the intermediate predicted frames output by the neural network can be obtained, which can be denoted as Frame_7', Frame_8', and Frame_9'. Then, the loss value is obtained based on the differences between Frame_7' and Frame_7, the differences between Frame_8' and Frame_8, and the differences between Frame_9' and Frame_9.

[0092] Based on the obtained loss value, the parameters of the neural network are gradually adjusted until the difference between the intermediate prediction frame output by the neural network and the corresponding training target reaches the preset condition, thereby completing the training of the intermediate frame prediction network.

[0093] In this embodiment, several ECT sample images are selected at intervals in the ECT sample image sequence to obtain the input image sequence; the training target is obtained based on the ECT sample images located between the selected ECT sample images; training is performed based on the input image sequence and the training target to obtain a more accurate intermediate frame prediction network, so that the estimated second ECT image sequence can better reflect the subtle changes in the concentration distribution of radioactive tracers at high temporal resolution, thereby improving the reconstruction quality.

[0094] In one embodiment, reconstructing a target second ECT image sequence using an estimated second ECT image sequence as a priori image includes: obtaining a first spatial kernel matrix using the estimated second ECT image sequence as a priori image; obtaining a first temporal kernel matrix based on second projection data; matching the single-frame duration corresponding to the second projection data with the temporal resolution of the second ECT image sequence; and reconstructing the target second ECT image sequence based on the first spatial kernel matrix and the first temporal kernel matrix.

[0095] Using the aforementioned example, after obtaining the estimated second ECT image sequence {Ⅰ_1,Ⅰ_1a,Ⅰ_1b,Ⅰ_1c,Ⅰ_2,Ⅰ_2a,Ⅰ_2b,Ⅰ_2c,Ⅰ_3,......,Ⅰ_20,Ⅰ_20a,Ⅰ_20b,Ⅰ_20c}, the estimated second ECT image sequence can be used as the prior image. Based on the estimated second ECT image sequence, a spatial kernel matrix is ​​obtained; for distinction, this spatial kernel matrix is ​​referred to as the first spatial kernel matrix. The spatial kernel matrix is ​​used to describe the correlation between adjacent pixels.

[0096] The formula for calculating the spatial kernel matrix is ​​as follows:

[0097] ;

[0098] in, Represents the space kernel function. This represents the spatial feature vector composed of the pixel values ​​corresponding to the j-th voxel in the ECT image. These are parameters used to adjust the weight calculation. The radial Gaussian kernel function is used to calculate the distance between the feature vectors of each pair of pixels, which serves as a description of spatial correlation. To simplify the calculation, the k-nearest neighbor algorithm from machine learning can be used to search for each pixel.

[0099] As in the previous example, the temporal resolution of the second ECT image sequence is 0.025 seconds per frame. In this embodiment, the projection data from second 40 to second 42 can be segmented at 0.025-second intervals to obtain 80 projection data points. Each projection data point corresponds to a single frame duration of 0.025 seconds. These projection data points can be referred to as second projection data, and the single frame duration corresponding to the second projection data points matches the temporal resolution of the second ECT image sequence. These 80 second projection data points each have their own timestamps. The timestamp of the first second projection data point corresponds to second 40 to second 40.025 seconds, and the timestamp of the second second projection data point corresponds to second 40.025 to second 40.05 seconds. Based on these 80 second projection data points, a temporal kernel matrix can be obtained; for distinction, this temporal kernel matrix is ​​referred to as the first temporal kernel matrix. The temporal kernel matrix is ​​used to describe the correlation of the same pixel at different times.

[0100] The formula for calculating the time kernel matrix is ​​as follows:

[0101] ;

[0102] in, Represents the time kernel function, This represents the temporal feature vector corresponding to the m-th frame. These are parameters for adjusting the weight calculation. It is the midpoint of the time of the m-th frame. Represents the width of the time window.

[0103] After obtaining the first spatial kernel matrix and the first temporal kernel matrix, the first spatiotemporal kernel matrix can be obtained based on the first spatial kernel matrix and the first temporal kernel matrix.

[0104] The spatiotemporal kernel matrix is ​​obtained from the spatial kernel matrix and the temporal kernel matrix, and the corresponding formula is as follows:

[0105]

[0106] in, ; Represents the space kernel matrix; Represents the time kernel matrix; This represents the Kronecker product.

