A method for generating normalized correction coefficients for PET imaging systems

Abstract

The present invention relates to a method for generating normalized correction coefficients for a PET imaging system. The method involves sequentially placing the same radioactive source at different axial positions of the PET system, with the radioactive sources at different positions overlapping axially. The method comprises: splicing and fusing the detection data from each position to obtain fused data with a complete axial field of view; physically correcting the fused data to obtain coincident data; calculating an inter-slice correction factor based on the number of intersections between response lines and the radioactive source; performing axial inter-slice correction on the coincident data based on the inter-slice correction factor to obtain corrected data; and analyzing the corrected data using a model-based normalized correction method to obtain the normalized correction coefficients. The method has the beneficial effect of being able to perform normalized correction of a long-axis PET system using the mechanical device and radioactive source of a short-axis PET system without requiring additional source costs or redesigning a complex mechanical device for fixing the radioactive source.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and in particular to a method for generating a normalized correction coefficient of a PET imaging system. Background Art

[0002] Positron Emission Tomography (PET) is a high-end nuclear medicine imaging diagnostic equipment. In actual operation, radionuclides (such as 18 F. 11 C, etc.) to label the metabolites and inject the nuclides into the human body, and then use PET to perform functional metabolic imaging on the patient to reflect the status of life metabolic activities, thereby achieving the purpose of diagnosis.

[0003] A PET imaging system typically consists of a ring-shaped detector system consisting of tens of thousands of detector units. Due to variations in geometry and performance, such as the crystal luminescence efficiency, crystal packaging, coupling between the crystal and PMT (photomultiplier tube) or silicon photomultiplier (SiPM), the electronics system, and the varying angles of incidence of the photon pairs, the detection efficiency of each detector unit varies. This results in its output not accurately reflecting the intensity of the input photon beam, which inevitably introduces artifacts during the reconstruction process. To accurately model the detection system, users must pre-calibrate the detector's detection efficiency, a process known as normalization.

[0004] Traditional normalization correction uses a uniform source placed at the center of the system (for example, a uniformly rotating rod source, a cylindrical barrel source filled with FDG solution, or a solid cylindrical Ge barrel source) to uniformly illuminate each detection unit, and uses a model-based normalization method to process the detection data. The intrinsic detection efficiency of the crystal, the normalization factor affected by system geometric factors, and the normalization factor related to the count rate are extracted from the data respectively, and finally they are integrated into a normalization correction factor reflecting the overall detection efficiency, and the measurement data is normalized and corrected.

[0005] Current PET systems have an increasingly larger axial field of view. To ensure uniform illumination of all detector crystals during normalized experiments, the axial dimensions of the sources have been correspondingly lengthened. This significantly increases their weight and makes their positioning and movement extremely difficult. Furthermore, the design and fabrication of the mechanisms that secure the source and support its stable, uniform reciprocating motion are extremely complex and expensive. Summary of the Invention

[0006] Technical problems to be solved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a method for generating normalization correction coefficients for a PET imaging system, which solves the technical problems that normalization correction currently requires additional source costs and the design and processing of the radiation source mechanical device is extremely complex and expensive.

[0008] Technical Solution

[0009] In order to achieve the above objectives, the main technical solutions adopted by the present invention include:

[0010] In a first aspect, the present invention provides a method for generating normalized correction coefficients for a PET imaging system. The method comprises placing the same radioactive source in succession at different axial positions of the PET system, where the radioactive sources at different positions overlap in the axial direction. The method comprises:

[0011] Modeling the PET acquisition process;

[0012] Calculate the acquisition time at each location based on the activity of the radioactive source at each location;

[0013] Collect detection data based on the collection time of each location;

[0014] The detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view;

[0015] Perform physical correction on the fused data to obtain conforming data;

[0016] The coordinates of the intersection of the response line and the radiation source are obtained based on the position information and system information recorded in each scan;

[0017] Determine whether the intersection exists based on the intersection coordinates, and calculate the interlayer correction factor based on the number of existing intersections;

[0018] Perform axial inter-slice correction on the conforming data based on the inter-slice correction factor to obtain corrected data;

[0019] The normalized correction coefficients were obtained by analyzing the correction data using a model-based normalized correction method.

