Limited angle projection phase contrast CT image reconstruction method and device based on deep learning

Abstract

The invention discloses a limited angle projection phase contrast CT image reconstruction method and device based on deep learning. The method comprises the following steps: acquiring limited angle projection data of X-ray phase contrast CT scanning through an equal angle increment strategy; performing flat field correction and dark field correction preprocessing on the limited angle projection data to obtain limited angle projection data before phase recovery; preprocessing the finite angle projection data through a phase recovery strategy to obtain finite angle projection data after phase recovery; cT reconstruction is carried out on the limited angle projection data before and after phase recovery through a filtering back projection strategy, and CT images of limited angle projection before and after phase recovery are obtained; a phase contrast CT intelligent image reconstruction network model is constructed and trained; and performing reconstruction processing on the CT image of the limited angle projection before and after phase recovery through the trained phase contrast CT intelligent image reconstruction network model to obtain a reconstructed phase contrast CT image. According to the method, phase contrast CT image reconstruction can be rapidly and efficiently carried out, finite angle reconstruction artifacts can be effectively inhibited, and the image quality is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radiological medical image processing, and in particular to a limited-angle projection phase-contrast CT image reconstruction method and device based on deep learning. Background Art

[0002] Compared with traditional X-ray absorption CT imaging technology, X-ray phase contrast CT imaging technology can produce higher contrast and spatial resolution for weakly absorbing samples, and has great application potential in biomedical imaging and clinical imaging. At present, long CT scanning time and high radiation dose are important issues that hinder the application of X-ray phase contrast CT imaging technology, and phase contrast CT reconstruction based on limited angle projection is one of the feasible strategies to solve the above problems.

[0003] Traditional X-ray phase contrast CT image reconstruction techniques can be mainly divided into three categories: (1) analytical methods based on the central slice theorem, such as the filtered back projection algorithm (FBP); (2) iterative reconstruction techniques based on precise mathematical modeling, including algebraic iterative reconstruction techniques (such as ART) and compressed sensing algorithms (such as total variation minimization algorithms TV); (3) reconstruction methods based on deep learning (such as FBPCovnet). Analytical methods rely on complete projection data to reconstruct high-quality phase contrast CT images, and face the problem of high radiation dose; iterative reconstruction techniques have high computational costs, resulting in long reconstruction times, and usually rely on available prior knowledge (such as compressed sensing algorithms) to optimize the reconstruction of CT images, which easily affects its generalization application capabilities. In addition, iterative reconstruction techniques have limited control over the quality of reconstructed images; with the development of deep learning technology, the use of deep neural networks to model and optimize X-ray phase contrast CT reconstruction problems from big data can more efficiently complete phase contrast CT reconstruction tasks, especially for its limited angle CT reconstruction needs. However, purely data-driven deep learning methods, such as DDNet and FBPCovnet, lack certain interpretability in the problem of limited-angle phase-contrast CT reconstruction and have limited reconstruction effects because they do not combine the physical imaging characteristics of phase-contrast CT imaging well.

[0004] Therefore, in view of the shortcomings of the existing technology, it is necessary to provide a deep neural network model that combines the physical imaging characteristics of X-ray phase contrast CT to solve the shortcomings of the existing technology in limited angle reconstruction of X-ray phase contrast CT images. Summary of the invention

[0005] To this end, the present invention provides a limited angle projection phase contrast CT image reconstruction method and device based on deep learning. The method designs the structure of a deep neural network by utilizing the physical imaging characteristics of X-ray phase contrast CT imaging, which can not only quickly and efficiently reconstruct phase contrast CT images, but also effectively suppress limited angle reconstruction artifacts and improve image quality.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a limited angle projection phase contrast CT image reconstruction method based on deep learning, comprising:

[0007] The limited angle projection data of X-ray phase contrast CT scanning is acquired by using the equal angle increment strategy;

[0008] The obtained limited angle projection data is preprocessed with flat field correction and dark field correction to obtain the limited angle projection data before phase recovery; the limited angle projection data before phase recovery is preprocessed with a phase recovery strategy to obtain the limited angle projection data after phase recovery;

