Method for ct data compression and method for obtaining raw projection data

Abstract

The present application relates to a kind of CT data compression method and the method of obtaining original projection data, belong to CT image processing technical field, solve the problem that original projection data cannot be obtained after CT data compression in prior art.The compression method includes: the first projection data of detector acquisition is sequentially pre-processed, reconstruction operation and compression, and the compressed second reconstruction data is stored to preset storage memory;Second reconstruction data is sequentially decompressed, projection operation and inverse operation processing of pre-processing, obtain fourth projection data;The fourth projection data and the first projection data are obtained first difference data by difference;Based on preset lossless compression algorithm, first difference data is compressed, and second difference data is obtained, and second difference data is stored to preset storage memory.Original projection data can be obtained, and the compression storage of CT data is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of CT image processing, and in particular to a CT data compression method and a method for obtaining original projection data. BACKGROUND

[0002] Computed Tomography (CT) generates a large amount of projection image data and reconstructed image data in use, but due to the subsequent viewing and comparison requirements, a large amount of storage space is occupied.

[0003] With the development of science and technology, CT data is increasingly widely used, and compression storage of CT data is becoming more and more important. Due to the characteristics of smooth image patches and small local fluctuations of reconstructed image data, a high compression ratio can be easily achieved; due to the influence of quantum noise in data acquisition, it is difficult to have a high compression ratio under lossless compression conditions, and the existing technology adopts the method of retaining reconstructed image data and discarding projection image data. However, due to subsequent data analysis, reconstruction algorithm update, hardware error troubleshooting and other reasons, it is still necessary to save the projection image data.

[0004] Therefore, there is an urgent need for a CT data compression method that can occupy a small amount of storage space while taking into account the subsequent data needs. SUMMARY

[0005] In view of the above analysis, the embodiments of the present application aim to provide a CT data compression method and a method for obtaining original projection data, to solve the problem that the CT data compressed in the prior art cannot obtain original projection data.

[0006] In one aspect, the embodiments of the present application provide a CT data compression method, the compression method comprising:

[0007] obtaining original projection data collected by a detector as first projection data, and pre-processing the first projection data to obtain second projection data;

[0008] reconstructing the second projection data based on a preset reconstruction algorithm to obtain initial reconstructed data;

[0009] compressing the initial reconstructed data based on a preset compression algorithm to obtain second reconstructed data, and storing the second reconstructed data in a preset storage memory;

[0010] decompressing the second reconstructed data based on the preset compression algorithm to obtain third reconstructed data;

[0011] projecting the third reconstructed data based on a preset forward projection algorithm to obtain third projection data;

[0012] performing inverse operation of the preprocessing on the third projection data to obtain fourth projection data;

[0013] obtaining first difference data by subtracting the fourth projection data from the first projection data; compressing the first difference data based on a preset compression algorithm to obtain second difference data, and storing the second difference data into the preset storage memory.

[0014] In further improvement of the above method, the compressing the initial reconstruction data based on the preset compression algorithm to obtain the second reconstruction data comprises:

[0015] mapping the initial reconstruction data to an integer domain based on a preset mapping algorithm to obtain first reconstruction data;

[0016] compressing the first reconstruction data based on the preset compression algorithm to obtain the second reconstruction data.

[0017] In further improvement of the above method, decompressing the second reconstruction data based on the preset compression algorithm to obtain third reconstruction data comprises:

[0018] decompressing the second reconstruction data based on the preset compression algorithm to obtain fourth reconstruction data;

[0019] performing inverse operation of the preset mapping algorithm on the fourth reconstruction data to obtain the third reconstruction data.

[0020] In further improvement of the above method, the compressing the first reconstruction data based on the preset compression algorithm to obtain the second reconstruction data comprises:

[0021] performing image noise reduction processing on the first reconstruction data based on a preset image processing algorithm to obtain fifth reconstruction data; the image processing algorithm is a three-dimensional block matching algorithm or a non-local average algorithm;

[0022] compressing the fifth reconstruction data based on the preset compression algorithm to obtain the second reconstruction data.

