Phase-contrast multi-material artifact removal method, electronic device, and storage medium
By acquiring and separating multi-material projection data in phase-contrast CT, and employing a two-step method and phase recovery technology, the problem of imaging artifacts in multi-material samples was solved, achieving accurate multi-material structure restoration and image quality improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI UNITED IMAGING RES INST OF INTELLIGENT IMAGING
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-31
AI Technical Summary
Existing phase-contrast CT algorithms can only select one interface for precise processing when dealing with multi-material samples, while other interfaces will exhibit imaging artifacts, making it difficult to accurately restore the true structure of multi-material samples.
By acquiring the initial projection data of the target phantom, the projection thickness of each initial material is extracted, and reconstruction is performed based on these thicknesses. A two-step method is used to separate the projection thickness of each initial material, and the FDK filtering back projection algorithm and phase recovery technology are used to eliminate artifacts.
It achieves accurate restoration of multi-material samples, eliminates artifacts at the source, significantly improves image quality, and enhances detail resolution and contrast.
Smart Images
Figure CN122492895A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of medical technology, and in particular to a method for eliminating phase-contrast multi-material artifacts, electronic devices, and storage media. Background Technology
[0002] In phase-contrast CT, the raw intensity images measured by the detector cannot directly provide quantitative information. Therefore, phase retrieval algorithms are needed to extract projected phase, attenuation, or thickness information from the intensity information measured by the detector. However, the currently widely used mainstream algorithms are only applicable to objects of the same material. When a sample has multiple interfaces of different materials, only one interface can be accurately processed, while imaging artifacts will appear on the other interfaces. Summary of the Invention
[0003] The main technical problem addressed in this application is to provide a phase-contrast multi-material artifact elimination method, electronic device, and storage medium that can accurately restore the true structure of each initial material of the target phantom and achieve artifact elimination.
[0004] To address the aforementioned technical problems, the first aspect of this application provides a method for eliminating phase-contrast multi-material artifacts. The method includes: acquiring initial projection data obtained by scanning a target phantom; wherein the target phantom includes at least two initial materials, and the initial projection data is obtained by electromagnetic waves passing through the target phantom; extracting the projection thickness corresponding to each initial material from the initial projection data; and reconstructing the target phantom based on the projection thickness corresponding to each initial material to obtain a target reconstruction image corresponding to the target phantom.
[0005] To address the aforementioned technical problems, a second aspect of this application provides an electronic device comprising a memory and a processor. The memory stores program instructions, and the processor executes the program instructions to implement the aforementioned phase-contrast multi-material artifact elimination method.
[0006] To address the aforementioned technical problems, a third aspect of this application provides a computer-readable storage medium for storing program instructions that can be executed to implement the aforementioned phase-contrast multi-material artifact elimination method.
[0007] In the above technical solution, the projection thickness corresponding to each initial material is an accurate projection thickness separated from the initial projection data. The signals of each initial material no longer interfere with each other. Image reconstruction based on the projection thickness corresponding to each initial material can accurately restore their true structure, and artifacts are eliminated from the source. Attached Figure Description
[0008] Figure 1 This is a schematic flowchart of an embodiment of the phase-contrast multi-material artifact elimination method provided in this application; Figure 2This is a schematic diagram of an embodiment of the target reconstruction image corresponding to the target phantom provided in this application; Figure 3 This is a schematic diagram of another embodiment of the target reconstruction image corresponding to the target phantom provided in this application; Figure 4 yes Figure 1 The flowchart of step S12 shown is a schematic diagram of one embodiment; Figure 5 yes Figure 4 The flowchart of step S41 shown is a schematic diagram of one embodiment. Figure 6 yes Figure 5 The flowchart of step S51 shown is a schematic diagram of an embodiment. Figure 7 yes Figure 1 The flowchart of step S13 shown is a schematic diagram of one embodiment. Figure 8 This is a schematic diagram of the structure of an embodiment of the phase-contrast multi-material artifact elimination device provided in this application; Figure 9 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application; Figure 10 This is a schematic diagram of an embodiment of the computer-readable storage medium provided in this application. Detailed Implementation
[0009] The embodiments of this application will now be described in detail with reference to the accompanying drawings.
[0010] In the following description, specific details such as particular system architectures, interfaces, and technologies are presented for illustrative purposes rather than for limiting purposes, in order to provide a thorough understanding of this application.
[0011] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, "many" in this document means two or more. Moreover, the term "at least one" in this document means any combination of at least two of any one or more of a plurality of objects. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.
