Image reconstruction method and CT imaging system

By dimensional downsampling and low-frequency filtering of detector data in CT imaging, the problem of CT image uniformity is solved, and more efficient image correction and quality improvement is achieved.

CN119048621APending Publication Date: 2024-11-29SHANGHAI UNITED IMAGING HEALTHCARE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411061950.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

In CT imaging, image reconstruction deviations and low-frequency ring band artifacts caused by low signal regions affect the uniformity of the image, and the existing correction methods are not effective.

Method used

By dimensional downsampling and low-frequency filtering of the data collected by the detector, the low-frequency components of the original data are replaced, and the statistics of the detector unit after the data is merged are improved, and image uniformity correction is achieved.

Benefits of technology

Improves uniformity of CT images, eliminates low-frequency ring band artifacts, enhances image accuracy and quality, and is suitable for various scanning scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119048621A_ABST
    Figure CN119048621A_ABST
Patent Text Reader

Abstract

The invention relates to an image reconstruction method and a CT imaging system. The method comprises the following steps: acquiring original scanning data of a target object; performing dimension downsampling processing on the original scanning data to obtain intermediate scanning data; the dimension downsampling processing comprises the step of carrying out addition processing on data in a preset merging range in the original scanning data; extracting low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data, and performing image reconstruction based on the intermediate low-frequency information and the original scanning data to obtain a target reconstruction image of the target object; the low-frequency information includes a low-frequency image or low-frequency data. According to the method, the statistical magnitude of a detector unit after data merging is improved by performing dimension downsampling on projection domain data collected by a detector; the method does not depend on interpolation prediction or model prediction and other external information, and can ensure the accuracy of subsequent correction and reconstruction, thereby improving the correction effect of the non-uniform part in the image, and improving the uniformity of the image.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of medical imaging technology, and in particular to an image reconstruction method and a CT imaging system. Background Art

[0002] During the process of computed tomography (CT) scanning, when X-rays pass through different substances, there will be different degrees of attenuation. When the attenuation is large, the signal received by the detector will be very small, or even close to 0. For this type of signal, it is very likely to cause deviations in the subsequent correction and reconstruction algorithms, thereby generating low-frequency ring artifacts in the reconstructed image, affecting the uniformity of the CT reconstructed image.

[0003] Traditionally, methods for correcting the uniformity of CT reconstructed images include: in the image reconstruction process, using the adjacent channel interpolation method to predict and correct this part of the signal, or predicting and correcting the signal based on external reference information or a physical model, so that when the image is reconstructed based on the corrected signal, the uniformity of the reconstructed image can be improved.

[0004] However, the uniformity of the CT reconstructed image obtained by using the traditional image uniformity correction processing method is still not good. Summary of the invention

[0005] Based on this, it is necessary to provide an image reconstruction method, apparatus, CT imaging system, computer equipment, computer-readable storage medium and computer program product that can improve the uniformity of the reconstructed image in response to the above technical problems.

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

[0007] Obtaining raw scan data of the target object;

[0008] Performing dimensionality downsampling processing on the original scan data to obtain intermediate scan data; the dimensionality downsampling processing includes summing the data in the original scan data that is within a preset merging range;

[0009] Extracting low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; the low-frequency information includes a low-frequency image or low-frequency data;

[0010] Image reconstruction is performed based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object.

[0011] In one embodiment, the original scanning data includes original scanning data of multiple scanning angles collected by the medical scanning device during the rotation scanning process, and the intermediate scanning data includes intermediate scanning data of each scanning angle; the original scanning data is subjected to dimensionality downsampling processing to obtain the intermediate scanning data, including:

[0012] For the original scan data at each scanning angle, in the merging area of ​​the original scan data, multiple data within a preset merging range are summed to obtain merged data;

[0013] The original scan data is scan data in a non-merge area, and the scan data in the non-merge area is combined with the merged data as intermediate scan data corresponding to the scan angle.

[0014] In one of the embodiments, when the dimensionality downsampling process is performed along the channel direction of the detector in the medical scanning device, the merged area is a partial area along the channel direction in the original scanning data;

[0015] When the dimensionality downsampling process is performed along the row direction of the detector, the merged area is a partial area along the row direction in the original scan data.

[0016] In one embodiment, the low-frequency information is a low-frequency image; extracting the low-frequency information based on the intermediate scanning data to obtain the intermediate low-frequency information corresponding to the intermediate scanning data includes:

[0017] Perform image reconstruction based on the intermediate scanning data to obtain an intermediate reconstructed image;

[0018] The low-frequency information is extracted from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

[0019] In one embodiment, performing image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object includes:

[0020] Perform image reconstruction based on the original scan data to obtain an original reconstructed image;

[0021] Extract low-frequency information from the original reconstructed image to obtain an original low-frequency image corresponding to the original scan data;

[0022] The original low-frequency image is replaced by the intermediate low-frequency image in the original reconstructed image to obtain a target reconstructed image of the target object.

[0023] In one embodiment, if the size of the intermediate reconstructed image is smaller than the size of the original reconstructed image, extracting low-frequency information from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data includes:

[0024] Extract low-frequency information from the intermediate reconstructed image to obtain a candidate low-frequency image;

[0025] The candidate low-frequency images are interpolated to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

[0026] In one embodiment, the low-frequency information is low-frequency data; extracting the low-frequency information based on the intermediate scanning data to obtain the intermediate low-frequency information corresponding to the intermediate scanning data includes:

[0027] The low-frequency information is extracted from the intermediate scanning data to obtain intermediate low-frequency data corresponding to the intermediate scanning data.

[0028] In one embodiment, performing image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object includes:

[0029] Extract low-frequency information from the original scanning data to obtain original low-frequency data corresponding to the original scanning data;

[0030] The original low-frequency data in the original scanning data is replaced by the intermediate low-frequency data to obtain the target scanning data;

[0031] Image reconstruction is performed based on the target scanning data to obtain a target reconstructed image of the target object.

[0032] In a second aspect, the present application further provides an image reconstruction device, comprising:

[0033] An acquisition module, used for acquiring original scan data of a target object;

[0034] A processing module is used to perform dimensionality downsampling processing on the original scan data to obtain intermediate scan data; the dimensionality downsampling processing includes adding data in the original scan data that is within a preset merging range;

[0035] An extraction module, used to extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; the low-frequency information includes a low-frequency image or low-frequency data;

[0036] The reconstruction module is used to reconstruct the image based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object.