[0023] In further improvement of the above method, when the preset compression algorithm adopts a lossless compression algorithm, the initial reconstruction data is taken as the third reconstruction data.

[0024] In further improvement of the above method, the preprocessing comprises one or more of dark field correction, gain correction, log calculation, scatter correction and hardening correction.

[0025] Based on the further improvement of the above method, the preset reconstruction algorithm is a filtered back-projection algorithm or an iterative algorithm.

[0026] Based on the further improvement of the above method, the preset compression algorithm is a lossless compression algorithm or a lossy compression algorithm, the lossless compression algorithm includes a JPEG2000 lossless compression algorithm and a JPEG-LS algorithm, and the lossy compression algorithm includes an SPIHT algorithm and a JPEG2000 lossy compression algorithm.

[0027] Based on the further improvement of the above method, the preset forward projection algorithm is a point-driven projection algorithm, a line-driven projection algorithm or a distance-driven projection algorithm.

[0028] In another aspect, the embodiment of the present application provides a method for obtaining original projection data based on CT compressed data, and the method for obtaining original projection data comprises:

[0029] The projection difference data and the compressed reconstruction data are obtained from the memory, and the projection difference data is decompressed to obtain first difference data;

[0030] The compressed reconstruction data is decompressed based on a preset compression algorithm to obtain third reconstruction data;

[0031] The third reconstruction data is projected based on a preset forward projection algorithm to obtain third projection data;

[0032] The third projection data is processed by inverse operation of the preprocessing to obtain fourth projection data;

[0033] The first difference data and the fourth projection data are summed to obtain original projection data.

[0034] Compared with the prior art, the present application can achieve at least one of the following beneficial effects:

[0035] 1、The first projection data collected is sequentially preprocessed, reconstructed and compressed, and the second reconstruction data after compression is stored in a preset storage memory, and the second reconstruction data is sequentially decompressed, projected and processed by inverse operation of the preprocessing to obtain fourth projection data; the fourth projection data is subtracted from the first projection data to obtain first difference data; the first difference data is compressed based on a preset lossless compression algorithm to obtain second difference data, and the second difference data is stored in the preset storage memory, thereby realizing the compressed storage of CT data.

[0036] 2、The application obtains first difference data by decompressing the compressed projection difference data, and obtains fourth projection data by sequentially decompressing, projecting and pre-processing the inverse operation of the compressed reconstruction data, and obtains the original projection data by summing the first difference data and the fourth projection data.

[0037] 3、The application improves the efficiency of data processing and reduces the time of CT data compression by mapping the initial reconstruction data to the integer domain.

[0038] In the application, the above technical solutions can be combined with each other to realize more preferred combination solutions. Other features and advantages of the application will be described in the subsequent description, and some advantages will become apparent from the description or be understood by implementing the application. The purposes and other advantages of the application can be realized and obtained from the contents specifically indicated in the description and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0039] The accompanying drawings are included to provide a further understanding of the application and are incorporated herein and constitute a part of the application. The same reference numbers in the drawings represent the same elements throughout the drawings.

[0040] Figure 1 A flowchart of a CT data compression method provided by an embodiment of the application is shown in the figure.

[0041] Figure 2 A data flow conversion schematic diagram of the CT data compression method provided by the embodiment of the application is shown in the figure.

[0042] Figure 3 A flowchart of a method for obtaining original projection data based on CT compressed data provided by an embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0043] The preferred embodiments of the application are specifically described below with reference to the accompanying drawings, which form a part of the application and are used to explain the principles of the application together with the embodiments of the application, but are not used to limit the scope of the application.