[0012] Please see Figure 1 , Figure 1 This is a schematic flowchart of an embodiment of the phase-contrast multi-material artifact elimination method provided in this application. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily reflect that outcome. Figure 1 The illustrated process sequence is limited. For example... Figure 1 As shown, this embodiment includes: In traditional computed tomography (CT), the imaging principle is based on the intensity attenuation of X-rays as they pass through an object due to absorption. Different tissues or materials within an object have varying absorption capacities for X-rays, and the intensity changes received by the detector reflect the density distribution within the object, thus obtaining a structural image. However, for soft tissues (such as cartilage, blood vessels, and some biological tissues) or materials composed of light elements and exhibiting minimal differences in X-ray absorption, the imaging contrast is extremely low, making it difficult to achieve accurate structural resolution.
[0013] Phase-contrast computed tomography (PCT) employs a completely different physical mechanism based on the wave nature of X-rays. When X-rays pass through an object, not only does their intensity attenuate, but their wavefront phase also undergoes an extremely small shift due to interaction with the matter. PCT can detect and reconstruct this phase change through phase-sensitive imaging mechanisms such as interference, diffraction, or propagation. Compared to absorption effects, the phase modulation effect of light elements is typically thousands of times stronger. Therefore, PCT can significantly enhance the imaging performance of soft tissues or low-density materials (such as biological soft tissues, polymers, and composite materials) with higher contrast, clearly revealing fine structures that are difficult to resolve with conventional absorption CT. Thus, it has broad application prospects in medical imaging, materials testing, and microstructure observation.
[0014] It's important to note that in phase-contrast CT, the raw intensity images measured by the detector cannot directly provide quantitative information. Therefore, phase retrieval algorithms are needed to extract projection phase, attenuation, or thickness information from the intensity information measured by the detector. However, currently widely used mainstream algorithms are only applicable to objects of the same material. When a sample contains multiple interfaces of different materials, only one interface can be accurately processed, while the remaining interfaces will exhibit imaging artifacts. For example, for projection images of objects containing both bone and soft tissue, if the parameters are set to bone values, the edge enhancement effect cannot be fully corrected; conversely, if they are set to soft tissue values, the bone region will exhibit overly smoothed blurring after recovery.
[0015] Step S11: Obtain the initial projection data obtained by scanning the target phantom.
[0016] In this embodiment, initial projection data obtained by scanning the target phantom is acquired. The target phantom comprises at least two initial materials, and the initial projection data is obtained by electromagnetic waves passing through the target phantom. Specifically, the target phantom is placed between an electromagnetic wave (e.g., X-ray) emission source and a detector. The electromagnetic wave emission source and the detector rotate synchronously 360° around the target phantom, and the original intensity information of the electromagnetic waves after passing through the target phantom and propagating freely is collected at different angles (θ), resulting in initial projection data I(r,z,θ) containing spatial position (r) and propagation distance (z).
[0017] It should be noted that since the target phantom includes at least two initial materials, it is a multi-material phantom. Therefore, the initial projection data obtained by electromagnetic waves passing through the target phantom includes the attenuation, phase shift, and thickness information of each initial material, providing basic data for subsequent phase recovery and material separation.
[0018] This does not limit the initial materials included in the target phantom. For example, such as Figure 2 As shown, Figure 2 This is a schematic diagram of an embodiment of the target phantom corresponding to the target reconstruction image provided in this application. The target phantom consists of four aluminum balls placed in a water tray, and the two initial materials included in the target phantom are the water tray and the aluminum balls. For example, as... Figure 3 As shown, Figure 3 This is a schematic diagram of another embodiment of the target phantom corresponding to the target reconstruction image provided in this application. The target phantom is a skull, and the two initial materials included in the target phantom can be bone and soft tissue, respectively. Furthermore, the number of initial materials included in the target phantom is not limited and can be specifically set according to actual usage needs.
[0019] Furthermore, the phase-contrast imaging system uses a micro-focus electromagnetic wave source with a focal size of 50-100µm, a detector resolution of 20-100µm, a detector distance of 1-2 meters from the light source, and a detector distance of 0.5-1 meter from the target phantom. The discrete Fresnel number F can range from 4 to 100. For example, with a detector resolution of 50µm, a propagation distance of 1 meter, and a magnification of 2x, F = 17.65.
[0020] Step S12: Extract the projected thickness of each initial material from the initial projection data.
[0021] In this embodiment, the projected thickness corresponding to each initial material is extracted from the initial projection data. Since the target phantom includes at least two initial materials, it is a multi-material phantom; therefore, the initial projection data obtained by electromagnetic waves passing through the target phantom is the sum of the effects of all initial materials on the electromagnetic waves, and includes the attenuation, phase shift, and thickness information of each initial material. Thus, the projected thickness corresponding to each initial material can be extracted from the initial projection data.