[0037] In a third aspect, the present application further provides a CT imaging system, including a photon counting detector array, a processing circuit, and a reconstruction system, characterized in that the photon counting detector array includes a plurality of voxels arranged along a channel direction and a row direction, the plurality of voxels are divided into a plurality of sub-box voxels along the channel direction and / or the row direction, and each sub-box voxel includes a plurality of voxels;

[0038] A voxel is used to obtain the energy of photon information and generate a detection signal when a CT scan is performed on a target object;

[0039] A processing circuit, for acquiring raw scan data of the target object according to the detection signal generated by each voxel; the raw scan data includes data corresponding to each voxel;

[0040] The processing circuit is also used to obtain intermediate scanning data according to the original scanning data; the intermediate scanning data includes data corresponding to each binned voxel, and the data corresponding to the binned voxel is obtained by adding the data of each voxel corresponding to the binned voxel in the original scanning data;

[0041] The reconstruction system is used to extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; and to reconstruct an image based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object; the low-frequency information includes a low-frequency image or low-frequency data.

[0042] In one embodiment, the binned voxels are an m×n voxel array; wherein m=1 and n≥2; or, m≥2 and n=1; or, m≥2 and n≥2;

[0043] There are no overlapping voxels between the voxels in each bin, or there are overlapping voxels.

[0044] In a fourth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the image reconstruction method in the first aspect when executing the computer program.

[0045] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the image reconstruction method in the first aspect are implemented.

[0046] In a sixth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the image reconstruction method in the first aspect.

[0047] The above-mentioned image reconstruction method, device, CT imaging system, computer equipment, storage medium and computer program product, the computer equipment obtains the original scanning data of the target object and performs dimensionality downsampling processing on the original scanning data to obtain intermediate scanning data; wherein the dimensionality downsampling processing includes summing the data in the original scanning data that are within a preset merging range; then, based on the intermediate scanning data, low-frequency information is extracted to obtain intermediate low-frequency information corresponding to the intermediate scanning data, and image reconstruction is performed based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object; wherein the low-frequency information includes a low-frequency image or low-frequency data. That is to say, in the image reconstruction method proposed in the embodiment of the present application, the statistics of the detector unit after data merging are improved by downsampling the dimension of the data in the projection domain collected by the detector; that is, in the present application, the data at the uneven location are statistically fused to increase the statistics at the uneven location, and the statistics are obtained by merging based on the real collected data, rather than the traditional method of increasing the statistics at the uneven location through interpolation prediction, or prediction based on external reference information or physical model; in other words, the method proposed in the present embodiment merges and downsamples the data collected by the detector, without relying on external information such as interpolation prediction or model prediction, so as to ensure the accuracy of subsequent correction and reconstruction, thereby improving the correction effect of the uneven locations in the image and improving the uniformity of the image; in addition, the method proposed in the embodiment of the present application can be generally applied to image reconstruction of various scanning scenarios, with high universality and a wide range of applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A diagram showing an application environment of an image reconstruction method in an embodiment;

[0050] Figure 2 is a schematic flow chart of an image reconstruction method in one embodiment;

[0051] FIG3( a ) is a schematic diagram of dimensional downsampling in a channel direction in one embodiment;

[0052] FIG3( b ) is a schematic diagram of dimensional downsampling in the row direction in one embodiment;

[0053] Figure 4 is a schematic flow chart of an image reconstruction method in another embodiment;

[0054] Figure 5 is a schematic flow chart of an image reconstruction method in another embodiment;

[0055] Figure 6 is a schematic flow chart of an image reconstruction method in another embodiment;

[0056] Figure 7 is a schematic diagram of a process of performing uniformity correction based on an image domain in one embodiment;

[0057] FIG8( a ) is a schematic diagram of an original reconstructed image after conventional correction and reconstruction in one embodiment;

[0058] FIG8( b ) is a schematic diagram of an original low-frequency image after low-frequency filtering in one embodiment;

[0059] FIG9( a ) is a schematic diagram of an intermediate reconstructed image after dimensionality downsampling in one embodiment;

[0060] FIG9( b ) is a schematic diagram of an intermediate low-frequency image after dimensional downsampling in one embodiment;

[0061] Fig.10 is a schematic diagram of a target reconstructed image after low-frequency correction in one embodiment;

[0062] Fig.11 It is a schematic diagram of a process of performing uniformity correction based on a projection domain in one embodiment;

[0063] Fig.12 is a structural block diagram of an image reconstruction device in one embodiment;

[0064] Fig.13 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] During CT scanning, when the signal received by the detector is very small or close to 0, it may cause deviations in the subsequent correction and reconstruction algorithms, thereby generating low-frequency ring artifacts in the reconstructed image, affecting the uniformity of the CT image. Currently, the correction methods for image inhomogeneity mostly use adjacent channel interpolation methods for prediction, or algorithms based on external information or physical models to predict and correct inhomogeneities. There is currently no similar method that uses dimensional downsampling and low-frequency filtering.

[0067] Moreover, the correction method of adjacent channel interpolation has a better interpolation effect when the number of invalid channels corresponding to the unevenness is small, that is, the uniformity correction effect is better; while when the number of adjacent invalid channels corresponding to the unevenness is large, the interpolation effect of the invalid channels adjacent to the valid channels at the uniformity is better, while the interpolation effect of the invalid channels far away from the valid channels is not good, and some artifacts still exist in the reconstructed image, resulting in poor uniformity correction effect.

[0068] The correction method based on external information or physical model requires additional external information or pre-prediction of the physical model, and theoretical prediction through reference data or physical model to correct the unevenness. It requires large amount of calculation and low processing efficiency.

[0069] Based on this, an embodiment of the present application proposes an image reconstruction method, which downsamples the projection domain data to improve the statistics of the detector unit after data merging, thereby ensuring the accuracy of subsequent correction and reconstruction algorithms, and then performs low-frequency filtering on the merged data to replace the low-frequency components of the original data, ultimately achieving the purpose of correcting the image uniformity and improving the uniformity of the reconstructed image.

[0070] In addition, the method of the present application is also applicable to the correction and reconstruction process of photon counting computed tomography (PCCT) images. Due to the low-dose and high-resolution mode of PCCT, the material decomposition algorithm based on the maximum likelihood method under low counts cannot guarantee unbiasedness, and may even lead to ray starvation, ultimately making it impossible to perform quantitative calculation of the energy spectrum in the high-resolution mode. When the method of the present application is used for image reconstruction, dimensional downsampling and low-frequency filtering replacement can solve the image artifacts caused by ray starvation, thereby improving the uniformity of the image, solving the PCCT image reconstruction problems currently faced by the industry, and improving the image quality of PCCT in the high-resolution mode.

[0071] The image reconstruction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the computer device 102 may include a medical imaging device, and may also include a reconstruction device that is communicatively connected to the medical imaging device; illustratively, the medical imaging device may include but is not limited to a computed tomography (CT) device, a photon counting computed tomography (PCCT) device, and the like.