[0107] After obtaining the first spatiotemporal kernel matrix, iterative reconstruction can be performed using the EM algorithm to obtain the second ECT image sequence of the target from second 40 to second 42. EM stands for Expectation-maximization algorithm.

[0108] In the PET image reconstruction scenario, the ECT image includes the PET image, and the update formula used in the EM algorithm is as follows:

[0109]

[0110] Where and respectively represent the iterative results of the coefficient images at the nth and (n + 1)th times, 𝑇 represents the transpose of the matrix, 𝑦 is the projection data collected by the system, 𝑃 is the system matrix, 𝐴 is the injection dose, 𝑆 is the system sensitivity, and 𝑟 is the expectation of random events and scatter events. <00​​​​​​​​​​​​​​​​​​​In this embodiment, estimating the second ECT image sequence can more fully reflect the subtle changes in the concentration distribution of radioactive tracers at high temporal resolution. The estimated second ECT image sequence is used as the prior image, and the prior image is used as the initial image for iteration. The initial image for iteration is input into the reconstruction algorithm, and the target second ECT image sequence with relatively low noise level can be reconstructed without generating temporal and spatial kernel matrices, thereby improving image quality.

[0118] In one embodiment, the method provided by this application further includes: acquiring a third ECT image sequence; the temporal resolution of the third ECT image sequence is lower than that of the first ECT image sequence; reconstructing the first ECT image sequence based on the third ECT image sequence and the first projection data; the single frame duration corresponding to the first projection data matches the temporal resolution of the first ECT image sequence.

[0119] The terms "third ECT image," "first ECT image," and "second ECT image" are relative concepts. A single-frame ECT image is reconstructed from projection data, which has a corresponding single-frame duration. If a single-frame duration is longer than the aforementioned first single-frame duration, it can be called the third single-frame duration. In some scenarios, the first single-frame duration is on the order of a few tenths of a second (e.g., 0.1 seconds), the second single-frame duration is on the order of a few tenths of a second (e.g., 0.025 seconds), and the third single-frame duration is typically on the order of several minutes or more (e.g., 5 minutes, 10 minutes). If the ECT image reconstructed from the projection data of the first single-frame duration is called the first ECT image, then the ECT image reconstructed from the projection data of the third single-frame duration is called the third ECT image; correspondingly, the temporal resolution of the third ECT image sequence is lower than that of the first ECT image sequence.

[0120] For example, the duration of the first single frame is 0.1 seconds, the duration of the second single frame is 0.025 seconds, and the duration of the third single frame is several minutes or more. After obtaining the projection data for a set duration, an ECT image can be reconstructed based on the projection data from minute 0 to minute 5 within the set duration, combined with the OSEM algorithm. The single frame duration of this ECT image is 5 minutes, which belongs to the third single frame duration and can be called the third ECT image, denoted as Ⅲ_1. An ECT image can also be reconstructed using the projection data from the 5th to the 20th minute within a set time period, combined with the OSEM algorithm. The duration of a single frame of this ECT image is 15 minutes, which belongs to the third single frame duration and can be called the third ECT image, denoted as Ⅲ_2. Similarly, an ECT image can be reconstructed using the projection data from the 20th to the 40th minute within a set time period, combined with the OSEM algorithm. The duration of a single frame of this ECT image is 20 minutes, which belongs to the third single frame duration and can be called the third ECT image, denoted as Ⅲ_3. Furthermore, an ECT image can be reconstructed using the projection data from the 40th to the 60th minute within a set time period, combined with the OSEM algorithm. The duration of a single frame of this ECT image is 20 minutes, which belongs to the third single frame duration and can be called the third ECT image, denoted as Ⅲ_4.

[0121] Based on Ⅲ_1, Ⅲ_2, Ⅲ_3, and Ⅲ_4, a third ECT image sequence {Ⅲ_1, Ⅲ_2, Ⅲ_3, Ⅲ_4} is formed.

[0122] The projection data from the 40th to the 42nd second can be divided into 20 first projection data at 0.1-second intervals, with each first projection data corresponding to a single frame duration of 0.1 seconds. These 20 first projection data have their own timestamps. For example, the timestamp of the first first projection data corresponds to the 40th to the 40th.1st second, and the timestamp of the second first projection data corresponds to the 40th.1st to the 40th.2nd second.

[0123] Based on the third ECT image sequence and 20 first projection data, the first ECT image sequence from the 40th to the 42nd second can be reconstructed.