[0044] One specific embodiment of the application discloses a CT data compression method, as shown in the figure, the compression method comprises: Figure 1

[0045] Step S11: Obtain the original projection data collected by the detector as first projection data, and pre-process the first projection data to obtain second projection data;

[0046] Step S12: Based on a preset reconstruction algorithm, the second projection data is reconstructed to obtain initial reconstruction data; ​

[0047] Step S13: based on a preset compression algorithm, compressing the initial reconstruction data to obtain second reconstruction data, and storing the second reconstruction data to a preset storage memory;

[0048] Step S14: based on the preset compression algorithm, decompressing the second reconstruction data to obtain third reconstruction data;

[0049] Step S15: based on a preset forward projection algorithm, performing a projection operation on the third reconstruction data to obtain third projection data;

[0050] Step S16: performing an inverse operation of the preprocessing on the third projection data to obtain fourth projection data;

[0051] Step S17: subtracting the fourth projection data from the first projection data to obtain first difference data; based on a preset lossless compression algorithm, compressing the first difference data to obtain second difference data, and storing the second difference data to the preset storage memory.

[0052] Specifically, the reconstructed image data can be generated by the projection image data through a reconstruction algorithm, but the projection image data is difficult to have a high compression ratio under lossless compression due to the influence of quantum noise in the sampling. Therefore, separately storing the projection image data after compression still occupies a large amount of memory space, and cannot achieve the purpose of storing CT data occupying a small amount of memory space.

[0053] Meanwhile, the reconstructed image data can easily achieve a high compression ratio due to its characteristics of smooth image slices and small local fluctuations. Separately storing the reconstructed image data after compression can certainly make CT data occupy a small amount of memory space and achieve compression of CT data. However, due to subsequent data analysis, reconstruction algorithm update, hardware error troubleshooting, and other reasons, the original projection data is still needed, but the original projection data cannot be obtained from the compressed and stored reconstructed image data, and thus subsequent data analysis and other operations cannot be performed. The present application compresses the first difference data obtained by subtracting the fourth projection data from the first projection data to obtain second difference data and stores it, and subsequently restores the original projection data based on the second difference data, thereby achieving compression storage of the reconstructed data and the projection data.

[0054] Specifically, as Figure 1 and Figure 2As shown, in step S11, the raw projection data collected by the detector is acquired, and the raw projection data is taken as the first projection data P1. The raw projection data is the raw data collected by different detector units in different environments. Exemplarily, the size of the raw projection data is 300*64*1800, indicating the data of the raw projection data in 1800 different positions, and a two-dimensional array of 300*64 is used to represent the data of each position.

[0055] Due to the inconsistency between different detector units, the raw projection data needs to be preprocessed in step S11. It can be understood that the inconsistency mainly reflects the difference in the dark current response of different detector units in the dark field environment, and the difference in the inconsistency of the ray response caused by the difference in the ray distribution and the inconsistency between the backend electronic modules.

[0056] Preferably, the preprocessing includes one or more of offset calibration, gain calibration, log calculation, scatter correction, and hardening correction.

[0057] Specifically, for the difference in the dark current response of different detector units in the dark field environment, offset calibration can be performed. It can be understood that offset calibration usually adopts the method of subtracting dark field data from the collected data to eliminate the difference caused by the dark current of the detector unit.

[0058] For the difference in the inconsistency of the ray response caused by the difference in the ray distribution and the inconsistency between the backend electronic modules, gain calibration can be performed. It can be understood that gain calibration usually adopts the "two-point correction method", and the gain coefficient for correction is obtained by calculating the ratio of the set point value and the data after gain correction under the premise that the detector pixels are assumed to be linearly responsive.

[0059] Specifically, in order to eliminate the influence of scattered rays contained in the first projection data P1, scatter correction can be used. It can be understood that in the process of interaction between X-rays and matter, non-coherent scattering or Compton scattering occurs, that is, when X-ray photons collide with electrons, part of the energy of the incident photons is transferred to the electrons, and due to the conservation of momentum, the path of the photons is deviated. Therefore, part of the signal collected by the final detector is derived from scattering, thereby deviating from the true measurement result of the X-ray intensity, resulting in numerical deviation and artifacts in the reconstructed image. Scatter correction is to process the projection value to minimize the influence of scattering on the reconstruction result.