[0072] In addition, the reconstruction device may include but is not limited to a reconstruction terminal and a reconstruction server, and the reconstruction terminal may be but is not limited to a personal computer, a laptop, a smart phone, a tablet computer and a portable wearable device, such as a smart watch, a smart bracelet, a head-mounted device, etc. The reconstruction server may be implemented by an independent server or a server cluster composed of multiple servers, and may be a local server, a remote server, a cloud server, etc.

[0073] In an exemplary embodiment, Figure 2 As shown, an image reconstruction method is provided, which is applied to Figure 1 The computer device in the example is used to illustrate, including the following steps 202 to 206. Among them:

[0074] Step 202: Acquire original scan data of the target object.

[0075] The target object may be a living animal or human body, or a non-living model, which is used in scenarios such as research and development, testing, and maintenance. In addition, the target object may be the entire object, or a local area of ​​the object, such as a local tissue or organ. The embodiment of the present application does not specifically limit the type of the target object.

[0076] Exemplarily, a medical scan may be performed on a target object through a medical scanning device to obtain raw scan data of the target object; the raw scan data may be scan data collected by a detector in the medical scanning device, and may also be referred to as raw scan data. Exemplarily, a computer device may obtain raw scan data of a target object from a medical scanning device, or may obtain raw scan data of a target object from a picture archiving and communication system (PACS), etc. In the embodiments of the present application, the manner in which a computer device obtains raw scan data of a target object is not specifically limited.

[0077] Step 204 , downsampling the original scan data to obtain intermediate scan data.

[0078] The dimensionality downsampling process may include performing summation processing on the data in the original scan data that is within a preset merging range. Exemplarily, the preset merging range may be predetermined based on the dimensionality downsampling ratio, for example, when the dimensionality downsampling ratio is 4, the preset merging range may include 2×2, 1×4, 4×1, etc.

[0079] Exemplarily, the computer device can perform dimensionality downsampling processing on the original scan data according to the preset merging range, and perform summation processing on the data in the preset merging range in the original scan data by scanning row by row / column by column, so as to obtain the intermediate scan data after dimensionality downsampling; wherein, when performing the summation processing, multiple data in the original scan data within the preset merging range can be summed to meet the statistical requirements. For example, if the original scan data is 4×12 scan data, and the preset merging range is 2×2, that is, every 2×2 data in the 4×12 original scan data is statistically summed, the intermediate scan data after dimensionality downsampling can be obtained, that is, the intermediate scan data is 2×6 scan data.

[0080] Step 206: extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data.

[0081] The low-frequency information may include a low-frequency image or low-frequency data; that is, the intermediate low-frequency information may be an intermediate low-frequency image or intermediate low-frequency data.

[0082] Exemplarily, when intermediate scanning data after dimensional downsampling is obtained, low-frequency information can be extracted from the intermediate scanning data to obtain intermediate low-frequency data corresponding to the intermediate scanning data; alternatively, the intermediate scanning data can be first subjected to preset data processing, and then low-frequency information can be extracted from the processed intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; exemplarily, the preset data processing can include but is not limited to performing image reconstruction processing on the intermediate scanning data to obtain a reconstructed intermediate scanning image, and then low-frequency information can be extracted from the reconstructed intermediate scanning image to obtain an intermediate low-frequency image.

[0083] Step 208 : performing image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object.

[0084] Exemplarily, when image reconstruction is performed based on the intermediate low frequency information and the original scan data, the original low frequency information in the original scan data can be replaced based on the intermediate low frequency information during the image reconstruction process, thereby obtaining a target reconstructed image after the low frequency information is replaced.

[0085] Exemplarily, the computer device can replace low-frequency information in the image domain, and can also replace low-frequency information in the projection domain; for example: when replacing low-frequency information in the image domain, the computer device can first reconstruct the image based on the intermediate scanning data after dimensionality reduction processing to obtain an intermediate reconstructed image, and then extract the intermediate low-frequency information from the intermediate reconstructed image, and replace the original low-frequency information in the original reconstructed image reconstructed based on the original scanning data based on the intermediate low-frequency information, so as to obtain a target reconstructed image of the target object.

[0086] Exemplarily, when performing low-frequency information replacement in the projection domain, the computer device can first extract the intermediate low-frequency information from the intermediate scanning data after dimensionality reduction processing, and replace the original low-frequency information in the original scanning data based on the intermediate low-frequency information, and then perform image reconstruction based on the original scanning data with the low-frequency information replaced, and finally obtain the target reconstructed image of the target object.

[0087] In the above-mentioned image reconstruction method, the computer device obtains the original scanning data of the target object and performs dimensionality downsampling processing on the original scanning data to obtain intermediate scanning data; wherein the dimensionality downsampling processing includes summing the data in the original scanning data that is within a preset merging range; then, low-frequency information is extracted based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data, and image reconstruction is performed based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object; wherein the low-frequency information includes a low-frequency image or low-frequency data. That is to say, in the image reconstruction method proposed in the embodiment of the present application, the statistics of the detector unit after data merging are improved by downsampling the dimension of the data in the projection domain collected by the detector; that is, in the present application, the data at the uneven location are statistically fused to increase the statistics at the uneven location, and the statistics are obtained by merging based on the real collected data, rather than the traditional method of increasing the statistics at the uneven location through interpolation prediction, or prediction based on external reference information or physical model; in other words, the method proposed in the present embodiment merges and downsamples the data collected by the detector, without relying on external information such as interpolation prediction or model prediction, so as to ensure the accuracy of subsequent correction and reconstruction, thereby improving the correction effect of the uneven locations in the image and improving the uniformity of the image; in addition, the method proposed in the embodiment of the present application can be generally applied to image reconstruction of various scanning scenarios, with high universality and a wide range of applicability.

[0088] In an exemplary embodiment, during the process of rotational scanning, the medical scanning device usually scans the target object when rotating to different scanning angles (views). For example, taking clockwise rotation as an example, the medical scanning device can scan at multiple scanning angles, namely, 0° (the tube is vertically directly above and the detector is vertically directly below), 90° (i.e., the tube is horizontally on the right and the detector is horizontally on the left), 180° (i.e., the tube is vertically directly below and the detector is vertically directly above), and 270° (i.e., the tube is horizontally on the left and the detector is horizontally on the right).

[0089] That is, the original scan data of the target object may include the original scan data of multiple scan angles collected by the medical scanning device during the rotational scanning process; then, when performing the dimensionality downsampling process in the projection domain, it is necessary to perform dimensionality downsampling on the original scan data of each scan angle view by view, so as to obtain the intermediate scan data of each scan angle. Based on this, the above step 204 may include:

[0090] For the original scanning data at each scanning angle, multiple data in the original scanning data within a preset merging range are summed up to obtain intermediate scanning data corresponding to the scanning angle.