[0020] Optionally, modeling the PET acquisition process includes: modeling the PET acquisition process as follows:

[0021]

[0022] in, represents the expected value of the detected data sinusoidal response line; r represents the variable index of the shortest distance between the response line and the radial center of the system, t represents the variable index of the azimuth angle of the response line, p represents the variable index of the axial fault plane to which the response line belongs, M represents the dimension of the sinusoidal graph in the r direction, S represents the dimension of the sinusoidal graph in the t direction, N represents the number of axial faults in the long-axis PET sinusoidal graph, and the single quote superscript represents the matrix transposition operation; x = [x1, x2, …, x j ,…,x Q ] ′ represents the unknown PET radioactivity distribution, j represents the number of the spatial point source, and Q represents the size of the radioactivity distribution image space; A = [A jrtp ] is the system matrix, r=[r 111 ,r 112 ,…,r rtp ,…,r MSN ] ′ represents the average value of random noise and scattered noise; is the normalized correction factor matrix.

[0023] Optionally, the acquisition time for each location is calculated based on the activity of the radioactive source at each location, including:

[0024] According to the formula Calculate the acquisition time for each position, where T = [T1, T2, ..., T i ,…T L ] represents the acquisition time of the radioactive source at each position, α=[α1,α2,…,α i ,…α L ] represents the initial activity of the radioactive source at each location before collection, L represents the number of locations where the radioactive source is collected, and the subscript i represents the number of the location where the radioactive source is placed.

[0025] Optionally, the detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view, including:

[0026] According to the formula The detection data are spliced ​​and fused to obtain fused data covering the complete axial field of view, where: To fuse data, For detection data.

[0027] Optionally, physical correction is performed on the fused data to obtain conforming data, including:

[0028] Perform attenuation correction, scatter correction, system response function correction, time drift correction, motion correction, noise correction and / or slice correction on the fused data to obtain the conforming data.

[0029] Optionally, the coordinates of the intersection of the response line and the radiation source are obtained based on the position information and system information recorded in each scan, including:

[0030] The response line at any radial distance number r on the outermost plane of the PET system can determine a plane coordinate system O with the system axis. hoz , the h-axis coincides with the response line, the z-axis is always parallel to the system axis and falls on the inner surface of the detector, and the h-axis and the z-axis are perpendicular to each other;

[0031] In O hoz The coordinates of the crystal pair corresponding to any response line are (z1,h1) and (z2,h2), and the coordinate origin o is defined on the crystal surface. R scan Indicates the annular inner diameter of the PET system, R rho Indicates the actual distance between the response line numbered r and the center of the field of view. z1 and z2 are determined according to the axial position of the crystal corresponding to the response line;

[0032] The formula for determining the response line according to the two crystal coordinates is:

[0033] h=k·z+b

[0034] in

[0035] The rotation trajectory of the radiation source at position i and the plane coordinate system O hoz There are two intersecting lines, which are two parallel straight lines in the coordinate system. The intersection line close to the positive direction of the coordinate axis h is defined as The intersection line close to the negative direction of the coordinate axis h is defined as

[0036] The intersection line at each position i is:

[0037]

[0038] in, Indicates intersection line and The axial coordinate close to the negative end of the coordinate axis z, Indicates intersection line and The axial coordinate close to the positive direction of the coordinate axis z;

[0039] Different radial positions, response lines and intersection lines and Axial intersection coordinates for:

[0040] Optionally, determining whether the intersection exists according to the intersection coordinates and calculating the interlayer correction factor according to the number of existing intersections includes:

[0041] if and All in If the response line is within the range, then the number of intersections between the response line and the radiation source line is N. i rp =2; if and Only one in Within the range, N i rp =1; if and None Within the range, N i rp =0;

[0042] The number of intersections N between all response lines of the sine graph of the detection data and the radiation source i rp , get N of the radiation source position i i rp ;

[0043] According to the formula Calculate the inter-layer correction factor, where W rp is the interlayer correction factor.

[0044] Optionally, performing axial inter-slice correction on the coincident data based on the inter-slice correction factor to obtain corrected data includes:

[0045] According to the formula Perform axial inter-slice correction on the coincident data, where To correct the data.

[0046] In a second aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the method for generating a normalized correction coefficient of a PET imaging system as described in any one of the first aspects above.