[0022] The projected thickness of each initial material is extracted from the initial projection data to separate the accurate projected thickness of each initial material, or in other words, to separate the accurate projection information of each initial material from the initial projection data. The original projection data is a superposition of signals from multiple materials, which can lead to problems such as stripe artifacts, blurred edges, and grayscale distortion in the reconstructed image. Separating the accurate projected thickness of each initial material from the initial projection data prevents interference between the signals of each initial material, allowing for accurate reconstruction of their respective true structures during subsequent image reconstruction, thus eliminating artifacts at their source. Furthermore, under the traditional single-material assumption, the signals of some initial materials (e.g., soft tissue) can be masked by the signals of other dominant initial materials (e.g., bone), resulting in blurred details. Separating the accurate projected thickness of each initial material from the initial projection data allows for the separate extraction of the projected thickness of the masked initial materials, clearly revealing even minor differences in their phase shifts, making previously blurred details (e.g., cartilage, blood vessel walls) clearly discernible.
[0023] In one embodiment, before extracting the projected thickness corresponding to each initial material from the initial projection data, the initial projection data is cleaned; the data cleaning includes at least one of the following: dead pixel removal, data correction, and data format conversion. By cleaning the initial projection data, it is ensured that the initial projection data used for subsequent phase recovery and initial material separation is clean, accurate, and effective.
[0024] The detector (equivalent to the digital camera sensor of the system) may have some permanently damaged or unresponsive pixels; these are dead pixels. Dead pixel removal involves determining the coordinates of the dead pixels and replacing their erroneous data with the average value of the surrounding normal pixels. This process is identical for the same detector. It essentially repairs dead pixels in a photograph, preventing these fixed, unrealistic black or white spots from interfering with subsequent accurate measurements and ensuring the integrity of the initial projection data after dead pixel removal.
[0025] In CT imaging, initial projection data refers to the raw signal directly acquired by the detector without any processing. However, initial projection data contains two types of irrelevant information: one is the detector's own electronic noise and environmental background interference (unrelated to electromagnetic waves and the target phantom); the other is the inhomogeneity of electromagnetic wave intensity and the sensitivity differences among detector pixels. If the initial projection data is used directly for subsequent processing, it will lead to false signals in the image (such as specks caused by detector noise, bright and dark stripes caused by electromagnetic wave inhomogeneity), completely obscuring the true structure of the target phantom. Therefore, by performing data correction on the initial projection data to eliminate all interference unrelated to the target phantom, such as detector noise, electromagnetic wave inhomogeneity, and pixel sensitivity differences, only the signal changes caused by the attenuation / phase shift of electromagnetic waves by the target phantom are retained, thus obtaining initial projection data that truly reflects the intensity changes of electromagnetic waves after passing through the target phantom.
[0026] Converting the data format of the initial projection data allows it to be linearly transformed into the required information on the target phantom's ability to block (absorb) electromagnetic waves, enabling the initial projection data to be used for subsequent phase recovery and initial material separation.
[0027] In one specific implementation, data correction includes dark-field correction. Specifically, a dark-field image is acquired, which is captured in complete darkness and records the detector's own electronic noise and background interference. By subtracting the dark-field image data from the initial projection data, the detector's own electronic noise and background interference can be eliminated. Through dark-field correction, the detector's own electronic noise and background interference, which are unrelated to electromagnetic waves and the target phantom, are removed from the initial projection data.
[0028] In one specific implementation, data correction includes flat-field correction. Specifically, a flat-field image is acquired, which is captured without placing the target phantom and records the inhomogeneity of the electromagnetic waves and the sensitivity differences between individual pixels of the detector. By dividing the initial projection data by the flat-field image data, the electromagnetic wave inhomogeneity and the sensitivity differences between individual pixels of the detector can be corrected.
[0029] In one specific implementation, data correction includes dark-field correction and flat-field correction. By performing dark-field and flat-field correction on the initial projection data, electronic noise and background interference from the detector itself, which are unrelated to electromagnetic waves and the target phantom, are removed from the initial projection data. This also corrects electromagnetic wave inhomogeneity and sensitivity differences between individual pixels of the detector. This is equivalent to performing a zero-calibration on the imaging system, aiming to eliminate inherent equipment defects and environmental interference, ensuring that the final initial projection data only reflects the characteristics of the target phantom itself.
[0030] In one specific implementation, the initial projection data can be formatted using post-log transformation, that is, a mathematical operation called taking the negative logarithm can be performed on the initial projection data.
[0031] In other specific implementations, the initial projection data can also be converted using methods such as standardized logarithmic transformation, base-10 logarithmic transformation, linearized intensity ratio transformation, or iterative absorption coefficient inversion. No specific method is specified here.
[0032] Step S13: Reconstruct the target image corresponding to the target phantom based on the projected thickness of each initial material.