[0091] It should be noted that when the original scan data of each scanning angle is subjected to dimensional downsampling, the preset merging range used can be the same, so that the original scan data of each scanning angle is subjected to the same dimensional downsampling. For example, the 4×12 original scan data of each scanning angle can be subjected to dimensional downsampling according to the preset merging range of 2×2, thereby obtaining 2×6 intermediate scan data of each scanning angle.

[0092] Exemplarily, when downsampling the original scan data of each scanning angle, the original scan data of the scanning angle may be downsampled as a whole, or the data of a partial area in the original scan data may be downsampled. For example, only the uneven area in the original scan data may be downsampled. Based on this, for the original scan data of each scanning angle, the computer device may first determine the merged area of ​​the original scan data, and then, in the merged area of ​​the original scan data, add and process multiple data within a preset merged range to obtain merged data; and determine the scan data of the original scan data in the non-merged area, and combine the scan data in the non-merged area with the merged data as the intermediate scan data corresponding to the scanning angle.

[0093] That is to say, the data in the merged area of ​​the original scan data can be downsampled, while the data in the non-merged area can not be downsampled, that is, some data in the uneven area are statistically merged to increase the data statistics in the uneven area.

[0094] Exemplarily, when dimensionality downsampling is performed on a partial area in the original scanning data, in the case where dimensionality downsampling is performed along the channel direction of the detector in the medical scanning device, the merged area can be a partial area in the original scanning data along the channel direction, that is, after dimensionality downsampling in the channel direction, the length of the merged data corresponding to the row direction of the detector is equal to the length of the original scanning data corresponding to the row direction; and in the case where dimensionality downsampling is performed along the row direction of the detector, the merged area can be a partial area in the original scanning data along the row direction, that is, after dimensionality downsampling in the row direction, the length of the merged data corresponding to the channel direction of the detector is equal to the length of the original scanning data corresponding to the channel direction.

[0095] Based on the above example, assuming that the original scan data is 4×12 scan data, when the dimension downsampling is performed on part of the merged area in the channel direction, referring to FIG3(a), assuming that the data of the middle 8 channels in the channel direction (the merged area surrounded by the dotted line) are dimensionally downsampled, then the preset merge range can be 1×n, where 2≤n≤8; when the preset merge range is 1×2, that is, the data of every 1×2 in the merged area (that is, the data of the middle 8 channels) is statistically added, so that the middle scan data after the channel direction dimension downsampling can be 4×8 scan data; this is equivalent to dimensional compression of the original scan data in the channel direction, while the dimension of the row direction of the original scan data remains unchanged.

[0096] In addition, when downsampling is performed on part of the merged area in the row direction, referring to FIG3(b), assuming that the data of the two middle channels in the row direction are downsampled, the preset merge range can be m×1, where m=2; when the preset merge range is 2×1, that is, the data of every 2×1 in the merged area (i.e., the data of the two middle rows) is statistically added, so that the intermediate scan data after the row direction dimensional downsampling is 3×12 scan data; this is equivalent to dimensional compression of the original scan data in the row direction, while the dimension of the original scan data in the channel direction remains unchanged.

[0097] In this example, the original scanning data at different scanning angles collected during the rotation process need to be downsampled separately, and when performing the downsampling process, part or all of the area in the channel direction can be downsampled to achieve dimensional compression in the channel direction; part or all of the area in the row direction can be downsampled to achieve dimensional compression in the row direction; or, the original scanning data can be downsampled as a whole; the downsampling processing method is relatively flexible, and in actual application, it can be flexibly processed according to actual needs.

[0098] In an exemplary embodiment, a low-frequency information replacement process based on an image domain is provided. In the case where the low-frequency information is a low-frequency image, such as Figure 4 As shown, the above step 206 may include step 402 and step 404. Among them:

[0099] Step 402: reconstruct an image based on the intermediate scanning data to obtain an intermediate reconstructed image.

[0100] Step 404: extract low-frequency information from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scan data.

[0101] The intermediate scanning data may include the intermediate scanning data after the dimensions corresponding to each scanning angle are downsampled.

[0102] Exemplarily, the computer device can perform image reconstruction based on the intermediate scanning data of each scanning angle to obtain the intermediate reconstructed image of each scanning angle; then, a low-frequency filtering method is used to extract low-frequency information from the intermediate reconstructed image of each scanning angle to obtain the intermediate low-frequency image containing low-frequency information of each scanning angle; among them, there are many low-frequency filtering methods. For example, two-dimensional Gaussian filtering can be selected in the image domain to extract low-frequency information from the intermediate scanning image. In the embodiment of the present application, there is no specific limitation on the low-frequency filtering processing method in the image domain, and there is no specific limitation on the cutoff frequency of the low-frequency filtering. In practical applications, it can be flexibly set and adjusted according to actual needs.

[0103] Accordingly, continue as Figure 4 As shown, the above step 208 may include steps 406 to 410, wherein:

[0104] Step 406: reconstruct an image based on the original scan data to obtain an original reconstructed image.

[0105] Step 408: extract low-frequency information from the original reconstructed image to obtain an original low-frequency image corresponding to the original scan data.

[0106] Exemplarily, when low-frequency information is replaced in the image domain, the computer device can also perform image reconstruction based on the original scanning data to obtain an original reconstructed image corresponding to the original scanning data; then, low-frequency information with the same cutoff frequency is extracted from the original reconstructed image to obtain an original low-frequency image containing the low-frequency information.

[0107] Exemplarily, when performing image reconstruction, the original scanning data can be corrected first, and then the image reconstruction can be performed based on the corrected original scanning data; it should be noted that the correction here can be a conventional projection data correction method, including but not limited to air correction, attenuation correction, scattering correction, etc.

[0108] Of course, when performing image reconstruction on the intermediate scanning data, conventional correction may be performed on the intermediate scanning data first, and then image reconstruction may be performed on the corrected intermediate scanning data to obtain an intermediate reconstructed image.

[0109] Step 410: Replace the original low-frequency image with the intermediate low-frequency image in the original reconstructed image to obtain a target reconstructed image of the target object.

[0110] Exemplarily, when extracting the original low-frequency image from the original reconstructed image and the intermediate low-frequency image from the intermediate reconstructed image, the computer device can first perform image subtraction on the original reconstructed image and the original low-frequency image, that is, remove the low-frequency information from the original reconstructed image to obtain a reconstructed image containing information other than the low-frequency information; then, perform image fusion on the reconstructed image containing information other than the low-frequency information and the intermediate low-frequency image, thereby achieving an image processing effect of replacing the original low-frequency information in the original reconstructed image with the intermediate low-frequency information, and obtaining a target reconstructed image of the target object.