[0047] In a third aspect, the present invention provides a storage device comprising a storage medium and a processor, wherein the storage medium stores a computer program, and when the program is executed by the processor, the method for generating a normalized correction coefficient of a PET imaging system as described in any one of the first aspects above is implemented.

[0048] Beneficial effects

[0049] The beneficial effects of the present invention are as follows: a method for generating normalized correction coefficients for a PET imaging system of the present invention, when acquiring radioactive sources for generating normalized correction parameters, places radioactive sources at different positions of the long-axis PET system for multiple scans, ensuring that the radioactive source position is the same distance from the crystal surface each time, and that the multiple placed radioactive sources can cover the entire axial field of view. The multiple acquired radioactive source data are spliced ​​and fused, and analyzed using a model-based normalization method to obtain normalized correction parameters such as the detection efficiency, geometric correction factor, and crystal interference correction factor of the long-axis PET system. This normalized correction method has no restrictions on the length of the radioactive source and no longer requires a radioactive source that is sufficient to cover the entire axial field of view. The mechanical devices and radioactive sources of a conventional short-axis PET system can be used to perform normalized correction on the long-axis PET system, without requiring additional source costs or the redesign of a complex mechanical device for fixing the radioactive source. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic flow chart of a method for generating a normalized correction coefficient for a PET imaging system provided in an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of two position scans of a rotating rod source provided in an embodiment of the present invention;

[0052] Figure 3 The plane coordinate system O provided by the embodiment of the present invention hoz Schematic diagram;

[0053] Figure 4 A schematic diagram of the PET normalized correction coefficient crystal detection efficiency provided by an embodiment of the present invention;

[0054] Figure 5 A schematic diagram of the radial geometric correction factor for the PET normalization correction coefficient provided in an embodiment of the present invention;

[0055] Figure 6 Schematic diagram of the PET normalization correction coefficient and crystal interference correction factor provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] To better explain the present invention and facilitate understanding, the present invention is described in detail below with reference to the accompanying drawings and through specific embodiments. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0057] First, refer to Figure 1This embodiment provides a method for generating a normalized correction coefficient for a PET imaging system. The method includes placing the same radioactive source at different axial positions of the PET system, where the radioactive sources at different positions overlap in the axial direction. The method includes:

[0058] S1, Modeling the PET acquisition process.

[0059] S2, calculate the acquisition time at each location based on the activity of the radioactive source at each location.

[0060] S3, collecting detection data according to the collection time of each location.

[0061] S4, the detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view.

[0062] S5, performing physical correction on the fused data to obtain conforming data.

[0063] S6, obtaining the coordinates of the intersection of the response line and the radiation source based on the position information and system information recorded in each scan.

[0064] S7, judging whether the intersection exists according to the intersection coordinates, and calculating the inter-layer correction factor according to the number of existing intersections.

[0065] S8, performing axial inter-slice correction on the conforming data based on the inter-slice correction factor to obtain corrected data.

[0066] S9, using a model-based normalization correction method to analyze the correction data to obtain a normalized correction coefficient.

[0067] This paper proposes a novel method for generating normalized correction parameters. This method uses multiple acquisitions of radioactive sources and splices the data based on position. This method, combined with a model-based normalization method, derives normalized correction parameters, including detection efficiency, geometric correction factors, and crystal interference correction factors. Compared to the traditional method of using a long-axis radioactive source that covers the entire axial field of view to acquire raw data for normalized correction, the proposed method uses a short-axis radioactive source for multiple acquisitions and then splices and fuses the data to generate normalized correction coefficients. This method offers significant advantages: It allows the use of universal short-axis radioactive sources, eliminating the need for long-axis source design specific to long-axis PET systems and reducing source costs. It also eliminates the need to design and fabricate mechanical devices to support the stabilization and uniform rotation of the long-axis source, simplifying equipment design and reducing costs.