[0060] Specifically, because the rays are not mono-energetic, there is a relatively wide energy spectrum, hardening correction can be used. It can be understood that, due to the assumption that the rays are mono-energetic in CT reconstruction, however, the X-ray generated by the X-ray tube covers a very wide spectrum. Low-energy X-ray photons are more easily absorbed by most materials than more energetic X-ray photons, and as a result, the proportion of high-energy photons in the X-ray becomes more (more penetrating, more "hard"). Hardening correction is a method of remapping the projections through the attenuation properties of the material, compensating for the hardening error.

[0061] Also, for the convenience of processing of the first projection data P1, log calculation can also be added at the end of the preprocessing, reducing the amount of data calculation and improving the processing efficiency of the data. It can be understood that CT reconstruction needs to rely on the measurement results of X-ray beams at different angles. In the case of assuming that X-ray photons are mono-energetic, the change of X-ray intensity obeys the Lambert-Beers law, that is, the rays present the law of exponential decay. Therefore, the log correction is represented as the line integral of the attenuation coefficient on the X-ray path by the logarithmic operation of the ratio of the incident and exit X-ray intensities, as the projection measurement value required in CT reconstruction.

[0062] It is worth noting that in step S11, when the first projection data P1 is preprocessed, the first projection data P1 can be detected in advance, and according to the detection result of the first projection data P1, the preprocessing process can be selected accordingly. For example, when the detection result shows that different detector units have differences in their own dark current response, the preprocessing needs to include dark field correction; when the detection result shows that different detector units are affected by scattered rays, the preprocessing needs to include scatter correction.

[0063] Specifically, as shown in Figure 1 and Figure 2 After the first projection data P1 is preprocessed, the preprocessed projection data is taken as the second projection data P2.

[0064] Specifically, as shown in Figure 1 and Figure 2As shown in step S12, the preset reconstruction algorithm is a reconstruction algorithm preset for the reconstruction operation on the second projection data P2. Preferably, the preset reconstruction algorithm can be a filtered back projection (FBP) or an iterative algorithm. It can be understood that the FBP has the advantage of fast reconstruction speed and needs to meet certain reconstruction conditions. The iterative algorithm obtains the final reconstruction result by iteration of data using an optimization algorithm, and the reconstructed image result is generally better than that of the FBP, but the reconstruction time is relatively long. In step S12, when the reconstruction operation is performed on the second projection data P2, a suitable reconstruction algorithm can be selected as the preset reconstruction algorithm according to actual conditions, and after the reconstruction operation is performed on the second projection data P2, the data obtained by the reconstruction operation is taken as the initial reconstruction data X.

[0065] Specifically, as shown in steps S13 and S14, Figure 1 and Figure 2 in step S13, the initial reconstruction data X is compressed based on a preset compression algorithm to obtain second reconstruction data X2, and the second reconstruction data X2 is stored in a preset storage memory.

[0066] Specifically, the preset compression algorithm is a compression algorithm preset for the compression of the initial reconstruction data X. Preferably, the preset compression algorithm is a lossless compression algorithm or a lossy compression algorithm. The lossless compression algorithm includes a JPEG2000 lossless compression algorithm and a JPEG-LS algorithm, and the lossy compression algorithm includes a SPIHT algorithm and a JPEG2000 lossy compression algorithm.

[0067] Specifically, the preset storage memory is a storage device for storing the compressed CT data. The storage device can be a random access memory (RAM), a read-only memory (ROM), a U disk, a magneto-optical disc (MO disc), or the like.

[0068] As shown in steps S13 and S14, Figure 2 after the second reconstruction data X2 is obtained, the second reconstruction data X2 is stored in the preset storage memory. It can be understood that the second reconstruction data X2 belongs to the reconstructed image data of the original projection data, and the reconstructed image data can achieve a high compression ratio. Storing the second reconstruction data X2 can reduce the memory space occupied by the storage of the original projection data.