[0111] In this embodiment, the low-frequency information of the merged data can be replaced in the image domain. Specifically, image reconstruction is first performed based on the intermediate scan data to obtain an intermediate reconstructed image, and low-frequency information is extracted from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scan data; and image reconstruction is performed based on the original scan data to obtain an original reconstructed image, and low-frequency information is extracted from the original reconstructed image to obtain an original low-frequency image corresponding to the original scan data; and then, the intermediate low-frequency image is used to replace the original low-frequency image in the original reconstructed image to obtain a target reconstructed image of the target object. By adopting this method, the low-frequency information in the original image can be replaced in the image domain, thereby correcting the uneven low-frequency information in the image to improve the uniformity of the reconstructed image.

[0112] In an exemplary embodiment, from the above Figure 4 It can be seen from the low-frequency information replacement process based on the image domain that in the process of low-frequency information replacement in the image domain, it is necessary to perform two image reconstruction processes, namely, image reconstruction of the original scan data to obtain the original reconstructed image, and image reconstruction of the intermediate scan data to obtain the intermediate reconstructed image; since the calculation process of image reconstruction is relatively complex and the amount of calculation is large, in order to further improve the image reconstruction efficiency, the intermediate reconstruction process can also be optimized in this embodiment to reduce the amount of calculation of the intermediate reconstruction.

[0113] Exemplarily, since the amount of data of the intermediate scan data is smaller than the amount of data of the original scan data, when image reconstruction is performed based on the intermediate scan data, image reconstruction of a preset size can be performed based on the intermediate scan data to obtain an intermediate reconstructed image, wherein the size of the intermediate reconstructed image is smaller than the size of the original reconstructed image.

[0114] Exemplarily, in order to speed up reconstruction, the reconstruction matrix can be reduced for the intermediate scanning data after dimensionality downsampling, thereby obtaining an intermediate reconstructed image that is smaller in size than the original reconstructed image; for example: for a 512×512 original reconstructed image, the intermediate scanning data after dimensionality downsampling can be reconstructed into a 128×128 matrix to obtain a 128×128 intermediate reconstructed image.

[0115] Based on this, Figure 5 As shown, the above step 404 may include:

[0116] Step 502: extract low-frequency information from the intermediate reconstructed image to obtain a candidate low-frequency image.

[0117] In which, when the intermediate reconstructed image is a reconstructed image of a preset size, the candidate low-frequency image may be a low-frequency image of the preset size.

[0118] Exemplarily, when an intermediate reconstructed image of a smaller size is obtained, low-frequency filtering can be used to extract low-frequency information from the intermediate reconstructed image, thereby obtaining a candidate low-frequency image of a preset size containing low-frequency information; for example: obtaining a 128×128 candidate low-frequency image containing low-frequency information.

[0119] Step 504: perform interpolation processing on the candidate low-frequency image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

[0120] Since the original reconstructed image is 512×512, the original low-frequency image extracted from the original reconstructed image will also be 512×512. When the original low-frequency image in the original reconstructed image is replaced by the intermediate low-frequency image, it should be ensured that the size of the intermediate low-frequency image is consistent with the size of the original low-frequency image; therefore, when a candidate low-frequency image of a smaller size is obtained, the candidate low-frequency image can be processed by interpolation to obtain an intermediate low-frequency image of the same size as the original low-frequency image.

[0121] In this embodiment, when low-frequency information is replaced in the image domain, in order to speed up the image reconstruction speed, for the intermediate scan data after dimensional downsampling, its reconstruction matrix can be reduced to obtain an intermediate reconstructed image of a smaller size; then, for the candidate low-frequency image extracted from the intermediate reconstructed image of the smaller size, an intermediate low-frequency image with the same size as the original reconstructed image can be obtained by interpolation, so that the low-frequency information in the original reconstructed image can be replaced by the intermediate low-frequency image with the same size subsequently; while speeding up the reconstruction speed, image uniformity correction is achieved, which can not only improve the image reconstruction efficiency, but also improve the uniformity processing effect of image reconstruction.

[0122] In an exemplary embodiment, a low-frequency information replacement process based on a projection domain is provided. When the low-frequency information is low-frequency data, such as Figure 6 As shown, the above step 206 may include step 602. Among them:

[0123] Step 602: extract low-frequency information from the intermediate scanning data to obtain intermediate low-frequency data corresponding to the intermediate scanning data.

[0124] Exemplarily, the computer device can directly extract low-frequency information in the data domain, that is, extract low-frequency information from the intermediate scanning data, thereby obtaining intermediate low-frequency data in the intermediate scanning data. For example, a low-frequency filtering method for data can be used to perform low-frequency filtering on the intermediate scanning data, thereby obtaining intermediate low-frequency data corresponding to the intermediate scanning data.

[0125] Accordingly, continue as Figure 6 As shown, the above step 208 may include steps 604 to 608, wherein:

[0126] Step 604: extract low-frequency information from the original scan data to obtain original low-frequency data corresponding to the original scan data.

[0127] Exemplarily, the low-frequency filtering method for data in the related art can also be used to perform low-frequency filtering on the original scan data, so as to obtain the original low-frequency data corresponding to the original scan data. In addition, it should be noted that when the intermediate scan data and the original scan data are subjected to low-frequency filtering, the low-frequency cutoff frequency corresponding to the low-frequency information extracted from the two should be consistent, that is, the low-frequency information with the same cutoff frequency is extracted to facilitate the subsequent replacement of the low-frequency information.

[0128] Step 606: Replace the original low-frequency data in the original scan data with the intermediate low-frequency data to obtain the target scan data.

[0129] Exemplarily, in the case where the original scanning data includes original scanning data at multiple different scanning angles, dimensionality downsampling can be performed on the original scanning data at each scanning angle, so as to obtain intermediate scanning data corresponding to the original scanning data at each scanning angle; on this basis, for the original scanning data at each scanning angle, the intermediate low-frequency data extracted from the corresponding intermediate scanning data is used to replace the original low-frequency data in the corresponding original scanning data, so as to obtain the target scanning data at each scanning angle.

[0130] That is to say, when there are raw scan data at multiple scan angles, it is necessary to replace the low-frequency information of the raw scan data at each scan angle, that is, perform correction view by view, and finally obtain the target scan data corresponding to each scan angle.

[0131] Step 608: Perform image reconstruction based on the target scanning data to obtain a target reconstructed image of the target object.

[0132] The target scanning data includes target scanning data at multiple scanning angles.

[0133] Exemplarily, a computer device may perform image reconstruction based on the target scanning data to obtain a corrected target reconstructed image; wherein, for the image reconstruction method, a filtered back projection reconstruction method or an iterative reconstruction method may be used, and the embodiments of the present application do not make specific limitations on this.