[0124] In this embodiment, a first ECT image sequence with a low noise level can be reconstructed based on the third ECT image sequence and the first projection data.

[0125] In one embodiment, reconstructing the first ECT image sequence based on the third ECT image sequence and the first projection data includes: obtaining a second spatial kernel matrix based on the third ECT image sequence; obtaining a second temporal kernel matrix based on the first projection data; and reconstructing the first ECT image sequence based on the second spatial kernel matrix and the second temporal kernel matrix.

[0126] Based on the third ECT image sequence {Ⅲ_1,Ⅲ_2,Ⅲ_3,Ⅲ_4}, a spatial kernel matrix can be obtained; for differentiation, this spatial kernel matrix can be called the second spatial kernel matrix. Similarly, based on the 20 first projection data points from second 40 to second 42, a temporal kernel matrix can be obtained; for differentiation, this temporal kernel matrix can be called the second temporal kernel matrix. Using the second spatial kernel matrix and the second temporal kernel matrix, combined with the EM algorithm, the first ECT image sequence from second 40 to second 42 can be reconstructed.

[0127] In this embodiment, a second spatial kernel matrix is ​​obtained based on the third ECT image sequence; a second temporal kernel matrix is ​​obtained based on the first projection data; based on the second spatial kernel matrix and the second temporal kernel matrix, a high-quality first ECT image sequence can be reconstructed, so that the subsequently obtained estimated second ECT image sequence can more fully reflect the subtle changes in the concentration distribution of radioactive tracers at high temporal resolution.

[0128] To better understand the above method, the following details an application example of the ECT image reconstruction method of this application. This application example includes dividing the image into a third ECT image, a first ECT image, and a second ECT image based on the length of a single frame. The single frame duration corresponding to the third ECT image is on the order of several minutes or more, the single frame duration corresponding to the first ECT image is on the order of a fraction of a second, and the single frame duration corresponding to the second ECT image is on the order of a fraction of a second. For ease of understanding, in this embodiment, the third ECT image is referred to as a long-frame ECT image, the first ECT image as a short-frame ECT image, and the second ECT image as an ultra-short-frame ECT image. Correspondingly, the projection data whose single frame duration matches the sequence of the third ECT image is called long-frame projection data, the projection data whose single frame duration matches the sequence of the first ECT image is called short-frame projection data, and the projection data whose single frame duration matches the sequence of the second ECT image is called ultra-short-frame projection data.

[0129] Reference Figure 3After obtaining projection data for a set duration, if the set duration is one hour, the long-frame ECT image Ⅲ_1 can be reconstructed using the projection data from minute 0 to minute 5 (which belongs to long-frame projection data) combined with the OSEM algorithm; the long-frame ECT image Ⅲ_2 can be reconstructed using the projection data from minute 5 to minute 20 (which belongs to long-frame projection data) combined with the OSEM algorithm; the long-frame ECT image Ⅲ_3 can be reconstructed using the projection data from minute 20 to minute 40 (which belongs to long-frame projection data) combined with the OSEM algorithm; and the long-frame ECT image Ⅲ_4 can be reconstructed using the projection data from minute 40 to minute 60 (which belongs to long-frame projection data) combined with the OSEM algorithm. This yields the long-frame ECT image sequence {Ⅲ_1,Ⅲ_2,Ⅲ_3,Ⅲ_4}; and the second spatial kernel matrix Ks can be calculated based on the long-frame ECT image sequence {Ⅲ_1,Ⅲ_2,Ⅲ_3,Ⅲ_4}, CT images, and MR images. CT stands for Computed Tomography, while MR stands for Magnetic Resonance.

[0130] The projection data of the target time period is divided into several short frame projection data within the target time period at a time interval of 0.1 seconds. The second time kernel matrix Kt is calculated based on the several short frame projection data within the target time period.

[0131] The second spatiotemporal kernel matrix K can be obtained from the second spatial kernel matrix Ks and the second temporal kernel matrix Kt. Based on the second spatiotemporal kernel matrix K, the short-frame ECT image sequence of the target time period can be reconstructed using the EM algorithm.

[0132] Based on the short-frame ECT image sequence of the target time period and the pre-constructed intermediate frame prediction network, the estimated ultra-short-frame ECT image sequence of the target time period can be obtained.

[0133] After obtaining the estimated ultra-short frame ECT image sequence for the target time period, two methods can be used to obtain the target ultra-short frame ECT image sequence for the target time period.