[0033] In this embodiment, the target reconstructed image of the target phantom is obtained by reconstructing the image based on the projected thickness corresponding to each initial material. The projected thickness corresponding to each initial material is an accurate projection thickness separated from the initial projection data. The signals of each initial material no longer interfere with each other. Image reconstruction based on the projected thickness corresponding to each initial material can accurately restore their true structure, and artifacts are eliminated at the source. In addition, the projected thickness corresponding to each initial material is separated from the initial projection data, rather than an approximation under the single-material assumption. The grayscale value of the target reconstructed image based on the projected thickness corresponding to each initial material is linearly related to parameters such as the true thickness and electron density of the material.
[0034] Please see Figure 4 , Figure 4 yes Figure 1 The flowchart shown is a schematic diagram of one embodiment of step S12. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow the same pattern. Figure 4 The illustrated process sequence is limited. For example... Figure 4 As shown, this embodiment includes: Step S41: Extract the projected thickness of the first target material from the initial projection data.
[0035] In this embodiment, the projected thickness corresponding to the first target material is extracted from the initial projection data; wherein, the first target material is an initial material. That is, based on the initial projection data, the projected thickness of an initial material is first determined.
[0036] In one embodiment, the first target material may be any one of at least two initial materials included in the target phantom.
[0037] In other embodiments, the first target material may be the initial material among the at least two initial materials included in the target phantom that has higher imaging contrast, clearer structure, or is easier to accurately segment. For example, if the target phantom is a skull, and the skull includes bone and soft tissue as the two initial materials, the skull can be selected as the first target material.
[0038] Step S42: Using the initial projection data and the projection thickness corresponding to the first target material, determine the projection thickness corresponding to the remaining initial material.
[0039] In this embodiment, the projection thickness of the remaining initial materials is determined using the initial projection data and the projection thickness corresponding to the first target material. That is, based on the initial projection data and the projection thickness of one initial material, the projection thickness of the remaining initial materials is determined; i.e., the projection thickness of each initial material is separated from the initial projection data in two steps. The first step separates the projection thickness of the first target material, which is equivalent to removing the interference of the first target material from the initial projection data. The second step derives the projection thickness of the remaining initial materials, which completely decouples the projection thicknesses of the two initial materials, preventing mutual interference during reconstruction and eliminating artifacts at the source. Furthermore, after separation using the two-step method, the projection thickness of the initial materials whose signals are easily masked is extracted separately, and the slight differences in their phase shift can be clearly presented, making originally blurry details clearly discernible, and improving contrast and detail resolution several times over.
[0040] The projected thicknesses of the remaining initial materials are directly derived using the initial projection data and the projected thicknesses of the first target material. This direct derivation method decouples the projected thicknesses of different initial materials, preventing misidentification of one initial material signal as another during subsequent reconstruction and eliminating artifacts at their source. Furthermore, this direct derivation method requires no additional iterative optimization or complex physical model fitting, resulting in low computational cost and high efficiency.
[0041] In one embodiment, the target phantom comprises two initial materials. The first step is to separate the projected thickness of one initial material, and the second step is to derive the projected thickness of the other initial material.
[0042] Please see Figure 5 , Figure 5 yes Figure 4 The diagram shows a flowchart of one embodiment of step S41. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow the same pattern. Figure 5 The illustrated process sequence is limited. For example... Figure 5 As shown, the target phantom comprises two initial materials, and this embodiment includes: Step S51: Reconstruct using the initial projection data to obtain the initial reconstructed image.
[0043] In this embodiment, initial projection data is used for reconstruction to obtain an initial reconstructed image. The two-dimensional initial projection data is reconstructed into a three-dimensional initial reconstructed image, which serves as the starting point for subsequent segmentation and iteration.
[0044] In one embodiment, the FDK filter back projection algorithm can be used to reconstruct the image based on the initial projection data to obtain the initial reconstructed image.
[0045] In one embodiment, reconstruction can be performed directly based on the initial projection data to obtain the initial reconstructed image.
[0046] Of course, in other embodiments, the initial reconstructed image can also be obtained by reconstructing the projection data obtained by performing single initial material phase recovery on the initial projection data.
[0047] Step S52: From the initial reconstructed image, segment the image region belonging to the first target material, and use it as the segmented image of the first target material.
[0048] In this embodiment, the image region belonging to the first target material is segmented from the initial reconstructed image and used as the segmented image of the first target material. The image region belonging to the first target material is segmented from the initial reconstructed image to obtain a segmented image containing only the first target material; that is, the first target material is first separated from the initial reconstructed image of the mixed materials.