[0009] The limited angle projection data before phase recovery and the limited angle projection data after phase recovery are respectively subjected to CT reconstruction by a filtering back-projection strategy to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection;

[0010] Constructing a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; training the edge enhancement network model and the multi-scale reconstruction network model respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model;

[0011] The CT image before phase recovery and the CT image after phase recovery of limited angle projection are input into the trained phase contrast CT intelligent image reconstruction network model; and the reconstructed phase contrast CT image is obtained through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

[0012] As a preferred solution of the limited angle projection phase contrast CT image reconstruction method based on deep learning, in the process of acquiring limited angle projection data of X-ray phase contrast CT scanning by the equal angle increment strategy, the calculation formula of the equal angle increment strategy is:

[0013]

[0014] Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

[0015] As a preferred solution of the limited angle projection phase contrast CT image reconstruction method based on deep learning, in the process of performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery, the correction expressions of the flat field correction and the dark field correction are:

[0016]

[0017] In the formula, I tomo Phase contrast CT projection image before correction; I flat is the average image of the flat-field image, I dark is the average image of the dark field images;

[0018] In the process of preprocessing the limited angle projection data before phase recovery by using the phase recovery strategy to obtain the limited angle projection data after phase recovery, the phase recovery strategy expression is:

[0019]

[0020] Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

[0021] As a preferred solution of the limited angle projection phase contrast CT image reconstruction method based on deep learning, in the process of respectively training the edge enhancement network model and the multi-scale reconstruction network model through the setting data, the edge enhancement network model is trained through the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model is trained through the CT image after phase recovery and the CT image after phase recovery of the full projection.

[0022] As a preferred solution of the limited angle projection phase contrast CT image reconstruction method based on deep learning, the loss function expression of the phase contrast CT intelligent image reconstruction network model is:

[0023]

[0024] In the formula, They are pixel consistency loss, perceptual loss and edge loss respectively. α1 and α2 are used to control the weights of perceptual loss and edge loss respectively.

[0025] The present invention also provides a limited angle projection phase contrast CT image reconstruction device based on deep learning, which adopts the limited angle projection phase contrast CT image reconstruction method based on deep learning described above, including:

[0026] A limited angle projection data acquisition module, used for acquiring limited angle projection data of X-ray phase contrast CT scanning by using an equal angle increment strategy;

[0027] A projection data preprocessing module is used to perform flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery; and preprocess the limited angle projection data before phase recovery through a phase recovery strategy to obtain the limited angle projection data after phase recovery;

[0028] A CT image reconstruction module, used to perform CT reconstruction on the limited angle projection data before phase recovery and the limited angle projection data after phase recovery by using a filtering back projection strategy, respectively, to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection;

[0029] A phase contrast CT intelligent image reconstruction network model construction and training module is used to construct a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; the edge enhancement network model and the multi-scale reconstruction network model are trained respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model;

[0030] The reconstructed phase contrast CT image acquisition module is used to input the CT image before phase recovery and the CT image after phase recovery of limited angle projection into the trained phase contrast CT intelligent image reconstruction network model; and obtain the reconstructed phase contrast CT image through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

[0031] As a preferred solution of the limited angle projection phase contrast CT image reconstruction device based on deep learning, in the limited angle projection data acquisition module, in the process of acquiring the limited angle projection data of the X-ray phase contrast CT scan by the equal angle increment strategy, the calculation formula of the equal angle increment strategy is:

[0032]

[0033] Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

[0034] As a preferred solution of the limited angle projection phase contrast CT image reconstruction device based on deep learning, in the projection data preprocessing module, in the process of performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery, the correction expressions of the flat field correction and the dark field correction are:

[0035]

[0036] In the formula, I tomo Phase contrast CT projection image before correction; I flat is the average image of the flat-field image, I dark is the average image of the dark field images;

[0037] In the process of preprocessing the limited angle projection data before phase recovery by using the phase recovery strategy to obtain the limited angle projection data after phase recovery, the phase recovery strategy expression is:

[0038]

[0039] Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

[0040] As a preferred solution of the limited angle projection phase contrast CT image reconstruction device based on deep learning, in the phase contrast CT intelligent image reconstruction network model construction and training module, in the process of respectively training the edge enhancement network model and the multi-scale reconstruction network model through the setting data, the edge enhancement network model is trained through the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model is trained through the CT image after phase recovery and the CT image after phase recovery of the full projection.