[0069] In step S13, as shown in steps S13 and S14, Figure 2 preferably, the compression of the initial reconstruction data based on the preset compression algorithm to obtain the second reconstruction data includes:

[0070] mapping the initial reconstruction data to an integer domain based on a preset mapping algorithm to obtain first reconstruction data;

[0071] compressing the first reconstruction data based on the preset compression algorithm to obtain the second reconstruction data.

[0072] Specifically, the preset mapping algorithm is a mapping algorithm that is preset for mapping the initial reconstruction data X to the integer domain to obtain the first reconstruction data X1. The initial reconstruction data X is of a floating-point type and is inconvenient for image processing. After the initial reconstruction data X is mapped to the integer domain to obtain the first reconstruction data X1, the data amount can be further reduced, and the efficiency of image processing can be improved.

[0073] The preset mapping algorithm can be a linear mapping algorithm, and the initial reconstruction data is mapped by the following formula:

[0074] X1=A*X+B;

[0075] wherein X1 is the mapped first reconstruction data, X is the initial reconstruction data, and A and B are mapping parameters, which are selected according to the situation.

[0076] For example, the data range of the initial reconstruction data X is between 0.001 and 0.1. It is not too convenient to perform data analysis and processing in this data range. After the initial reconstruction data X is multiplied by 10000, the initial reconstruction data X is all mapped to an integer.

[0077] In step S13, as shown in the figure, Figure 2 Preferably, the step of compressing the first reconstruction data based on the preset compression algorithm to obtain the second reconstruction data comprises:

[0078] performing image denoising processing on the first reconstruction data based on a preset image processing algorithm to obtain fifth reconstruction data; the image processing algorithm is a three-dimensional block matching algorithm or a non-local average algorithm;

[0079] compressing the fifth reconstruction data based on the preset compression algorithm to obtain the second reconstruction data.

[0080] Specifically, the preset image processing algorithm is an algorithm that is preset for image processing of the first reconstruction data X1, further improves the image quality of the first reconstruction data X1, and facilitates subsequent data analysis to facilitate the discovery of dangerous goods or pathological changes and other information in the data.

[0081] The preset image processing algorithm can select a block matching 3D (BM3D) algorithm or a non-local means (NLMS) algorithm. It can be understood that the BM3D algorithm is a relatively good image denoising algorithm. By matching with adjacent image blocks, a plurality of similar blocks are integrated into a three-dimensional matrix, and filtering processing is performed in the three-dimensional space, and then the result is inversely transformed and fused into two dimensions to form a denoised image. The algorithm has a significant denoising effect and can obtain the highest peak signal-to-noise ratio, but has a high time complexity. The NLMS algorithm is an image denoising algorithm. Compared with a local algorithm (such as Gaussian blur and anisotropic diffusion), the NLMS algorithm only uses points near each target pixel to smooth the image to remove noise. The non-local means algorithm defines a block around each target pixel, and gives a weight to all pixels in the image according to the similarity between the block around the target pixel and the target pixel block, and then performs averaging. In this way, the processed image is clearer and has less loss of details.

[0082] Specifically, after obtaining the fifth reconstructed data X5 after image processing, the fifth reconstructed data X5 is compressed according to a preset compression algorithm to obtain the second reconstructed data X2.

[0083] Specifically, as shown in Figure 1 and Figure 2 In step S14, the second reconstructed data X2 is decompressed based on a preset compression algorithm to obtain the third reconstructed data X3.

[0084] It should be noted that the second reconstructed data X2 is data obtained by compressing the initial reconstructed data X according to a preset compression algorithm. When the second reconstructed data X2 is decompressed according to the preset compression algorithm, the third reconstructed data X3 after decompression can be obtained,

[0085] Preferably, decompressing the second reconstructed data based on the preset compression algorithm to obtain the third reconstructed data comprises:

[0086] Decompressing the second reconstructed data based on the preset compression algorithm to obtain the fourth reconstructed data;

[0087] Performing a mapping inverse operation on the fourth reconstructed data based on the preset mapping algorithm to obtain the third reconstructed data.

