CN113367717A - Cone beam X-ray fluorescence imaging method, system, terminal and storage medium - Google Patents
Cone beam X-ray fluorescence imaging method, system, terminal and storage medium Download PDFInfo
- Publication number
- CN113367717A CN113367717A CN202110576925.0A CN202110576925A CN113367717A CN 113367717 A CN113367717 A CN 113367717A CN 202110576925 A CN202110576925 A CN 202110576925A CN 113367717 A CN113367717 A CN 113367717A
- Authority
- CN
- China
- Prior art keywords
- image
- imaging
- fluorescence
- contrast image
- contrast
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/02—Devices for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
- A61B6/03—Computerised tomographs
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/40—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for generating radiation specially adapted for radiation diagnosis
- A61B6/4035—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for generating radiation specially adapted for radiation diagnosis the source being combined with a filter or grating
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/40—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for generating radiation specially adapted for radiation diagnosis
- A61B6/4064—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for generating radiation specially adapted for radiation diagnosis specially adapted for producing a particular type of beam
- A61B6/4085—Cone-beams
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/48—Diagnostic techniques
- A61B6/484—Diagnostic techniques involving phase contrast X-ray imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/48—Diagnostic techniques
- A61B6/485—Diagnostic techniques involving fluorescence X-ray imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5205—Devices using data or image processing specially adapted for radiation diagnosis involving processing of raw data to produce diagnostic data
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5211—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5211—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data
- A61B6/5229—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data combining image data of a patient, e.g. combining a functional image with an anatomical image
- A61B6/5235—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data combining image data of a patient, e.g. combining a functional image with an anatomical image combining images from the same or different ionising radiation imaging techniques, e.g. PET and CT
Abstract
The application relates to a cone beam X-ray fluorescence imaging method, a system, a terminal and a storage medium. The method comprises the following steps: acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device, and synchronously acquiring a surface fluorescence image of the imaging target; carrying out information separation and extraction on the phase contrast projection image to obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of an imaging target, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image; inputting the three-dimensional reconstruction images of the absorption contrast image, the refraction contrast image and the scattering contrast image and the surface fluorescence image into a trained XLCT three-dimensional reconstruction depth convolution neural network model, and outputting a cone beam XLCT fluorescence image of an imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model; the method and the device can effectively reduce the ill-conditioned degree of the solution of the cone beam XLCT three-dimensional reconstruction equation and realize the fast and high-resolution cone beam X-ray fluorescence imaging.
Description
Technical Field
The application belongs to the technical field of optical molecular imaging, and particularly relates to a cone beam X-ray fluorescence imaging method, a cone beam X-ray fluorescence imaging system, a cone beam X-ray fluorescence imaging terminal and a storage medium.
Background
The optical molecular imaging technology is an imaging technology which utilizes a specific molecular probe to emit light, uses an optical imaging system to detect an optical signal emitted by a molecular marker, and recovers the spatial distribution of the molecular probe through algorithm reconstruction so as to obtain a functional image of a cell or a tissue. The optical molecular imaging technology is an important technical means for qualitative or quantitative research on the cellular molecular level of physiological processes and pathological change processes of biological tissues in a living body state. In the last two decades, the bio-optical molecular imaging technology is rapidly developed and applied in the fields of tumor research, bio-pharmaceuticals and the like, and according to the difference of imaging mechanisms, the currently representative optical molecular imaging technology mainly comprises: autofluorescence tomography (BLT) and Fluorescence Molecular Tomography (FMT). Where BLT is bioluminescent imaging without the need for external light source excitation, while FMT requires an external light source to excite fluorescent molecules in the organism to emit light. FMT has the advantages that the available fluorescent probes are much more abundant than BLT, have a wider corresponding luminescence spectrum, and can simultaneously label a plurality of target molecules, but has the disadvantages that the external excitation light source not only excites the molecular probes, but also can simultaneously excite other substances of organisms such as hair, cartilage and the like, i.e. autofluorescence phenomenon occurs, so that serious noise is generated, and the imaging signal-to-noise ratio is greatly reduced. The BLT has advantages in that it does not require excitation by an in vitro light source, and no auto-fluorescence phenomenon occurs, and thus its signal-to-noise ratio and sensitivity are higher, but has disadvantages in that the kinds of probes are very limited, and the fluorescence intensity is also generally very weak, and thus the requirements for an imaging apparatus are very severe.