[0049] In one embodiment, a hard thresholding algorithm can be used to segment the image region belonging to the first target material from the initial reconstructed image, which is then used as the segmented image of the first target material. Specifically, a segmentation threshold is set, and the relationship between the pixel value of each pixel in the initial reconstructed image and the segmentation threshold is determined. The image region formed by pixels with pixel values greater than the segmentation threshold is then used as the segmented image of the first target material. The size of the segmentation threshold is not limited and can be set as needed.
[0050] Of course, in other implementations, a soft thresholding segmentation algorithm can also be used to segment the image region belonging to the first target material from the initial reconstructed image, and use it as the segmented image of the first target material.
[0051] In one specific implementation, the soft threshold segmentation formula is as follows:
[0052]
[0053]
[0054] in, This represents the pixel values in the initial reconstructed image; This represents the linear attenuation coefficient of the first target material; This represents the linear decay coefficient of the remaining initial material; This represents the relative material coefficient corresponding to each pixel in the segmented image; This represents the image region corresponding to the first target material.
[0055] Step S53: Perform forward projection on the segmented image of the first target material to obtain the projected thickness of the first target material.
[0056] In this embodiment, a forward projection is performed on the segmented image of the first target material to obtain the projected thickness of the first target material. The forward projection of the segmented image of the first target material projects the three-dimensional segmented image back onto a two-dimensional plane, obtaining the projected thickness L1(r,θ), that is, the thickness distribution of the first target material at each projection angle. This quantifies the contribution of the first target material at each projection angle and provides a basis for calculating the projected thickness of another initial material.
[0057] At this point, the projection thickness of the remaining initial materials is determined using the initial projection data and the projection thickness corresponding to the first target material. Specifically, the projection thickness of the remaining initial materials is determined by directly deriving the projection thickness from the initial projection data and the projection thickness corresponding to the first target material. This direct derivation method decouples the projection thicknesses of different initial materials, preventing misidentification of one initial material signal as another during subsequent reconstruction, thus eliminating artifact problems at their source. Furthermore, this direct derivation method requires no additional iterative optimization or complex physical model fitting, resulting in low computational cost and high computational efficiency.
[0058] In one embodiment, the target phantom comprises two initial materials. Using the initial projection data and the projection thickness corresponding to the first target material, the projection thickness corresponding to the remaining initial materials is determined. Specifically, this can be achieved by subtracting the linear attenuation, phase shift, and propagation effects of the first target material from the initial projection data to obtain the projection thickness corresponding to the remaining initial materials. Combining the attenuation and phase shift effects of electromagnetic waves and their propagation effects in the medium, and utilizing the frequency domain computational capabilities of Fourier transform, the projection thickness of the remaining initial materials is directly calculated by subtracting the attenuation, phase shift, and propagation effects of the first target material from the initial projection data containing multi-material superposition information and the known projection thickness of the first target material, within the physical framework of propagation phase-contrast imaging. The physical processes of absorption and phase shift when electromagnetic waves pass through a multi-material target phantom are quantitatively modeled. By separating the superposition effects of multiple materials on the target phantom, accurate extraction of single-material information is achieved, eliminating imaging artifacts caused by multi-material mixing at the source.
[0059] In one specific embodiment, the target phantom comprises two initial materials. The specific calculation formula for determining the projection thickness of the remaining initial materials using the initial projection data and the projection thickness corresponding to the first target material is as follows:
[0060] Where L2 represents the projected thickness of the remaining initial material; Indicates Fourier transform; represents the inverse Fourier transform; I represents the initial projection data; δ1 represents the linear attenuation coefficient of the first target material; δ1 represents the phase shift coefficient of the first target material; 4π 2 The constant term is represented by f, which represents the spatial frequency. δ1 represents the linear attenuation coefficient of the remaining initial material; δ2 represents the phase shift coefficient of the remaining initial material; L1 represents the projected thickness of the first target material; z represents the propagation distance. The first value is obtained by performing a Fourier transform on the projected thickness L1 corresponding to the first target material; The second value represents the linear attenuation coefficient of the first target material. Phase-related terms The sum is obtained.
[0061] In one embodiment, after reconstructing the target phantom based on the projected thickness of each initial material to obtain the target reconstructed image, the target reconstructed image is used as the new initial reconstructed image. The process of segmenting the image region belonging to the first target material from the initial reconstructed image is repeated at least once, serving as the segmented image of the first target material and subsequent steps. The first reconstructed target image may still contain a small number of residual artifacts. Through repeated execution and iterative optimization, the blurred signals in the boundary areas are further clarified, making the boundaries of different initial materials sharper and the internal grayscale more uniform. Ultimately, the residual artifacts are continuously compressed, and the purity of the target reconstructed image is significantly improved.
[0062] There is no limit to the number of times it can be repeated; the number can be set according to actual usage needs. For example, it can be repeated 1 time, 2 times, or 5 times.