[0088] It is worth mentioning that if the initial reconstruction data X is mapped to the integer domain to obtain the first reconstruction data X1, and then the first reconstruction data X1 is compressed to obtain the second reconstruction data, in the step S14, the second reconstruction data X2 is decompressed to obtain the third reconstruction data X3, it is necessary to include decompressing the second reconstruction data X2 to obtain the fourth reconstruction data X4, and then performing the inverse operation of mapping on the fourth reconstruction data X4 to obtain the third reconstruction data X3.

[0089] Preferably, when the preset compression algorithm adopts a lossless compression algorithm, the initial reconstruction data is taken as the third reconstruction data.

[0090] It is worth mentioning that the lossless compression algorithm is to compress by using statistical redundancy of data, and the original data can be completely restored without causing any distortion. Therefore, if the initial reconstruction data X is compressed to obtain the second reconstruction data X2 in the step S13, the initial reconstruction data X can be taken as the third reconstruction data X3 in the step S14, and the two are completely consistent.

[0091] Specifically, as shown in Figure 1 and Figure 2 In the step S15, the third reconstruction data is projected based on a preset forward projection algorithm to obtain third projection data.

[0092] Specifically, the preset forward projection algorithm is a forward projection algorithm that is set in advance for the projection operation of the third reconstruction data X3, and the third projection data P3 is obtained after the projection operation.

[0093] Preferably, the preset forward projection algorithm is a point-driven projection algorithm, a line-driven projection algorithm or a distance-driven projection algorithm.

[0094] It can be understood that in the point-driven (Pixel Driven) projection, the rays pass through the center points of the ray source and the pixels, and then are projected onto the detector. In the forward projection process, the pixel value is interpolated to the adjacent detector cells by means of the projection of the rays on the detector. The point-driven projection brings high-frequency noise to the projection data, and in addition, if the pixel size and the size of the detector cells are not suitable, it will cause the projection data to be too sparse or dense.

[0095] It can be understood that in the line-driven (Ray Driven) projection, the selected rays pass through the center points of the ray source and the detector cells, and the projection contribution to the detector cells is calculated according to the length of the rays passing through the pixels. There are mainly three calculation methods for the length of the rays passing through the pixels, Joseph’s algorithm, Siddon’s algorithm and Cole’s algorithm. The Joseph algorithm interpolates the pixel points across each row or column as the projection result for the case that the ray is close to the coordinate axis. The Siddon algorithm calculates the intersection of the ray and each row and column of the image coordinate system, and then calculates the distance between two adjacent points. The Cole algorithm is for the three-dimensional case, which uses Simpson's rule to calculate the line integral of the ray. The Joseph algorithm is faster than the Siddon algorithm and the Cole algorithm, which need to calculate each pixel, but it needs to consider the principal direction of the ray.

[0096] It can be understood that the distance-driven projection combines the characteristics of point-driven and line-driven, projects the pixel points and the detector units onto a common axis, and takes the overlapping length as the weight coefficient. In the three-dimensional case, the grid formed after projection can be simplified as a regular rectangle to simplify the operation. The distance-driven algorithm needs to consider the principal direction of the ray to determine the pixel position used for projection and the selection of the common axis, like the Joseph algorithm. The distance-driven algorithm has the largest difference with the result of the analytical method at the position where the scanning angle is close to π / 4 times an odd number, mainly because at this angle, the projection of the pixel cannot be approximated by a square. Compared with the point-driven and line-driven algorithms, the distance-driven algorithm has more accurate results, does not need to consider the matching problem of the pixel size and the size of the detector unit, but the computational complexity is greatly improved.

[0097] Specifically, as shown in Figure 1 and Figure 2 , in step S16, an inverse operation process of the preprocessing is performed on the third projection data P3 to obtain fourth projection data P4.

[0098] It can be understood that if one or more of the dark field correction, the gain correction, the log calculation, the scatter correction and the hardening correction are included in the preprocessing, the inverse operation process of the preprocessing is the inverse operation process of the aforementioned preprocessing.