In recent years, the synthesis of fluorescent nanomaterials has brought about a new optical molecular imaging technique, X-ray fluorescence tomography (XLCT), which uses specific fluorescent nanoprobes injected into biological tissues to generate fluorescence under X-ray excitation. Such fluorescent nanomaterials mainly contain rare earth atoms such as Eu, Lu, Ce, Pr, etc., which generate near infrared light when irradiated with X-rays, thereby realizing imaging of specific targets. Compared with the traditional optical molecular imaging technology, the XLCT has the advantages that the available probes are rich in types, the excited target light source intensity is strong, the self-interference can be avoided, and the sensitivity is high. In addition, due to the strong penetrating power of X-rays, the XLCT can realize deeper molecular imaging, and meanwhile, the traditional X-ray imaging technology can be combined to provide more prior information for the reconstruction of a fluorescence tomography image, so that the ill-conditioned degree of the reconstruction problem solution is reduced.
Currently, the existing XLCT imaging methods are mainly classified into a narrow beam XLCT imaging method and a cone beam XLCT imaging method. The narrow beam XLCT imaging method is characterized in that a beam limiter is used for restricting an X-ray light source into a beam line and then exciting fluorescent nano-particles, so that the X-ray only excites the fluorescent nano-particles on a path where the beam line passes through every imaging, then beam line scanning is carried out, and finally three-dimensional tomography is realized through a related reconstruction algorithm. Its formation of image is rebuild the process and is similar to traditional beam line X ray CT imaging technique, and the advantage can directly utilize the straight line penetrability of X ray bundle directly to fix a position the spatial position information of target light source, does not have the ill-conditioned nature that the fluorescence rebuild problem solved, and its spatial resolution depends on the cross-section size of X ray bundle, can realize higher spatial resolution. But the defects are that the scanning process is complex and tedious, the imaging time is long, the utilization efficiency of rays is low, and the like. The cone beam XLCT imaging method is a mainstream direction for developing the future XLCT imaging technology, the existing cone beam XLCT imaging method depends on the structural information of an imaging sample provided by the traditional cone beam X-ray CT imaging to serve as the prior information for XLCT reconstruction, and more key prior information such as material composition or charge density distribution of the sample cannot be provided, so that the existing cone beam XLCT reconstruction equation solution still faces serious ill-condition, the accuracy of three-dimensional reconstruction of the cone beam XLCT imaging method is greatly limited, and high-quality and high-resolution cone beam XLCT imaging is difficult to realize on a biological sample with complex structure and composition.
Disclosure of Invention
The present application provides a cone beam X-ray fluorescence imaging method, system, terminal and storage medium, which aims to solve at least one of the above technical problems in the prior art to a certain extent.
In order to solve the above problems, the present application provides the following technical solutions:
a cone-beam X-ray fluorescence imaging method, comprising:
acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device, and synchronously acquiring a surface fluorescence image of the imaging target;
carrying out information separation and extraction on the phase contrast projection image to obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of the imaging target, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image;
inputting the three-dimensional reconstruction images of the absorption contrast image, the refraction contrast image and the scattering contrast image and the surface fluorescence image into a trained XLCT three-dimensional reconstruction depth convolution neural network model, and outputting a cone beam XLCT fluorescence image of an imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the X-ray phase contrast-fluorescence imaging device comprises a phase contrast imaging light path and a fluorescence imaging light path, wherein the phase contrast imaging light path is a Talbot-Lau imaging device and is used for performing phase contrast imaging with three contrasts of absorption, refraction and scattering on an imaging target after an X-ray light source is started, and meanwhile, fluorescent nanoparticles in the imaging target generate excitation fluorescence under the excitation of the X-ray light source and are received by the fluorescence imaging light path to generate a surface fluorescence image of the imaging target.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the phase contrast imaging optical path comprises an X-ray light source, a source grating, a phase grating, a sample stage, an analytical grating and an X-ray flat panel detector, wherein the source grating, the phase grating, the sample stage, the analytical grating and the X-ray flat panel detector are all arranged on a three-axis precision motion platform, and the sample stage can horizontally and freely rotate by 360 degrees; the analytical grating can horizontally rotate at +/-45 degrees and deflect at +/-15 degrees in the vertical direction;
the fluorescence imaging optical path comprises a weak signal detector and a black box, and shares the same sample stage with the phase contrast imaging optical path; the black box is made of materials capable of shielding light waves in an infrared band and a visible light band, the sample stage and the weak signal detector are covered in the dark box, and an X-ray shielding layer is further arranged on the periphery of the weak signal detector.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the XLCT three-dimensional reconstruction depth convolutional neural network model comprises 12 convolutional layers, wherein the first convolutional layer is an input layer, the last convolutional layer is an output layer, and the other 10 convolutional layers are formed by serially or parallelly connecting multi-mode image information extraction structures;
the activating function of the XLCT three-dimensional reconstruction depth convolution neural network model is a Leaky-ReLU function.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the input layer comprises four input ends, namely a three-dimensional absorption contrast image input end, a three-dimensional refraction contrast image input end, a three-dimensional scattering contrast image input end and a surface fluorescence image input end.