[0068] Optionally, modeling the PET acquisition process includes: modeling the PET acquisition process as follows:

[0069]

[0070] in, Represents the expected value of the detected data sinogram response line LOR (line of response). The detected data sinogram response line LOR is scaled by three-dimensional coordinates: r represents the variable index of the shortest distance between the response line LOR and the radial center of the system, t represents the variable index of the response line LOR azimuth, p represents the variable index of the axial fault plane to which the response line LOR belongs, M represents the dimension of the sinogram in the r direction, S represents the dimension of the sinogram in the t direction, N represents the number of axial faults in the long-axis PET sinogram, and the single quote superscript represents the matrix transpose operation. x=[x1,x2,…,x j ,…,x Q ] ′ represents the unknown PET radioactivity distribution, j represents the number of the spatial point source, and Q represents the size of the radioactivity distribution image space. jrtp ] is the system matrix, which mathematically expresses the probability that a point source j at a spatial location in the PET system will be detected by the line of response (LOR) with coordinates (r, t, p). This includes the geometric detection efficiency factor, the photon detection point spread factor, and the photon attenuation factor during the photon pair detection process. The geometric detection efficiency is expressed as the probability that a photon pair reaches the detector pair surface, which depends on the variation of the geometric detection efficiency of the detector pair relative to each pixel position; the photon point spread function is expressed as the response function of the detector crystal to the incident photon; and the photon attenuation is expressed as the probability that a photon pair will decay before being detected. r = [r 111 ,r 112 ,…,r rtp ,…,r MSN ] ′ Represents the average value of random noise and scattered noise. The normalization correction factor matrix is ​​used to correct for uneven detector crystal efficiency. The normalization parameters are independent of the time-of-flight measurement and therefore do not consider the time-of-flight dimension. To reduce the amount of statistical data required, model-based methods are often used to evaluate the normalization correction factor. The present invention aims to optimize the process for solving the normalization correction factor for long-axis PET systems.

[0071] Optionally, the acquisition time for each location is calculated based on the activity of the radioactive source at each location, including:

[0072] According to the formula Calculate the acquisition time for each position, where T = [T1, T2, ..., T i ,…T L ] represents the acquisition time of the radioactive source at each position, α=[α1,α2,…,α i ,…α L] represents the initial activity of the radioactive source at each location before collection, L represents the number of locations where the radioactive source is collected, and the subscript i represents the number of the location where the radioactive source is placed.

[0073] The collected data is defined as L represents the number of locations where the radioactive source is collected, and the superscript i represents the number of the location where the radioactive source is placed. i ,…T L ] represents the acquisition time of the radioactive source at each location. Since the radioactive source decays irreversibly over time and its activity decreases continuously, the acquisition time at each location needs to be corrected in advance according to the activity. The initial activity of the radioactive source at each location is recorded before acquisition as α = [α1, α2, …, α i ,…α L ], calculate the acquisition time at each location and ensure that the acquisition counts at each location are roughly the same.

[0074] Optionally, the detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view, including:

[0075] According to the formula The detection data are spliced ​​and fused to obtain fused data covering the complete axial field of view, where: To fuse data, For detection data.

[0076] The detection data at each position are spliced ​​and fused to obtain a sinusoidal diagram of the detection data covering the complete axial field of view.

[0077] Optionally, physical correction is performed on the fused data to obtain conforming data, including:

[0078] Perform attenuation correction, scatter correction, system response function correction, time drift correction, motion correction, noise correction and / or slice correction on the fused data to obtain the conforming data.

[0079] Considering that attenuation, dead time, scattering or random events are inevitable during the scanning process, it is necessary to perform physical correction on the combined data to try to eliminate the effects of attenuation, dead time, scattering or random events, so as to obtain more accurate combined data.

[0080] The axial inter-slice correction factor is calculated by estimating the coincidence of the axial crystal pair when the radiation source is placed at different positions in the PET system. Taking the rotating rod source as an example, first, since the rod source is at a fixed radius R in the center of the field of view, rotFor rotation, LOR does not differ at different circumferential angles, so the effect of angle on the axial interslice correction factor is not considered. Secondly, for LOR at different radial positions, the distance between the radiation source and the crystal surface is different, so the axial coincidence at different radial positions needs to be considered.

[0081] Optionally, the coordinates of the intersection of the response line and the radiation source are obtained based on the position information and system information recorded in each scan, including:

[0082] The response line at any radial distance r on the outermost plane of the PET system can determine a plane coordinate system O with the system axis. hoz ,like Figure 3 As shown. The h-axis coincides with the response line, the z-axis is always parallel to the system axis and falls on the inner surface of the detector. The h-axis and z-axis are perpendicular to each other and the direction is user-defined.