[0041] As a preferred solution of the limited angle projection phase contrast CT image reconstruction device based on deep learning, in the phase contrast CT intelligent image reconstruction network model construction and training module, the loss function expression of the phase contrast CT intelligent image reconstruction network model is:

[0042]

[0043] In the formula, They are pixel consistency loss, perceptual loss and edge loss respectively. α1 and α2 are used to control the weights of perceptual loss and edge loss respectively.

[0044] The present invention has the following advantages: obtaining limited angle projection data of X-ray phase contrast CT scanning by means of an equal angle increment strategy; performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery; performing preprocessing on the limited angle projection data before phase recovery by means of a phase recovery strategy to obtain the limited angle projection data after phase recovery; performing CT reconstruction on the limited angle projection data before phase recovery and the limited angle projection data after phase recovery by means of a filtered back projection strategy respectively to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection; constructing a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model comprises an edge enhancement network model and a multi-scale reconstruction network model; respectively training the edge enhancement network model and the multi-scale reconstruction network model by means of setting data to obtain the trained phase contrast CT intelligent image reconstruction network model; inputting the CT image before phase recovery and the CT image after phase recovery of limited angle projection into the trained phase contrast CT intelligent image reconstruction network model; and obtaining a reconstructed phase contrast CT image by means of reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model. Compared with the traditional method, the present invention has high reconstruction quality. The clarity and contrast of the image are significantly improved, and the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) evaluation of the reconstructed image quality are higher, and the reconstruction error (MAE) evaluation is lower. Compared with the traditional method, the present invention has high reconstruction performance. In the limited angle scanning range of phase contrast CT such as [0, 150°], [0, 120°] and [0, 90°], better CT reconstruction results can be obtained, which can not only effectively suppress the limited angle CT reconstruction artifacts, but also better protect the edge features and detail information. At the same time, it also has the characteristics of high practical application value. The X-ray phase contrast CT intelligent reconstruction model can quickly obtain high-quality CT images, with high reconstruction efficiency and strong generalization application capabilities. The present invention designs a deep neural network model MSFR-EEG (with certain interpretability) based on the physical imaging characteristics of X-ray phase contrast CT, and provides a limited angle reconstruction system for X-ray phase contrast CT images. The system can quickly and efficiently reconstruct phase contrast CT images, effectively suppress limited angle reconstruction artifacts and improve image quality, and has high practical application value in the field of phase contrast CT imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the drawings required for the implementation methods or the prior art descriptions are briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0046] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0047] Figure 1 A schematic flow chart of a limited angle projection phase contrast CT image reconstruction method based on deep learning provided in Example 1 of the present invention;

[0048] Figure 2 A schematic diagram of a specific implementation process of the limited angle projection phase contrast CT image reconstruction method based on deep learning provided in Example 1 of the present invention;

[0049] Figure 3 A schematic diagram of the architecture of a multi-scale reconstruction network model module in a limited angle projection phase contrast CT image reconstruction method based on deep learning provided in Example 1 of the present invention;

[0050] Figure 4 A schematic diagram of some results of the phase contrast CT image intelligent reconstruction method and the prior art in the limited angle projection phase contrast CT image reconstruction method based on deep learning provided in Example 1 of the present invention;

[0051] Figure 5 A schematic diagram of the architecture of a limited-angle projection phase-contrast CT image reconstruction device based on deep learning provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0052] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Example 1

[0054] See also Figure 1 and Figure 2 Embodiment 1 of the present invention provides a limited angle projection phase contrast CT image reconstruction method based on deep learning, comprising:

[0055] S1, obtaining limited angle projection data of X-ray phase contrast CT scanning by equal angle increment strategy;

[0056] S2, performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery; performing preprocessing on the limited angle projection data before phase recovery by using a phase recovery strategy to obtain the limited angle projection data after phase recovery;

[0057] S3, reconstructing the limited angle projection data before phase recovery and the limited angle projection data after phase recovery by using a filtered back projection strategy to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection;

[0058] S4, constructing a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; training the edge enhancement network model and the multi-scale reconstruction network model respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model;

[0059] S5. Input the CT image before phase recovery and the CT image after phase recovery of limited angle projection into the trained phase contrast CT intelligent image reconstruction network model; obtain a reconstructed phase contrast CT image through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

[0060] In this embodiment, in step S1, the limited angle projection data I of the X-ray phase contrast CT scan is obtained by using the equal angle increment strategy. tomo ; Among them, the limited angle is the projection angle θ<180°.

[0061] The calculation formula of the equal angle increment strategy is:

[0062]

[0063] Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

[0064] In this embodiment, in step S2, the obtained limited angle projection data is preprocessed with flat field correction and dark field correction to obtain the limited angle projection data before phase recovery; the limited angle projection data before phase recovery is preprocessed with a phase recovery strategy to obtain the limited angle projection data after phase recovery;

[0065] Specifically, based on the flat-field image I acquired during X-ray phase contrast CT scanning flat and dark field image I dark , for the obtained projection data I tomo Flat field correction and dark field correction are performed to obtain I, and then phase recovery is performed on I through a phase recovery strategy to obtain φ;

[0066] Among them, the correction expressions of the flat field correction and the dark field correction are:

[0067]

[0068] In the formula, I tomo I is the phase contrast CT projection image before correction; flat is the average image of the flat-field image, I dark is the average image of the dark field images;

[0069] The phase recovery strategy expression is:

[0070]

[0071] Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

[0072] In this embodiment, in step S3, the limited angle projection data before phase recovery and the limited angle projection data after phase recovery are respectively subjected to CT reconstruction by a filtered back projection strategy to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection;

[0073] Specifically, the projection data before and after phase recovery (ie, I and φ) are respectively reconstructed by filtered back projection (FBP) to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection.

[0074] In this embodiment, in step S4, a phase contrast CT intelligent image reconstruction network model is constructed; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; the edge enhancement network model and the multi-scale reconstruction network model are trained respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model;

[0075] Specifically, the phase-contrast CT intelligent image reconstruction network model (multiscale reconstruction frameworkbased on edge-enhanced guidance, MSFR-EEG) uses the two physical imaging characteristics of X-ray phase-contrast CT imaging, "multiscale" and "edge enhancement", to design the network model, namely the multiscale reconstruction network model (multiscale reconstruction subnetwork, MSRSN) and the edge-enhanced subnetwork model (edge-enhanced subnetwork, EESN).

[0076] Among them, Figure 3 As shown in the figure, the multi-scale reconstruction network model MSRSN includes an edge attention module (EAT) and an edge aggregation module (EAG). EAT can establish attention weights of different sizes for edge information of different scales on phase contrast CT images, which is used to optimize the edges of reconstructed images and improve the quality and accuracy of reconstructed images. EAG can effectively utilize structural information of different scales encoded in phase contrast CT images. It is used to enhance the model's ability to encode edges and structural features of different scales.

[0077] In the process of training the edge enhancement network model EESN and the multi-scale reconstruction network model MSRSN, the edge enhancement network model EESN is trained by the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model MSRSN is trained by the CT image after phase recovery and the CT image after phase recovery of the full projection.