[0134] Figure 4 The first method is shown. (See reference.) Figure 4Based on the estimated ultra-short frame ECT image sequence for the target time period, the first spatial kernel matrix Ks' can be calculated. The projection data within the target time period is segmented at 0.025-second intervals to obtain several ultra-short frame projection data points. Based on these ultra-short frame projection data points, the first temporal kernel matrix Kt' is calculated. From the first spatial kernel matrix Ks' and the first temporal kernel matrix Kt', the first spatiotemporal kernel matrix K' can be obtained. Using the first spatiotemporal kernel matrix K' and the EM algorithm, the target ultra-short frame ECT image sequence for the target time period can be reconstructed.

[0135] Figure 5 The second method is shown. (See reference.) Figure 5 The projection data within the target time period can be segmented at 0.025-second intervals to obtain several ultra-short frame projection data within the target time period. Based on the several ultra-short frame projection data within the target time period and the estimated ultra-short frame ECT image sequence, the target ultra-short frame ECT image sequence of the target time period can be reconstructed by combining the OSEM algorithm.

[0136] This embodiment improves the temporal resolution of ECT image sequences by using intermediate frames generated by an AI network, such as the Intermediate Frame Prediction Network, as one of the prior images, achieving high temporal resolution image reconstruction of less than 0.1 s / frame. The introduction of these prior images helps reduce noise and improve image quality. Furthermore, this embodiment combines the prior images generated by the Intermediate Frame Prediction Network with non-deep learning reconstruction methods, fully leveraging the stability of non-deep learning reconstruction methods and the richer information provided by the prior images generated by the Intermediate Frame Prediction Network, thereby achieving higher quality image reconstruction. Additionally, the training set for the Intermediate Frame Prediction Network can be pre-trained using natural images, and the final reconstruction uses a non-deep learning reconstruction method; therefore, it maintains high performance and stability when processing projection data of different nuclides.

[0137] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0138] Based on the same inventive concept, this application also provides an ECT image reconstruction apparatus for implementing the ECT image reconstruction method described above. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in one or more ECT image reconstruction apparatus embodiments provided below can be found in the limitations of the ECT image reconstruction method described above, and will not be repeated here.

[0139] In one embodiment, such as Figure 6 As shown, an apparatus for reconstructing ECT images is provided, comprising:

[0140] The image acquisition module 601 is used to acquire a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device.

[0141] The intermediate frame prediction module 602 is used to input the first ECT image sequence into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence;

[0142] The reconstruction processing module 603 is used to reconstruct the target second ECT image sequence using the estimated second ECT image sequence as a priori image; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than the temporal resolution of the first ECT image sequence.

[0143] In one embodiment, the intermediate frame prediction module 602 is further configured to:

[0144] The first ECT image sequence is divided into multiple image groups; several first ECT images in the same image group are consecutive first ECT images; multiple image groups are respectively input into the intermediate frame prediction network to obtain intermediate frame sub-sequences corresponding to each image group; based on the intermediate frame sub-sequences corresponding to each image group, an intermediate frame sequence is obtained; the intermediate frame sequence is inserted into the first ECT image sequence to obtain an estimated second ECT image sequence.

[0145] In one embodiment, the apparatus further includes a model training module for:

[0146] In the ECT sample image sequence, several ECT sample images are selected at intervals to obtain the input image sequence; the training target is obtained based on the ECT sample images located between the selected ECT sample images; and the intermediate frame prediction network is obtained based on the input image sequence and the training target.

[0147] In one embodiment, the reconstruction processing module 603 is further configured to:

[0148] Using the estimated second ECT image sequence as a priori images, a first spatial kernel matrix is ​​obtained; a first temporal kernel matrix is ​​obtained based on the second projection data; the single-frame duration corresponding to the second projection data matches the temporal resolution of the second ECT image sequence; and the target second ECT image sequence is reconstructed based on the first spatial kernel matrix and the first temporal kernel matrix.

[0149] In one embodiment, the reconstruction processing module 603 is further configured to:

[0150] Using the estimated second ECT image sequence as the prior image, the prior image is used as the initial image for iteration; the initial image for iteration is input into the reconstruction algorithm to obtain the target second ECT image sequence.

[0151] In one embodiment, the reconstruction processing module 603 is further configured to:

[0152] A third ECT image sequence is acquired; the temporal resolution of the third ECT image sequence is lower than that of the first ECT image sequence; the first ECT image sequence is reconstructed based on the third ECT image sequence and the first projection data; the single frame duration corresponding to the first projection data matches the temporal resolution of the first ECT image sequence.