[0083] In o hoz The coordinates of the crystal pair corresponding to any response line are (z1, h1) and (z2, h2), the coordinate origin o is defined on the crystal surface, and h1 is set to 0; R scan Indicates the annular inner diameter of the PET system, R rho Indicates the actual distance between the response line numbered r and the center of the field of view. z1 and z2 are determined according to the axial position of the crystal corresponding to the response line;

[0084] The formula for determining the response line according to the two crystal coordinates is:

[0085] h=k·z+b

[0086] in

[0087] The rotation trajectory of the radiation source at position i and the plane coordinate system O hoz There are two intersecting lines, which are two parallel straight lines in the coordinate system. The intersection line close to the positive direction of the coordinate axis h is defined as The intersection line close to the negative direction of the coordinate axis h is defined as

[0088] The intersection line at each position i is:

[0089]

[0090] in, Indicates intersection line and The axial coordinate near the negative end of the coordinate axis z, z i+ Indicates intersection line and The axial coordinate close to the positive direction of the coordinate axis z;

[0091] For different radial positions, the above formulas can be combined to obtain the response line and the intersection line and Axial intersection coordinates for:

[0092] Optionally, determining whether the intersection exists according to the intersection coordinates and calculating the interlayer correction factor according to the number of existing intersections includes:

[0093] if and All in If the response line is within the range, then the number of intersections between the response line and the radiation source line is N. i rp =2; if and Only one in Within the range, N i rp =1; if and None Within the range, N i rp =0;

[0094] The number of intersections N between all response lines of the sine graph of the detection data and the radiation source i rp , get N of the radiation source position i i rp ;

[0095] According to the formula Calculate the inter-layer correction factor, where W rp is the interlayer correction factor.

[0096] N of all position rod sources i rp The final axial interlayer correction factor is obtained by summing them.

[0097] Optionally, performing axial inter-slice correction on the coincident data based on the inter-slice correction factor to obtain corrected data includes:

[0098] According to the formula Perform axial inter-slice correction on the coincident data, where To correct the data.

[0099] Since the commonly used short-axis radiation source cannot fully fill the long-axis PET system, a segmented detection method is used to obtain normalized and corrected raw data of the complete axial field of view. This method will result in uneven acquisition of the number of axial inter-slice response lines (LORs), and it is necessary to perform inter-slice count correction on the detection data according to various situations. The inter-slice correction factor is defined as Wrp , the dimension of the correction factor is M×N.

[0100] Dividing the stitched data from the sinogram domain by the axial inter-slice correction factor can solve the data duplication problem caused by the axial intersection of the two acquisitions, thereby obtaining long-axis data with uniform counts.

[0101] For data Y sum-c′ Using traditional model-based methods, the normalized correction factor A is obtained by analysis. norm The corrected and spliced ​​detection data can be equivalent to the data of a rotating rod source that uniformly fills the long-axis axial field of view, and the traditional model-based normalization correction method can be directly applied.

[0102] In addition, the radiation source in the present application can be a rotating rod source, a hollow cylindrical source, and a planar source. When the radiation source is a rotating rod source, the rotation center and the rotation radius must be the same.

[0103] Figure 2 Schematic diagram of a scan performed in two steps using a rotating rod source, with overlapping areas on both sides. Figure 4 To utilize the intrinsic detection efficiency obtained by the present invention, Figure 5 is the radial geometric correction factor obtained by using the present invention; Figure 6 is the crystal interference correction factor obtained by using the present invention.

[0104] Below is an annular inner diameter R scan The implementation steps of the present invention are described using a rod source normalization experiment performed on a PET system with a diameter of 370 mm and an axial field of view of 518 mm as an example. The steps include:

[0105] Prepare a Ge rod source with a length of 300 mm and a diameter of 2 mm. Rod sources are commonly used to obtain normalized correction parameters in short-axis PET systems. In this experiment, the rod source was placed at two locations for acquisition, depending on its length.

[0106] Use a robotic arm to fix the rod source and rotate it at a constant speed around the central axis of the PET system's acquisition field of view. When placing the rotating rod source, it is necessary to ensure that the rod sources on both sides have the same rotation center, and the rotation radius is set to R rot =330mm, and the two acquisition positions have overlapping parts in the axial field of view, ensuring that the two acquisitions cover the complete axial field of view of the PET system.