[0078] In this embodiment, in step S5, the CT image before phase recovery and the CT image after phase recovery of limited angle projection are input into the trained phase contrast CT intelligent image reconstruction network model; and the reconstructed phase contrast CT image is obtained through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

[0079] Specifically, EESN is first used to extract the edge prior of the phase contrast CT image, and then the MSRSN is guided to optimize the reconstruction through the two modules of EAT and EAG, and finally a high-quality phase contrast CT image is output. Figure 4 The quantitative evaluation scores are shown in Table 1;

[0080]

[0081] Table 1 Quantitative evaluation table

[0082] In summary, the present invention obtains limited-angle projection data of X-ray phase contrast CT scanning through an equal-angle increment strategy; performs flat-field correction and dark-field correction preprocessing on the obtained limited-angle projection data to obtain the limited-angle projection data before phase recovery; preprocesses the limited-angle projection data before phase recovery through a phase recovery strategy to obtain the limited-angle projection data after phase recovery; performs CT reconstruction on the limited-angle projection data before phase recovery and the limited-angle projection data after phase recovery through a filtered back-projection strategy, respectively, to obtain a CT image before phase recovery and a CT image after phase recovery of limited-angle projection; constructs a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; trains the edge enhancement network model and the multi-scale reconstruction network model respectively through setting data to obtain the trained phase contrast CT intelligent image reconstruction network model; inputs the CT image before phase recovery and the CT image after phase recovery of limited-angle projection into the trained phase contrast CT intelligent image reconstruction network model; and obtains a reconstructed phase contrast CT image through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model. Compared with the traditional method, the present invention has high reconstruction quality. The clarity and contrast of the image are significantly improved, and the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) evaluation of the reconstructed image quality are higher, and the reconstruction error (MAE) evaluation is lower. Compared with the traditional method, the present invention has high reconstruction performance. In the limited angle scanning range of phase contrast CT such as [0, 150°], [0, 120°] and [0, 90°], better CT reconstruction results can be obtained, which can not only effectively suppress the limited angle CT reconstruction artifacts, but also better protect the edge features and detail information. At the same time, it also has the characteristics of high practical application value. The X-ray phase contrast CT intelligent reconstruction model can quickly obtain high-quality CT images, with high reconstruction efficiency and strong generalization application capabilities. The present invention designs a deep neural network model MSFR-EEG (with certain interpretability) based on the physical imaging characteristics of X-ray phase contrast CT, and provides a limited angle reconstruction system for X-ray phase contrast CT images. The system can quickly and efficiently reconstruct phase contrast CT images, effectively suppress limited angle reconstruction artifacts and improve image quality, and has high practical application value in the field of phase contrast CT imaging.

[0083] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0084] It should be noted that the above describes some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] Example 2

[0086] See also Figure 5 Embodiment 2 of the present invention provides a limited angle projection phase contrast CT image reconstruction device based on deep learning, comprising:

[0087] The limited angle projection data acquisition module 001 is used to acquire limited angle projection data of X-ray phase contrast CT scanning by using an equal angle increment strategy;

[0088] The projection data preprocessing module 002 is used to perform flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery; and preprocess the limited angle projection data before phase recovery by using a phase recovery strategy to obtain the limited angle projection data after phase recovery;

[0089] The CT image reconstruction module 003 is used to perform CT reconstruction on the limited angle projection data before phase recovery and the limited angle projection data after phase recovery through a filtering back projection strategy to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection;

[0090] Phase contrast CT intelligent image reconstruction network model construction and training module 004, used to construct a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; the edge enhancement network model and the multi-scale reconstruction network model are trained respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model;

[0091] The reconstructed phase contrast CT image acquisition module 005 is used to input the CT image before phase recovery and the CT image after phase recovery of limited angle projection into the trained phase contrast CT intelligent image reconstruction network model; and obtain the reconstructed phase contrast CT image through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

[0092] In this embodiment, in the limited angle projection data acquisition module 001, in the process of acquiring limited angle projection data of X-ray phase contrast CT scanning by using the equal angle increment strategy, the calculation formula of the equal angle increment strategy is:

[0093]

[0094] Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

[0095] In this embodiment, in the projection data preprocessing module 002, in the process of performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery, the correction expressions of the flat field correction and the dark field correction are:

[0096]