[0099] Specifically, as shown in Figure 1 and Figure 2 , in step S17, the fourth projection data P4 and the first projection data P1 are subtracted to obtain first difference data B1; the first difference data B1 is compressed based on a preset lossless compression algorithm to obtain second difference data B2, and the second difference data B2 is stored in the preset storage memory.

[0100] Specifically, the subtraction of the fourth projection data P4 and the first projection data P1 is to subtract the elements at the same position of the fourth projection data P4 from the elements at the same position of the first projection data P1.

[0101] Exemplarily, the first projection data has a size of 300*64*1800, the fourth projection data has a size of 300*64*1800, and the first difference data can be obtained by respectively subtracting elements at the same position of the first projection data and the fourth projection data.

[0102] Specifically, the preset lossless compression algorithm is a preset lossless compression algorithm for compressing the first difference data B1, and the second difference data B2 obtained after compression can be obtained by decompression.

[0103] The second difference data B2 is stored in the preset storage memory.

[0104] Compared with the prior art, the CT data compression method provided by the embodiment of the application realizes compression and storage of CT data by sequentially performing preprocessing, reconstruction operation and compression on the collected first projection data, storing the compressed second reconstruction data in the preset storage memory, sequentially performing decompression, projection operation and inverse operation processing of preprocessing on the second reconstruction data to obtain fourth projection data, subtracting the fourth projection data from the first projection data to obtain first difference data, and compressing the first difference data based on a preset lossless compression algorithm to obtain second difference data and storing the second difference data in the preset storage memory. Since the second difference data is stored, the original projection data can be recovered based on the data. Compared with directly storing the compressed projection data, storing the compressed second difference data reduces the storage capacity and also stores the projection data. At the same time, the application improves the efficiency of data processing and reduces the time of CT data compression by mapping the initial reconstruction data to the integer domain.

[0105] One specific embodiment of the application discloses a method for obtaining original projection data based on CT compressed data, as shown in Figure 3 The method for obtaining original projection data comprises the following steps:

[0106] Step S21: obtaining projection difference data and compressed reconstruction data from the memory, and decompressing the projection difference data to obtain first difference data;

[0107] Step S22: decompressing the compressed reconstruction data based on a preset compression algorithm to obtain third reconstruction data;

[0108] Step S23: performing projection operation on the third reconstruction data based on a preset forward projection algorithm to obtain third projection data;

[0109] Step S24: performing inverse operation processing of preprocessing on the third projection data to obtain fourth projection data;

[0110] Step S25: summing the first difference data and the fourth projection data to obtain original projection data.

[0111] Specifically, in step S21, the memory is a storage device storing projection difference data corresponding to original projection data and compressed reconstruction data.

[0112] It can be understood that the projection difference data is the second difference data described above, and the compressed reconstruction data is the second reconstruction data described above.

[0113] The projection difference data and the compressed reconstruction data are read from the memory, the projection difference data is decompressed to obtain the first difference data. It should be noted that the first difference data is obtained by compressing the projection difference data by a lossless compression algorithm, so that the first difference data can be obtained after decompressing the projection difference data.

[0114] In step S22, the compressed reconstruction data is decompressed based on a preset compression algorithm to obtain third reconstruction data.

[0115] In step S23, the third reconstruction data is projected based on a preset forward projection algorithm to obtain third projection data.

[0116] In step S24, the third projection data is preprocessed to obtain fourth projection data. It can be understood that the preset compression algorithm in step S22, the preset forward projection algorithm in step S23 and the preprocessing in step S24 are all preset algorithms, which will not be described here.

[0117] After obtaining the fourth projection data and the first difference data, in step S25, the fourth projection data and the first difference data are summed to obtain the original projection data.

[0118] Compared with the prior art, the method for obtaining original projection data based on CT compressed data provided by the embodiment of the application obtains the first difference data by decompressing the compressed projection difference data, and obtains the fourth projection data by sequentially decompressing, projecting and preprocessing the compressed reconstruction data. The first difference data and the fourth projection data are summed to obtain the original projection data, and the original projection data can be used for subsequent data analysis and other processes.