[0107] Place the rod source in the first position, that is, fix it to one side of the PET system by the cantilever, ensuring that the rod source covers both the axial edge of the PET system and the center of the axial field of view, and record The rod source rotates uniformly at a speed of 20 revolutions per minute within the PET system acquisition field of view. The acquisition time T1 is set to 129600s. The initial activity is recorded as α1 = 300.73 μmCi. The acquisition data Y 1 .

[0108] After the acquisition at the first position is completed, remove the rod source and the fixing device, and place the rod source at the second position, that is, fix the rod source on the other side of the PET system with the cantilever, ensuring that the rod source covers the other axial edge of the PET system and the center of the axial field of view at the same time, and record The same rotation speed of 20 revolutions per minute is uniformly moving inside the system. At this time, the initial activity collected is α2=299.56μmCi. According to the formula The total time of the second acquisition is calculated to be T2 = 130106s, and the acquisition data Y 2 .

[0109] After the two positions are collected, the collected data Y 1 and Y 2 Splicing and fusion to get Y sum . Perform physical correction to get the data Y sum-c .

[0110] Substitute the location information and system information recorded from the two scans into the formula:

[0111]

[0112] The coordinates of the intersection of the LOR and the radiation source at any radial position r can be obtained:

[0113]

[0114] R rho is the actual distance between the LOR with radial distance number r and the center of the field of view. According to the LOR position coordinates, z1 and z1 can be determined. Traverse all LORs and determine and Is it in the range of [0,293.5]? Get N 1 rp ; then judge and Is it within the range of [230.0,518.0]? Get N 2 rp Based on the formula Calculate the accurate interlayer correction factor W rp .

[0115] Based on the formula For data Y sum-cPerform axial inter-slice correction. This way, you don’t need a long rod source that covers the entire axial range of the PET system to get an equivalent full axial field of view consistent with the data Y sum-c′ .

[0116] The model-based normalization correction method is used to sum-c′ By analyzing, we can get the normalized correction coefficients such as crystal detection efficiency, radial geometry correction factor and crystal interference correction factor. The results are shown in Figure 4 、 Figure 5 and Figure 6 The new method for generating the normalized correction coefficient of the PET imaging system proposed in the present invention effectively simplifies the equipment design and reduces the source manufacturing cost.

[0117] In a second aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements a method for generating a normalized correction coefficient of a PET imaging system as described in any one of the first aspects above.

[0118] In a third aspect, an embodiment of the present invention provides a storage device comprising a storage medium and a processor, wherein the storage medium stores a computer program, and when the program is executed by the processor, the method for generating a normalized correction coefficient of a PET imaging system as described in any one of the first aspects above is implemented.

[0119] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0120] Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if such modifications and variations fall within the scope of the claims and their equivalents, the present invention shall also include such modifications and variations.

[0121] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for generating a normalized correction coefficient for a PET imaging system, characterized in that: The same radioactive source is placed successively at different axial positions of the PET system, wherein the radioactive sources at different positions have overlapping portions in the axial direction. The method comprises: Modeling the PET acquisition process; Calculate the acquisition time at each location based on the activity of the radioactive source at each location; Collect detection data based on the collection time of each location; The detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view; Perform physical correction on the fused data to obtain conforming data; The coordinates of the intersection of the response line and the radiation source are obtained based on the position information and system information recorded in each scan; Determine whether the intersection exists based on the intersection coordinates, and calculate the interlayer correction factor based on the number of existing intersections; Perform axial inter-slice correction on the conforming data based on the inter-slice correction factor to obtain corrected data; The normalized correction coefficients were obtained by analyzing the correction data using a model-based normalized correction method.

2. The method for generating a normalized correction coefficient for a PET imaging system according to claim 1, wherein: Modeling the PET acquisition process includes: Modeling the PET acquisition process as follows: in, represents the expected value of the detected data sinusoidal response line; r represents the variable index of the shortest distance between the response line and the radial center of the system, t represents the variable index of the azimuth angle of the response line, p represents the variable index of the axial fault plane to which the response line belongs, M represents the dimension of the sinusoidal graph in the r direction, S represents the dimension of the sinusoidal graph in the t direction, N represents the number of axial faults in the long-axis PET sinusoidal graph, and the single quote superscript represents the matrix transposition operation; x = [x1, x2, …, x j ,…,x Q ] ′ represents the unknown PET radioactivity distribution, j represents the number of the spatial point source, and Q represents the size of the radioactivity distribution image space; A = [A jrtp ] is the system matrix, r=[r 111 ,r 112 ,…,r rtp ,…,r MSN ] ′ represents the average value of random noise and scattered noise; is the normalized correction factor matrix.