[0063] In one specific implementation, the repeated execution can be stopped when the target reconstruction image corresponding to the obtained target phantom meets the image quality requirements.
[0064] Of course, in other specific implementations, the repeated execution can also be stopped when the number of repeated executions reaches a threshold; in this case, the size of the threshold is not limited.
[0065] Please see Figure 6 , Figure 6 yes Figure 5 The flowchart shown is a schematic diagram of one embodiment of step S51. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow the same pattern. Figure 6 The illustrated process sequence is limited. For example... Figure 6 As shown, this embodiment includes: Step S61: Select one initial material from at least two initial materials as the second target material.
[0066] In this embodiment, one initial material is selected from at least two initial materials as the second target material.
[0067] In one embodiment, the second target material may be any one of at least two initial materials included in the target phantom.
[0068] In other embodiments, the second target material may be the initial material among the at least two initial materials included in the target phantom that has higher imaging contrast, clearer structure, or is easier to accurately segment. For example, if the target phantom is a skull, and the skull includes bone and soft tissue as the two initial materials, the skull can be selected as the second target material.
[0069] In one embodiment, the first target material and the second target material are the same. When the first target material and the second target material are the same, there is no need to switch between two different materials for processing. Using the same material as the reference material for phase retrieval, image segmentation, and forward projection reduces algorithm complexity. Furthermore, it avoids deviations introduced by differences in the attenuation and phase characteristics of different materials, improving the accuracy of thickness extraction. This allows for accurate restoration of the true structure of each image based on the projected thickness corresponding to each initial material, eliminating artifacts at their source.
[0070] Step S62: Perform phase recovery on the initial projection data with respect to the second target material to obtain the target projection data.
[0071] In this embodiment, phase recovery with respect to the second target material is performed on the initial projection data to obtain target projection data. By using phase recovery under the single-material assumption, the phase shift information can be decoupled and extracted from the original mixed initial projection data to generate target projection data that is more consistent with image reconstruction.
[0072] Step S63: Reconstruct based on the target projection data to obtain the initial reconstructed image.
[0073] In this embodiment, an initial reconstructed image is obtained by reconstructing based on the target projection data. This initial reconstructed image, dominated by phase information and with fewer artifacts, provides a clear benchmark for subsequent initial material segmentation and multi-material iteration. Furthermore, without phase recovery of the second target material, the initial reconstructed image directly reconstructed from the initial projection data will be riddled with artifacts, making it impossible for subsequent segmentation to accurately distinguish regions of different initial materials, rendering the entire multi-material artifact elimination process impossible.
[0074] Please see Figure 7 , Figure 7 yes Figure 1 The diagram shows a flowchart of one embodiment of step S13. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow that approach. Figure 7 The illustrated process sequence is limited. For example... Figure 7 As shown, this embodiment includes: Step S71: Synthesize the projected thicknesses of each initial material to obtain a multi-material projected thickness image of the target model.
[0075] In this embodiment, the projected thicknesses corresponding to each initial material are synthesized to obtain a multi-material projected thickness image corresponding to the target phantom. It should be noted that the multi-material projected thickness image can also be called a virtual monoenergetic image. The projected thickness can be regarded as length, and the linear attenuation coefficient is similar to density. In the final synthesized virtual monoenergetic image, each pixel value represents the total attenuation effect felt by the electromagnetic wave along that path. This is a virtual, standardized CT image value.
[0076] In one embodiment, the projected thickness of each initial material is multiplied by its respective linear attenuation coefficient, and then the two are added together to obtain a multi-material projected thickness image corresponding to the target phantom.
[0077] In one specific embodiment, the target phantom comprises two initial materials, and the specific calculation formula for synthesis is shown below:
[0078] in, This represents the multi-material projection thickness image corresponding to the target phantom. This represents the projected thickness corresponding to an initial material. This represents the linear decay coefficient corresponding to an initial material. This represents the projected thickness of the other initial material; This represents the linear decay coefficient corresponding to the other initial material.
[0079] Step S72: Reconstruct the multi-material projection thickness image to obtain the target reconstructed image.
[0080] In this embodiment, the multi-material projection thickness image is reconstructed to obtain the target reconstructed image. The synthesis operation integrates the independent projection thicknesses of each initial material according to their spatial positions, generating a multi-material projection thickness image that includes the spatial distribution of all initial materials. This synthesized multi-material projection thickness image retains the independent signals of each initial material (without cross-interference) and completely reproduces the overall structure of the target phantom. In addition, synthesizing the projection thicknesses corresponding to each initial material before reconstruction completes the integration of initial material information at the two-dimensional projection level. During three-dimensional reconstruction, it is based on the same complete set of projection data, ensuring that the two initial materials are completely aligned in spatial position, the transition of the boundary areas between different initial materials is natural, and there are no stitching artifacts. The spatial consistency and integrity of the target reconstructed image are greatly improved.