[0134] In this embodiment, the original scan data is downsampled to obtain intermediate scan data, and low-frequency information is extracted from the intermediate scan data to obtain intermediate low-frequency data corresponding to the intermediate scan data; and low-frequency information is extracted from the original scan data to obtain original low-frequency data corresponding to the original scan data; then, the original low-frequency data in the original scan data is replaced with the intermediate low-frequency data in the intermediate scan data to obtain target scan data; and then, image reconstruction is performed based on the target scan data to obtain a target reconstructed image of the target object. By adopting this method, the low-frequency information in the original scan data can be replaced in the projection domain, thereby realizing data correction at uneven locations and improving the uniformity of the reconstructed image.

[0135] In an exemplary embodiment, Figure 7 As shown, a reconstruction method for image uniformity correction based on the image domain is provided, and the method flow includes the following steps:

[0136] Step 1: Obtain the original scan data of the target object.

[0137] Step 2, performing conventional reconstruction processing on the original scan data, that is, first performing conventional correction on the original scan data, and then performing image reconstruction on the corrected original scan data to obtain an original reconstructed image, which is represented by I1.

[0138] The water phantom is used as an example for testing, where the window width WW and window level WL of the reconstructed image are [WW, WL] = [50,0] respectively. Referring to FIG8(a), it shows the original reconstructed image I1 obtained after conventional processing.

[0139] Step 3, perform low-frequency correction on the original scan data, that is, first perform dimensionality downsampling on the original scan data; illustratively, a suitable preset merging range can be selected, for example: a merging range of 2×2, and statistically sum each 2×2 data in the original scan data to meet the statistical requirements, ensuring that the subsequent correction and reconstruction are near the true value; the dimensionality downsampling of the original scan data is achieved through the preset merging range to obtain intermediate scan data; then, the intermediate scan data after dimensionality downsampling is subjected to conventional correction and image reconstruction to obtain an intermediate reconstructed image after dimensionality downsampling, which is represented as I2. Refer to FIG9(a), which shows the intermediate reconstructed image I2 obtained after low-frequency correction.

[0140] Step 4: extract low-frequency information from the original reconstructed image I1 to obtain the original low-frequency image LP(I1), as shown in FIG8(b), and extract low-frequency information with the same cutoff frequency from the intermediate reconstructed image I2 after dimensional downsampling to obtain the intermediate low-frequency image LP(I2), as shown in FIG9(b). Then, the intermediate low-frequency image LP(I2) based on the intermediate reconstructed image I2 replaces the original low-frequency image LP(I1) in the original reconstructed image I1 to obtain the corrected target reconstructed image I out ,like Fig.10 As shown, it can be expressed as:

[0141] I out = I1- LP(I1) + LP(I2)

[0142] Among them, I1 is the original reconstructed image, I2 is the intermediate reconstructed image, LP(I1) is the original low-frequency image, LP(I2) is the intermediate low-frequency image, I out Reconstruct the image for the target.

[0143] contrast Fig.10As shown in Figure 8(a), the signal in the central area of ​​the original reconstructed image in Figure 8(a) is weak, the pixel distribution between the central area and the surrounding area in the reconstructed original reconstructed image is uneven, and the image uniformity is poor; however, by adopting the method proposed in the embodiment of the present application, after dimensionality downsampling and low-frequency information replacement of the original scan data, the pixels in the central area of ​​the corrected target reconstructed image are consistent with the pixels in the surrounding area, and the uniformity of the overall image is greatly improved.

[0144] Furthermore, since image domain correction requires two reconstructions for each image, the reconstruction matrix can be reduced to speed up reconstruction. In the above step 3, for the intermediate scan data after dimensional downsampling, after conventional correction, a smaller-sized matrix reconstruction can be performed on the intermediate scan data at this time, and then the candidate low-frequency image obtained by filtering is interpolated to obtain an intermediate low-frequency image with the same size as the original reconstructed image, so as to subsequently replace the original low-frequency image in the original reconstructed image.

[0145] In an exemplary embodiment, Fig.11 As shown, a reconstruction method for image uniformity correction based on projection domain is provided, and the method flow includes the following steps:

[0146] Step 1: Obtain the original scan data of the target object.

[0147] Step 2: Perform conventional correction processing on the original scan data to obtain corrected original scan data.

[0148] Step 3: perform dimensionality downsampling processing on the original scan data to obtain intermediate scan data, and perform conventional correction processing on the intermediate scan data to obtain corrected intermediate scan data.

[0149] Step 4: extract low-frequency information from the corrected intermediate scanning data to obtain intermediate low-frequency data, and extract low-frequency information from the corrected original scanning data to obtain original low-frequency data; use the intermediate low-frequency data extracted from the intermediate scanning data to replace the original low-frequency data extracted from the corresponding original scanning data to obtain target scanning data.

[0150] Step 5: reconstruct the image based on the target scanning data to obtain a corrected target reconstructed image.

[0151] It should be noted that image domain correction is applied to a single reconstructed image, while projection domain correction requires correcting the original scan data at each scanning angle before reconstruction one by one, and then reconstructing the corrected target scan data at each scanning angle to obtain the corrected target reconstructed image.

[0152] The image reconstruction method proposed in the embodiment of the present application improves the statistics of the detector unit after data merging by dimensionally merging and downsampling the original scanning data collected by the detector, does not rely on external information such as interpolation or model prediction, can ensure the accuracy of subsequent correction and reconstruction, eliminate low-frequency ring artifacts, and improve the uniformity of the image; in addition, by adopting this method, by dimensional downsampling analysis of the original scanning data, the consistency of the low-frequency information and the low-frequency information structure of the original data can be ensured; and this method only changes the low-frequency components of the original image, and will not affect the noise and resolution of the original image, thereby achieving the correction of low-frequency artifacts and improving the quality of the reconstructed image. Furthermore, the low-frequency information of the merged data after dimensional downsampling can be replaced in the projection domain or in the image domain; this method is not limited to the type of detector, and can be used for traditional energy integrating detector (EID) CT, and can also be used for photon counting detector (PCD) CT, to solve the ray starvation problem that may occur in the low-dose, high-resolution mode of PCCT, and improve the image quality of PCCT in the high-resolution mode.

[0153] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0154] Based on the same inventive concept, the embodiment of the present application also provides an image reconstruction device for implementing the above-mentioned image reconstruction method. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above-mentioned method, so the specific limitations in one or more image reconstruction device embodiments provided below can refer to the limitations on the image reconstruction method above, and will not be repeated here.