3. The method for generating a normalized correction coefficient for a PET imaging system according to claim 2, wherein: Calculate the acquisition time at each location based on the activity of the radioactive source at each location, including: According to the formula Calculate the acquisition time for each position, where T = [T1, T2, ..., T i ,…T L ] represents the acquisition time of the radioactive source at each position, α=[α1,α2,Ω,α i ,…α L ] represents the initial activity of the radioactive source at each location before collection, L represents the number of locations where the radioactive source is collected, and the subscript i represents the number of the location where the radioactive source is placed.

4. The method for generating a normalized correction coefficient for a PET imaging system according to claim 3, wherein: The detection data at each position are stitched and fused to obtain fused data covering the complete axial field of view, including: According to the formula The detection data are spliced ​​and fused to obtain fused data covering the complete axial field of view, where: To fuse data, For detection data.

5. The method for generating a normalized correction coefficient for a PET imaging system according to claim 4, wherein: Physical correction is performed on the fused data to obtain conforming data, including: Perform attenuation correction, scatter correction, system response function correction, time drift correction, motion correction, noise correction and / or slice correction on the fused data to obtain the conforming data.

6. The method for generating a normalized correction coefficient for a PET imaging system according to claim 5, wherein: The coordinates of the intersection of the response line and the radiation source are obtained based on the position information and system information recorded in each scan, including: The response line at any radial distance number r on the outermost plane of the PET system can determine a plane coordinate system O with the system axis. hoz , the h-axis coincides with the response line, the z-axis is always parallel to the system axis and falls on the inner surface of the detector, and the h-axis and the z-axis are perpendicular to each other; In O hoz The coordinates of the crystal pair corresponding to any response line are (z1,h1) and (z2,h2), and the coordinate origin o is defined on the crystal surface. R scan Indicates the annular inner diameter of the PET system, R rho Indicates the actual distance between the response line numbered r and the center of the field of view. z1 and z2 are determined according to the axial position of the crystal corresponding to the response line; The formula for determining the response line according to the two crystal coordinates is: h=k·z+b in The rotation trajectory of the radiation source at position i and the plane coordinate system O hoz There are two intersecting lines, which are two parallel straight lines in the coordinate system. The intersection line close to the positive direction of the coordinate axis h is defined as The intersection line close to the negative direction of the coordinate axis h is defined as The intersection line at each position i is: in, Indicates intersection line and The axial coordinate close to the negative end of the coordinate axis z, Indicates intersection line and The axial coordinate close to the positive direction of the coordinate axis z; Different radial positions, response lines and intersection lines and Axial intersection coordinates for:

7. The method for generating a normalized correction coefficient for a PET imaging system according to claim 6, wherein: Determine whether the intersection exists based on the intersection coordinates, and calculate the interlayer correction factor based on the number of existing intersections, including: if and All in If the response line is within the range, then the number of intersections between the response line and the radiation source line is N. i rp =2; if and Only one in Within the range, N i rp =1; if and None Within the range, N i rp =0; The number of intersections N between all response lines of the sine graph of the detection data and the radiation source i rp , get N of the radiation source position i i rp ; According to the formula Calculate the inter-layer correction factor, where W rp is the interlayer correction factor.

8. The method for generating a normalized correction coefficient for a PET imaging system according to claim 7, wherein: Perform axial inter-slice correction on the coincident data based on the inter-slice correction factor to obtain corrected data, including: According to the formula r=1,…,M;t=1,…,S;p=1,…,N Perform axial inter-slice correction on the coincident data, where To correct the data.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for generating a normalized correction coefficient for a PET imaging system according to any one of claims 1 to 8 is implemented.

10. A storage device comprising a storage medium and a processor, wherein the storage medium stores a computer program, wherein: When the processor executes the computer program, the method for generating a normalized correction coefficient for a PET imaging system according to any one of claims 1 to 8 is implemented.

Images