[0081] In one embodiment, the FDK filtering back projection algorithm can be used to reconstruct the multi-material projection thickness image to obtain the target reconstructed image.
[0082] Please see Figure 8 , Figure 8This is a schematic diagram of an embodiment of the phase-contrast multi-material artifact elimination device provided in this application. The phase-contrast multi-material artifact elimination device 80 includes an acquisition module 81, an extraction module 82, and a reconstruction module 83; the acquisition module 81 is used to acquire initial projection data obtained by scanning a target phantom; wherein, the target phantom includes at least two initial materials, and the initial projection data is obtained by electromagnetic waves passing through the target phantom; the extraction module 82 is used to extract the projection thickness corresponding to each initial material from the initial projection data; the reconstruction module 83 is used to reconstruct the target phantom based on the projection thickness corresponding to each initial material to obtain a target reconstruction image corresponding to the target phantom.
[0083] The extraction module 82 is used to extract the projection thickness corresponding to each initial material from the initial projection data, including: extracting the projection thickness corresponding to the first target material from the initial projection data; wherein the first target material is an initial material; and using the initial projection data and the projection thickness corresponding to the first target material, determining the projection thickness corresponding to the remaining initial materials.
[0084] The target phantom includes two initial materials. The extraction module 82 is used to extract the projection thickness corresponding to the first target material from the initial projection data, including: reconstructing using the initial projection data to obtain an initial reconstructed image; segmenting the image region belonging to the first target material from the initial reconstructed image as a segmented image of the first target material; and projecting the segmented image of the first target material forward to obtain the projection thickness of the first target material. The extraction module 82 is used to determine the projection thickness corresponding to the remaining initial materials using the initial projection data and the projection thickness corresponding to the first target material, including: determining the projection thickness corresponding to the remaining initial materials using the initial projection data and the projection thickness corresponding to the first target material.
[0085] The extraction module 82 is used to determine the projection thickness of the remaining initial material by using the initial projection data and the projection thickness of the first target material, including: subtracting the linear attenuation, phase shift and propagation effects of the first target material from the initial projection data to obtain the projection thickness of the remaining initial material.
[0086] The steps for determining the linear attenuation, phase shift, and propagation effects of the first target material include: performing a Fourier transform on the projected thickness of the first target material to obtain a first value; and summing the linear attenuation coefficient and phase correlation term of the first target material to obtain a second value; wherein the phase correlation term of the first target material is obtained by multiplying the propagation distance, the phase shift coefficient of the first target material, the spatial frequency, and a preset constant term; and the product of the first value and the second value is used as the linear attenuation, phase shift, and propagation effects of the first target material.
[0087] The extraction module 82 is used to reconstruct the image using the initial projection data, including: selecting one initial material from at least two initial materials as the second target material; performing phase recovery on the initial projection data with respect to the second target material to obtain target projection data; and reconstructing the image based on the target projection data to obtain the initial reconstructed image.
[0088] The first target material is the same as the second target material.
[0089] The phase-contrast multi-material artifact elimination device 80 also includes an iterative optimization module 84. The iterative optimization module 84 is used to reconstruct the target reconstruction image corresponding to the target phantom based on the projection thickness of each initial material, and then use the target reconstruction image as a new initial reconstruction image. The module 84 also repeats the process at least once to segment the image region belonging to the first target material from the initial reconstruction image as the segmentation image of the first target material and its subsequent steps.
[0090] The reconstruction module 83 is used to reconstruct the target phantom based on the projection thickness of each initial material to obtain the target reconstruction image, including: synthesizing the projection thickness of each initial material to obtain a multi-material projection thickness image of the target phantom; and reconstructing the multi-material projection thickness image to obtain the target reconstruction image.
[0091] The extraction module 82 is used to perform data cleaning on the initial projection data before extracting the projection thickness corresponding to each initial material from the initial projection data; wherein the data cleaning includes at least one of the following: bad pixel elimination, data correction, and data format conversion.
[0092] Please see Figure 9 , Figure 9 This is a schematic diagram of an embodiment of the electronic device provided in this application. The electronic device 90 includes a memory 91 and a processor 92 coupled to each other. The processor 92 is used to execute program instructions stored in the memory 91 to implement the steps of any of the above-described embodiments of the phase-contrast multi-material artifact elimination method. In a specific implementation scenario, the electronic device 90 may include, but is not limited to, a microcomputer or a server. In addition, the electronic device 90 may also include mobile devices such as laptops and tablets, which are not limited here.