[0153] In one embodiment, the reconstruction processing module 603 is further configured to:

[0154] Based on the third ECT image sequence, a second spatial kernel matrix is ​​obtained; based on the first projection data, a second temporal kernel matrix is ​​obtained; based on the second spatial kernel matrix and the second temporal kernel matrix, the first ECT image sequence is reconstructed.

[0155] The modules in the aforementioned ECT image reconstruction device 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 corresponding operations of each module.

[0156] In one exemplary embodiment, a computer device is provided, the internal structure of which can be as shown in the figure. Figure 7As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores the data involved in the aforementioned methods. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for reconstructing ECT images.

[0157] Those skilled in the art will understand that Figure 7 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.

[0158] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps in the various method embodiments described above.

[0159] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the various method embodiments described above.

[0160] In one embodiment, a computer program product is provided having a computer program stored thereon, the computer program being executed by a processor of the steps described in the various method embodiments above.

[0161] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0162] 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 above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0163] 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 application.

[0164] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. 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 application should be determined by the appended claims.

Claims

1. A method for reconstructing ECT images, characterized in that, The method includes: Acquire a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device; The first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence; Using the estimated second ECT image sequence as a priori images, a target second ECT image sequence is reconstructed; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

2. The method of claim 1, wherein, The first ECT image sequence is input into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence, including: The first ECT image sequence is divided into multiple image groups; several first ECT images in the same image group are consecutive first ECT images. The multiple image groups are respectively input into the intermediate frame prediction network to obtain the intermediate frame subsequence corresponding to each image group; The intermediate frame sequence is obtained based on the intermediate frame sub-sequence corresponding to each of the image groups; The intermediate frame sequence is inserted into the first ECT image sequence to obtain the estimated second ECT image sequence.

3. The method of claim 1, wherein, Obtain the pre-built intermediate frame prediction network, including: In the ECT sample image sequence, several ECT sample images are selected at intervals to obtain the input image sequence; The training target is obtained from the ECT sample images located between the selected ECT sample images; Based on the input image sequence and the training objective, an intermediate frame prediction network is obtained.

4. The method of claim 1, wherein, Using the estimated second ECT image sequence as a priori images, the target second ECT image sequence is reconstructed, including: Using the estimated second ECT image sequence as the prior image, the first spatial kernel matrix is ​​obtained; The first time kernel matrix is ​​obtained based on the second projection data; the single frame duration corresponding to the second projection data matches the temporal resolution of the second ECT image sequence. Based on the first spatial kernel matrix and the first temporal kernel matrix, the target second ECT image sequence is reconstructed.

5. The method of claim 1, wherein, Using the estimated second ECT image sequence as a priori images, the target second ECT image sequence is reconstructed, including: Using the estimated second ECT image sequence as the prior image, the prior image is used as the initial image for iteration; The iterative initial image is input into the reconstruction algorithm to obtain the target second ECT image sequence.

6. The method of claim 1, wherein, The method further includes: Acquire a third ECT image sequence; the temporal resolution of the third ECT image sequence is lower than that of the first ECT image sequence. Based on the third ECT image sequence and the first projection data, the first ECT image sequence is reconstructed; the single frame duration corresponding to the first projection data matches the temporal resolution of the first ECT image sequence.

7. The method of claim 6, wherein, Based on the third ECT image sequence and the first projection data, the first ECT image sequence is reconstructed, including: The second spatial kernel matrix is ​​obtained based on the third ECT image sequence; Based on the first projection data, the second time kernel matrix is ​​obtained; The first ECT image sequence is reconstructed based on the second spatial kernel matrix and the second temporal kernel matrix.

8. An apparatus for reconstruction of an ECT image, characterized by The device includes: An image acquisition module is used to acquire a first ECT image sequence; the first ECT image sequence is obtained by scanning with an ECT scanning device. The intermediate frame prediction module is used to input the first ECT image sequence into a pre-constructed intermediate frame prediction network to obtain an estimated second ECT image sequence; The reconstruction processing module is used to reconstruct the target second ECT image sequence using the estimated second ECT image sequence as a priori images; the estimated second ECT image sequence and the target second ECT image sequence belong to the second ECT image sequence; the temporal resolution of the second ECT image sequence is higher than that of the first ECT image sequence.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements 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, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images