[0097] In the formula, I tomo I is the phase contrast CT projection image before correction; flat is the average image of the flat-field image, I dark is the average image of the dark field images;

[0098] In the process of preprocessing the limited angle projection data before phase recovery by using the phase recovery strategy to obtain the limited angle projection data after phase recovery, the phase recovery strategy expression is:

[0099]

[0100] Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

[0101] In this embodiment, in the phase contrast CT intelligent image reconstruction network model construction and training module 004, in the process of respectively training the edge enhancement network model and the multi-scale reconstruction network model through the setting data, the edge enhancement network model is trained through the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model is trained through the CT image after phase recovery and the CT image after phase recovery of the full projection.

[0102] In this embodiment, in the phase contrast CT intelligent image reconstruction network model construction and training module 004, the loss function expression of the phase contrast CT intelligent image reconstruction network model is:

[0103]

[0104] In the formula, They are pixel consistency loss, perceptual loss and edge loss respectively. α1 and α2 are used to control the weights of perceptual loss and edge loss respectively.

[0105] It should be noted that the information interaction, execution process, etc. between the modules / units of the above-mentioned system are based on the same concept as the method embodiment in Example 1 of the present application, and the technical effects they bring are the same as those of the method embodiment of the present application. For specific contents, please refer to the description in the method embodiment shown in the previous part of the present application, and will not be repeated here.

[0106] Example 3

[0107] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code of a limited-angle projection phase-contrast CT image reconstruction method based on deep learning is stored. The program code includes instructions for executing the limited-angle projection phase-contrast CT image reconstruction method based on deep learning of Embodiment 1 or any possible implementation thereof.

[0108] The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0109] Example 4

[0110] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0111] The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the limited angle projection phase contrast CT image reconstruction method based on deep learning of Example 1 or any possible implementation thereof.

[0112] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor implemented by reading software codes stored in a memory. The memory can be integrated in the processor or can be located outside the processor and exist independently.

[0113] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center.

[0114] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0115] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A limited angle projection phase contrast CT image reconstruction method based on deep learning, characterized in that: include: The limited angle projection data of X-ray phase contrast CT scanning is acquired by using the equal angle increment strategy; The obtained limited angle projection data is preprocessed with flat field correction and dark field correction to obtain the limited angle projection data before phase recovery; the limited angle projection data before phase recovery is preprocessed with a phase recovery strategy to obtain the limited angle projection data after phase recovery; The limited angle projection data before phase recovery and the limited angle projection data after phase recovery are respectively subjected to CT reconstruction by a filtering back-projection strategy to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection; Constructing a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; training the edge enhancement network model and the multi-scale reconstruction network model respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model; The CT image before phase recovery and the CT image after phase recovery of limited angle projection are input into the trained phase contrast CT intelligent image reconstruction network model; and the reconstructed phase contrast CT image is obtained through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

2. The method for limited angle projection phase contrast CT image reconstruction based on deep learning according to claim 1, characterized in that: In the process of acquiring limited angle projection data of X-ray phase contrast CT scanning by using the equal angle increment strategy, the calculation formula of the equal angle increment strategy is: Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

3. The method for limited angle projection phase contrast CT image reconstruction based on deep learning according to claim 2, characterized in that: In the process of performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery, the correction expressions of the flat field correction and the dark field correction are: In the formula, I tomo Phase contrast CT projection image before correction; I flat is the average image of the flat-field image, I dark is the average image of the dark field images; In the process of preprocessing the limited angle projection data before phase recovery by using the phase recovery strategy to obtain the limited angle projection data after phase recovery, the phase recovery strategy expression is: Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

4. The method for limited angle projection phase contrast CT image reconstruction based on deep learning according to claim 3, characterized in that: In the process of respectively training the edge enhancement network model and the multi-scale reconstruction network model through the setting data, the edge enhancement network model is trained through the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model is trained through the CT image after phase recovery and the CT image after phase recovery of the full projection.