[0093] Specifically, processor 92 controls itself and memory 91 to implement the steps of any of the phase-contrast multi-material artifact elimination method embodiments described above. Processor 92 can also be referred to as a CPU (Central Processing Unit). Processor 92 may be an integrated circuit chip with signal processing capabilities. Processor 92 can also be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor. Furthermore, processor 92 can be implemented using integrated circuit chips.
[0094] Please see Figure 10 , Figure 10 This is a schematic diagram of an embodiment of the computer-readable storage medium provided in this application. The computer-readable storage medium 100 of this application embodiment stores program instructions 101. When executed, these program instructions 101 implement the methods provided in any embodiment of the contrast-contrast multi-material artifact elimination method and any non-conflicting combination thereof. The program instructions 101 can be formed into a program file and stored in the aforementioned computer-readable storage medium 100 in the form of a software product, so that a computer device (which may be a personal computer, server, or network device, etc.) can execute all or part of the steps of the methods of various embodiments of this application. The aforementioned computer-readable storage medium 100 includes various media capable of storing program code, such as a USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or terminal devices such as computers, servers, mobile phones, and tablets.
[0095] If the technical solution of this application involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.
[0096] The above are merely embodiments of this application and do not limit the scope of this patent application. Any equivalent structural or procedural changes made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of this application.
Claims
1. A phase-contrast multi-material artifact removal method, characterized by, The method includes: Acquire initial projection data obtained by scanning a target phantom; wherein the target phantom comprises at least two initial materials, and the initial projection data is obtained by electromagnetic waves passing through the target phantom; Extract the projected thickness of each initial material from the initial projection data; Based on the projected thickness of each of the initial materials, a target reconstruction image corresponding to the target phantom is obtained.
2. The method of claim 1, wherein, The step of extracting the projected thickness corresponding to each initial material from the initial projection data includes: The projection thickness corresponding to the first target material is extracted from the initial projection data; wherein the first target material is the initial material. Using the initial projection data and the projection thickness corresponding to the first target material, the projection thickness corresponding to the remaining initial material is determined.
3. The method of claim 2, wherein, The target phantom comprises two initial materials; the step of extracting the projection thickness corresponding to the first target material from the initial projection data includes: The initial projection data is used to reconstruct the image, resulting in an initial reconstructed image. From the initial reconstructed image, the image region belonging to the first target material is segmented as the segmented image of the first target material; The segmented image of the first target material is projected forward to obtain the projected thickness of the first target material; The step of determining the remaining projection thickness of the initial material using the initial projection data and the projection thickness corresponding to the first target material includes: Using the initial projection data and the projection thickness corresponding to the first target material, the projection thickness corresponding to the remaining initial material is determined.
4. The method of claim 3, wherein, The step of determining the remaining projection thickness of the initial material using the initial projection data and the projection thickness corresponding to the first target material includes: From the initial projection data, the linear attenuation, phase shift, and propagation effects of the first target material are subtracted to obtain the remaining projection thickness of the initial material. And / or, the step of reconstructing using the initial projection data to obtain an initial reconstructed image includes: From the at least two initial materials, one of the initial materials is selected as the second target material; Phase recovery with respect to the second target material is performed on the initial projection data to obtain target projection data; The initial reconstructed image is obtained by reconstructing based on the target projection data.
5. The method of claim 4, wherein, The steps for determining the linear attenuation, phase shift, and propagation effects of the first target material include: Perform a Fourier transform on the projected thickness corresponding to the first target material to obtain a first value; Furthermore, the second value is obtained by summing the linear attenuation coefficient and the phase correlation term of the first target material; wherein the phase correlation term of the first target material is obtained by multiplying the propagation distance, the phase shift coefficient of the first target material, the spatial frequency, and a preset constant term. The product of the first value and the second value is taken as the linear attenuation, phase shift, and propagation effects of the first target material.
6. The method according to claim 4, characterized in that, The first target material is the same as the second target material.
7. The method of claim 3, wherein, After reconstructing the target phantom image based on the projected thickness corresponding to each of the initial materials, the method further includes: The target reconstructed image is used as a new initial reconstructed image, and the process of segmenting the image region belonging to the first target material from the initial reconstructed image is repeated at least once, as the segmented image of the first target material and subsequent steps.
8. The method of claim 1, wherein, The process of reconstructing the target phantom based on the projected thickness of each of the initial materials to obtain the target reconstructed image includes: The projected thicknesses of each of the initial materials are synthesized to obtain a multi-material projected thickness image of the target phantom. The target reconstructed image is obtained by reconstructing the multi-material projection thickness image.
9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory being used to store program instructions, and the processor being used to execute the program instructions to implement the phase-contrast multi-material artifact elimination method as described in any one of claims 1-8.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program instructions that can be executed to implement the phase-contrast multi-material artifact elimination method as described in any one of claims 1-8.