[0155] In an exemplary embodiment, Fig.12 As shown, an image reconstruction device is provided, including: an acquisition module 1202, a processing module 1204, an extraction module 1206 and a reconstruction module 1208, wherein:

[0156] The acquisition module 1202 is used to acquire the original scan data of the target object.

[0157] The processing module 1204 is used to perform dimensionality downsampling processing on the original scan data to obtain intermediate scan data; the dimensionality downsampling processing includes adding the data in the original scan data that is within a preset merging range.

[0158] An extraction module 1206 is used to extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; the low-frequency information includes a low-frequency image or low-frequency data;

[0159] The reconstruction module 1208 is used to perform image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object.

[0160] In one embodiment, the raw scan data includes raw scan data of multiple scan angles acquired by the medical scanning device during the rotational scanning process, and the intermediate scan data includes intermediate scan data of each scan angle; the processing module 1204 includes:

[0161] A processing unit, for performing sum processing on a plurality of data within a preset merging range in a merging area of ​​the original scanning data at each scanning angle, to obtain merged data;

[0162] The determining unit is used to determine the scanning data of the original scanning data in the non-merging area, and combine the scanning data of the non-merging area with the merged data as the intermediate scanning data corresponding to the scanning angle.

[0163] In one of the embodiments, when dimensionality downsampling is performed along the channel direction of the detector in the medical scanning device, the merged area is a partial area along the channel direction of the original scanning data; when dimensionality downsampling is performed along the row direction of the detector, the merged area is a partial area along the row direction of the original scanning data.

[0164] In one embodiment, the low-frequency information is a low-frequency image; the extraction module 1206 includes:

[0165] A first reconstruction unit, configured to perform image reconstruction based on the intermediate scanning data to obtain an intermediate reconstructed image;

[0166] The first extraction unit is used to extract low-frequency information from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

[0167] In one embodiment, the reconstruction module 1208 includes:

[0168] A second reconstruction unit, used for performing image reconstruction based on the original scan data to obtain an original reconstructed image;

[0169] A second extraction unit is used to extract low-frequency information from the original reconstructed image to obtain an original low-frequency image corresponding to the original scan data;

[0170] The first replacement unit is used to replace the original low-frequency image with the intermediate low-frequency image in the original reconstructed image to obtain a target reconstructed image of the target object.

[0171] In one of the embodiments, if the size of the intermediate reconstructed image is smaller than the size of the original reconstructed image, the first extraction unit is used to extract low-frequency information from the intermediate reconstructed image to obtain a candidate low-frequency image; and perform interpolation processing on the candidate low-frequency image to obtain an intermediate low-frequency image.

[0172] In one embodiment, the low-frequency information is low-frequency data; the extraction module 1206 is used to extract the low-frequency information from the intermediate scanning data to obtain intermediate low-frequency data corresponding to the intermediate scanning data.

[0173] In one embodiment, the reconstruction module 1208 includes:

[0174] A third extraction unit is used to extract low-frequency information from the original scanning data to obtain original low-frequency data corresponding to the original scanning data;

[0175] A second replacement unit is used to replace the original low-frequency data in the original scanning data with the intermediate low-frequency data to obtain the target scanning data;

[0176] The third reconstruction unit is used to perform image reconstruction based on the target scanning data to obtain a target reconstructed image of the target object.

[0177] Each module in the above-mentioned image reconstruction device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module above.

[0178] In an exemplary embodiment, a CT imaging system is provided, including a photon counting detector array, a processing circuit, and a reconstruction system, wherein the photon counting detector array includes a plurality of voxels arranged along a channel direction and a row direction, and the plurality of voxels are divided into a plurality of boxed voxels along the channel direction and / or the row direction, and each boxed voxel includes a plurality of voxels.

[0179] Voxels are used to obtain the energy of photon information and generate detection signals when performing CT scanning on a target object.

[0180] The processing circuit is used to obtain the original scanning data of the target object according to the detection signal generated by each voxel; the original scanning data includes the data corresponding to each voxel.

[0181] The processing circuit is also used to obtain intermediate scanning data according to the original scanning data; the intermediate scanning data includes data corresponding to each boxed voxel, and the data corresponding to the boxed voxel is obtained by adding the data of each voxel corresponding to the boxed voxel in the original scanning data.

[0182] The reconstruction system is used to extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; and to reconstruct an image based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object; the low-frequency information includes a low-frequency image or low-frequency data.

[0183] In this example, the medical imaging device may be a PCCT device, i.e., a photon counting CT device; the PCCT device is equipped with a CT imaging system, which may include but is not limited to a photon counting detector array, a processing circuit, and a reconstruction system; wherein the photon counting detector array is composed of a plurality of voxels, and the plurality of voxels form an M×N detector array along a channel direction and a row direction, where M may be the number of voxels in the channel direction, i.e., the number of voxels along the circumferential direction of the PCCT device frame; and N may be the number of voxels in the row direction, i.e., the number of voxels along the axial direction of the PCCT device frame.

[0184] Exemplarily, for a plurality of voxels in a photon counting detector array, the plurality of voxels may be divided into a plurality of binned voxels along a channel direction and / or a row direction, that is, each binned voxel includes a plurality of voxels. Exemplarily, for each binned voxel, the binned voxel may be an m×n voxel array, wherein m is the number of voxels in the channel direction in the binned voxel, and n is the number of voxels in the row direction in the binned voxel.

[0185] Exemplarily, m=1 and n≥2; or, m≥2 and n=1; or, m≥2 and n≥2; that is, the binned voxel can be a row vector voxel, that is, a binned voxel composed of multiple voxels in a single channel direction, that is, m=1 and n≥2; the binned voxel can also be a column vector voxel, that is, a binned voxel composed of multiple voxels in a single row direction, that is, m≥2 and n=1; the binned voxel can also be a matrix voxel, that is, a binned voxel composed of voxels in multiple channel directions and multiple row directions, that is, m≥2 and n≥2.

[0186] Exemplarily, for each binned voxel, there may be no overlapping voxels between the binned voxels, or there may be overlapping voxels. If there are no overlapping voxels between the binned voxels, then for an M×N detector array, each m×n voxel can be divided into a binned voxel; for example, a 4×12 detector array can be divided according to a size of 2×2, thereby obtaining 12 binned voxels, that is, 2×6 binned voxels. If there are overlapping voxels between the binned voxels, then for an M×N detector array, the M×N detector array can be binned according to a bin size of m×n and an overlap size of p×q, thereby obtaining a plurality of binned voxels; for example, for a 4×12 detector array, assuming that the bin size is 4×2 and the overlap size is 4×1, 4×11 binned voxels can be obtained, that is, 44 binned voxels.

[0187] For each binned voxel, during subsequent data processing, the data of all voxels in the binned voxel can be summed to obtain the summed data corresponding to the binned voxel; the summed data corresponding to the multiple binned voxels can constitute the original scan data after dimensionality downsampling, that is, the intermediate scan data.