5. The method for limited angle projection phase contrast CT image reconstruction based on deep learning according to claim 4, characterized in that: The loss function expression of the phase contrast CT intelligent image reconstruction network model is: In the formula, They are pixel consistency loss, perceptual loss and edge loss respectively; α1 and α2 are used to control the weights of perceptual loss and edge loss respectively.

6. A limited angle projection phase contrast CT image reconstruction device based on deep learning, adopting the limited angle projection phase contrast CT image reconstruction method based on deep learning according to any one of claims 1 to 5, characterized in that: include: A limited angle projection data acquisition module, used for acquiring limited angle projection data of X-ray phase contrast CT scanning by using an equal angle increment strategy; A projection data preprocessing module is used to perform flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery; and preprocess the limited angle projection data before phase recovery through a phase recovery strategy to obtain the limited angle projection data after phase recovery; A CT image reconstruction module, used to perform CT reconstruction on the limited angle projection data before phase recovery and the limited angle projection data after phase recovery by using a filtering back projection strategy, respectively, to obtain a CT image before phase recovery and a CT image after phase recovery of limited angle projection; A phase contrast CT intelligent image reconstruction network model construction and training module is used to construct a phase contrast CT intelligent image reconstruction network model; the phase contrast CT intelligent image reconstruction network model includes an edge enhancement network model and a multi-scale reconstruction network model; the edge enhancement network model and the multi-scale reconstruction network model are trained respectively by setting data to obtain the trained phase contrast CT intelligent image reconstruction network model; The reconstructed phase contrast CT image acquisition module is used to input the CT image before phase recovery and the CT image after phase recovery of limited angle projection into the trained phase contrast CT intelligent image reconstruction network model; and obtain the reconstructed phase contrast CT image through reconstruction processing by the trained phase contrast CT intelligent image reconstruction network model.

7. The apparatus for reconstructing limited angle projection phase contrast CT images based on deep learning according to claim 6, characterized in that: In the limited angle projection data acquisition module, in the process of acquiring limited angle projection data of X-ray phase contrast CT scanning by using the equal angle increment strategy, the calculation formula of the equal angle increment strategy is: Where Δd is the angle increment; N is the number of detector units corresponding to the maximum thickness of the sample at a single projection angle.

8. The apparatus for reconstructing limited angle projection phase contrast CT images based on deep learning according to claim 7, characterized in that: In the projection data preprocessing module, in the process of performing flat field correction and dark field correction preprocessing on the obtained limited angle projection data to obtain the limited angle projection data before phase recovery, the correction expressions of the flat field correction and the dark field correction are: In the formula, I tomo Phase contrast CT projection image before correction; I flat is the average image of the flat-field image, I dark is the average image of the dark field images; In the process of preprocessing the limited angle projection data before phase recovery by using the phase recovery strategy to obtain the limited angle projection data after phase recovery, the phase recovery strategy expression is: Where (x, y) is the coordinate of the detector plane; (ξ, v) is the coordinate corresponding to (x, y) in the frequency domain; is the light intensity recorded by the detector when the distance from the sample to the detector is D and the projection angle is θ; for Image after phase recovery; γ is a constant coefficient; λ is the wavelength of X-rays; F and F -1 is the Fourier transform and its inverse transform.

9. The apparatus for reconstructing limited angle projection phase contrast CT images based on deep learning according to claim 8, characterized in that: In the phase contrast CT intelligent image reconstruction network model construction and training module, in the process of respectively training the edge enhancement network model and the multi-scale reconstruction network model through the setting data, the edge enhancement network model is trained through the CT image before phase recovery and the real edge image data; the multi-scale reconstruction network model is trained through the CT image after phase recovery and the CT image after phase recovery of the full projection.

10. The apparatus for reconstructing limited angle projection phase contrast CT images based on deep learning according to claim 9, characterized in that: In the phase contrast CT intelligent image reconstruction network model construction and training module, the loss function expression of the phase contrast CT intelligent image reconstruction network model is: In the formula, They are pixel consistency loss, perceptual loss and edge loss respectively; α1 and α2 are used to control the weights of perceptual loss and edge loss respectively.

Images