[0119] Those skilled in the art can understand that all or part of the processes of the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. The computer readable storage medium is a disk, an optical disk, a read-only memory or a random access memory, etc.

[0120] The above description is only the preferred embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A method for compressing CT data, characterized in that, The compression method includes: The original projection data collected by the detector is obtained as the first projection data. The first projection data is detected, and the preprocessing process is determined based on the retrieval results. The first projection data is then preprocessed to obtain the second projection data. Based on a preset reconstruction algorithm, the second projection data is reconstructed to obtain initial reconstruction data; Based on a preset compression algorithm, the initial reconstructed data is compressed to obtain second reconstructed data, and the second reconstructed data is stored in a preset storage memory. Based on the preset compression algorithm, the second reconstructed data is decompressed to obtain the third reconstructed data; Based on a preset orthographic projection algorithm, the third reconstructed data is projected to obtain the third projected data; Perform the inverse operation of the preprocessing on the third projection data to obtain the fourth projection data; The first difference data is obtained by subtracting the fourth projection data from the first projection data; the first difference data is compressed based on a preset lossless compression algorithm to obtain the second difference data, and the second difference data is stored in the preset storage memory.

2. The compression method according to claim 1, characterized in that, The initial reconstructed data is compressed based on a preset compression algorithm to obtain second reconstructed data, including: Based on a preset mapping algorithm, the initial reconstructed data is mapped to an integer field to obtain the first reconstructed data; Based on the preset compression algorithm, the first reconstructed data is compressed to obtain the second reconstructed data.

3. The compression method according to claim 2, characterized in that, Based on the preset compression algorithm, the second reconstructed data is decompressed to obtain the third reconstructed data, including: Based on the preset compression algorithm, the second reconstructed data is decompressed to obtain the fourth reconstructed data; Based on the preset mapping algorithm, the fourth reconstructed data is subjected to a mapping inverse operation to obtain the third reconstructed data.

4. The compression method according to claim 2, characterized in that, The step of compressing the first reconstructed data based on the preset compression algorithm to obtain the second reconstructed data includes: Based on a preset image processing algorithm, the first reconstructed data is subjected to image noise reduction processing to obtain the fifth reconstructed data; the image processing algorithm is a three-dimensional block matching algorithm or a non-local averaging algorithm. Based on the preset compression algorithm, the fifth reconstructed data is compressed to obtain the second reconstructed data.

5. The compression method according to any one of claims 1, characterized in that, When the preset compression algorithm adopts a lossless compression algorithm, the initial reconstructed data is used as the third reconstructed data.

6. The compression method according to any one of claims 1-5, characterized in that, The preprocessing includes one or more of the following: dark field correction, gain correction, log calculation, scattering correction, and hardening correction.

7. The compression method according to any one of claims 1-5, characterized in that, The preset reconstruction algorithm is either a filtered back-projection algorithm or an iterative algorithm.

8. The compression method according to any one of claims 1-5, characterized in that, The preset compression algorithm is a lossless compression algorithm or a lossy compression algorithm. The lossless compression algorithm includes the JPEG2000 lossless compression algorithm and the JPEG-LS algorithm, and the lossy compression algorithm includes the SPIHT algorithm and the JPEG2000 lossy compression algorithm.

9. The compression method according to any one of claims 1-5, characterized in that, The preset orthographic projection algorithm is a point-driven projection algorithm, a line-driven projection algorithm, or a distance-driven projection algorithm.

10. A method for acquiring raw projection data based on CT compressed data, characterized in that, The method for obtaining the original projection data includes: The projection difference data and compressed reconstruction data are retrieved from memory, and the projection difference data is decompressed to obtain the first difference data. Based on a preset compression algorithm, the compressed and reconstructed data is decompressed to obtain the third reconstructed data; Based on a preset orthographic projection algorithm, the third reconstructed data is projected to obtain the third projected data; The third projection data is preprocessed in reverse order to obtain the fourth projection data; The original projection data is obtained by summing the first difference data and the fourth projection data.

Images