[0188] It should be noted that, for multiple binned voxels with overlapping voxels, the intermediate scan data obtained after the summing process has a smaller dimension reduction than the original scan data; while for multiple binned voxels without overlapping voxels, the intermediate scan data obtained after the summing process has a larger dimension reduction than the original scan data. Compared with the image resolution obtained after image uniformity correction based on the intermediate scan data obtained from multiple binned voxels without overlapping voxels, the image resolution obtained after image uniformity correction based on the intermediate scan data obtained from multiple binned voxels with overlapping voxels is higher, that is, the overlapping binning method can improve the image resolution.

[0189] Next, when the binned voxels are summed to obtain intermediate scan data, the reconstruction system can extract low-frequency information from the intermediate low-frequency information, and in the process of image reconstruction based on the original scan data, the low-frequency information extracted from the intermediate low-frequency information is used to replace the low-frequency information in the original scan data, thereby obtaining a target reconstructed image after uniformity correction.

[0190] The specific implementation process of image reconstruction based on low-frequency information may refer to the relevant description of the image reconstruction method in any of the above embodiments, and will not be repeated here.

[0191] By adopting the CT imaging system proposed in the embodiment of the present application, when performing CT imaging, especially PCCT imaging, it is possible to improve the uniformity correction effect and obtain a reconstructed image with higher uniformity, thereby improving the quality of the reconstructed image.

[0192] In an exemplary embodiment, a computer device is provided. The computer device may be a medical scanning device, or a terminal or a server connected to the medical scanning device for communication. Exemplarily, when the computer device is a terminal, its internal structure diagram may be as follows: Fig.13 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, an image reconstruction method is realized. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0193] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0194] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the image reconstruction method in any of the above embodiments are implemented.

[0195] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the image reconstruction method in any of the above embodiments are implemented.

[0196] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of the image reconstruction method in any of the above embodiments.

[0197] It should be noted that the data involved in this application (including but not limited to data used for analysis, data stored, data displayed, etc.) are information and data fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0198] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive 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. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0199] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0200] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. An image reconstruction method, characterized in that: The method comprises: Obtaining raw scan data of the target object; Performing dimensionality downsampling processing on the original scan data to obtain intermediate scan data; the dimensionality downsampling processing includes summing up the data in the original scan data that is within a preset merging range; Extracting low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; the low-frequency information includes a low-frequency image or low-frequency data; Image reconstruction is performed based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object.

2. The method according to claim 1, characterized in that The original scanning data includes original scanning data of multiple scanning angles collected by the medical scanning device during the rotation scanning process, and the dimensionality downsampling process is performed on the original scanning data to obtain the intermediate scanning data, including: For each of the original scan data at the scanning angle, in a merging area of ​​the original scan data, a plurality of data in the preset merging range are summed to obtain merged data; The original scan data is determined to be scan data in a non-merging area, and the scan data in the non-merging area is combined with the merged data as intermediate scan data corresponding to the scan angle.

3. The method according to claim 2, characterized in that In the case where dimensionality downsampling processing is performed along the channel direction of the detector in the medical scanning device, the merged area is a partial area of ​​the original scanning data along the channel direction; In the case where dimensionality downsampling processing is performed along the row direction of the detector, the merged area is a partial area along the row direction in the original scanning data.

4. The method according to any one of claims 1 to 3, characterized in that The low-frequency information is a low-frequency image; extracting the low-frequency information based on the intermediate scanning data to obtain the intermediate low-frequency information corresponding to the intermediate scanning data includes: Performing image reconstruction based on the intermediate scanning data to obtain an intermediate reconstructed image; Low-frequency information is extracted from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

5. The method according to claim 4, characterized in that The performing image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object includes: Performing image reconstruction based on the original scan data to obtain an original reconstructed image; Extracting low-frequency information from the original reconstructed image to obtain an original low-frequency image corresponding to the original scan data; The intermediate low-frequency image is used to replace the original low-frequency image in the original reconstructed image to obtain a target reconstructed image of the target object.

6. The method according to claim 5, characterized in that If the size of the intermediate reconstructed image is smaller than the size of the original reconstructed image, extracting low-frequency information from the intermediate reconstructed image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data includes: Extracting low-frequency information from the intermediate reconstructed image to obtain a candidate low-frequency image; An interpolation process is performed on the candidate low-frequency image to obtain an intermediate low-frequency image corresponding to the intermediate scanning data.

7. The method according to any one of claims 1 to 3, characterized in that The low-frequency information is low-frequency data; the step of extracting the low-frequency information based on the intermediate scanning data to obtain the intermediate low-frequency information corresponding to the intermediate scanning data includes: Low-frequency information is extracted from the intermediate scanning data to obtain intermediate low-frequency data corresponding to the intermediate scanning data.

8. The method according to claim 7, characterized in that The performing image reconstruction based on the intermediate low-frequency information and the original scan data to obtain a target reconstructed image of the target object includes: Extracting low-frequency information from the original scan data to obtain original low-frequency data corresponding to the original scan data; Replacing the original low-frequency data in the original scan data with the intermediate low-frequency data to obtain target scan data; Image reconstruction is performed based on the target scanning data to obtain a target reconstructed image of the target object.

9. A CT imaging system comprising a photon counting detector array, a processing circuit and a reconstruction system, characterized in that: The photon counting detector array comprises a plurality of voxels arranged along a channel direction and a row direction, wherein the plurality of voxels are divided into a plurality of boxed voxels along the channel direction and / or the row direction, and each of the boxed voxels comprises a plurality of the voxels; The voxel is used to obtain the energy of photon information and generate a detection signal when a CT scan is performed on the target object; The processing circuit is used to obtain raw scan data of the target object according to the detection signal generated by each voxel; the raw scan data includes data corresponding to each voxel; The processing circuit is further used to obtain intermediate scanning data according to the original scanning data; the intermediate scanning data includes data corresponding to each of the boxed voxels, and the data corresponding to the boxed voxels are obtained by adding the data of each of the voxels corresponding to the boxed voxels in the original scanning data; The reconstruction system is used to extract low-frequency information based on the intermediate scanning data to obtain intermediate low-frequency information corresponding to the intermediate scanning data; and perform image reconstruction based on the intermediate low-frequency information and the original scanning data to obtain a target reconstructed image of the target object; the low-frequency information includes a low-frequency image or low-frequency data.

10. The system according to claim 9, characterized in that The binned voxels are m×n voxel arrays; wherein m=1 and n≥2; or, m≥2 and n=1; or, m≥2 and n≥2; There are no overlapping voxels between the boxed voxels, or there are overlapping voxels.