The technical scheme adopted by the embodiment of the application further comprises the following steps: each multi-mode image information extraction structure is provided with three parallel convolution operation channels: the size of convolution kernels of two series convolution layers in a channel 1 is 1 x 1, and the channel 1 is used for ensuring that the resolution of a network output image is the same as that of an input image; the sizes of convolution kernels of two series convolution layers in the channel 2 are 1 x 3 and 3 x 1 respectively, the sizes of convolution kernels of two series convolution layers in the channel 3 are 1 x 5 and 5 x 1 respectively, and the channel 2 and the channel 3 are used for reflecting the corresponding relation possibly included in the operation between the neighborhood pixels in the two different mode images.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the cone beam XLCT fluorescence image of the imaging target output by the XLCT three-dimensional reconstruction depth convolution neural network model specifically comprises the following steps:
acquiring the internal structure information of the imaging target from the absorption contrast image, the refraction contrast image and the scattering contrast image, solving the charge density distribution information and the equivalent atomic number distribution information of the imaging target from the absorption contrast image and the refraction contrast image, taking the internal structure information, the charge density distribution information and the equivalent atomic number distribution information as priori knowledge, acquiring the three-dimensional space distribution information of the corresponding target fluorescent light source from the surface fluorescent image of the imaging target, and performing cone beam XLCT fluorescent image reconstruction on the imaging target.
Another technical scheme adopted by the embodiment of the application is as follows: a cone-beam X-ray fluorescence imaging system, comprising:
an image acquisition module: the device comprises a phase contrast imaging device, a fluorescence imaging device and a data processing device, wherein the phase contrast imaging device is used for acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device and synchronously acquiring a surface fluorescence image of the imaging target;
a phase contrast image processing module: the phase contrast projection image is used for carrying out information separation and extraction on the phase contrast projection image to obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of the imaging target, and the absorption contrast image, the refraction contrast image and the scattering contrast image are respectively subjected to three-dimensional reconstruction;
a fluorescence imaging module: and the XLCT three-dimensional reconstruction depth convolution neural network model is used for inputting the three-dimensional reconstruction images and the surface fluorescence images of the absorption contrast image, the refraction contrast image and the scattering contrast image into the trained XLCT three-dimensional reconstruction depth convolution neural network model and outputting the cone beam XLCT fluorescence image of the imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
The embodiment of the application adopts another technical scheme that: a terminal comprising a processor, a memory coupled to the processor, wherein,
the memory stores program instructions for implementing the cone-beam X-ray fluorescence imaging method;
the processor is configured to execute the program instructions stored by the memory to control cone beam X-ray fluorescence imaging.
The embodiment of the application adopts another technical scheme that: a storage medium having stored thereon program instructions executable by a processor for performing the cone-beam X-ray fluorescence imaging method.
Compared with the prior art, the embodiment of the application has the advantages that: the cone beam X-ray fluorescence imaging method, the cone beam X-ray fluorescence imaging system, the cone beam X-ray fluorescence imaging terminal and the storage medium of the embodiment of the application utilize an X-ray phase contrast-fluorescence imaging device to obtain a surface fluorescence image, an absorption contrast image, a refraction contrast image and a scattering contrast image of an imaging target, obtain internal structure information of the imaging target, solve charge density distribution information and equivalent atomic number distribution information of an imaging sample from the absorption contrast image and the refraction contrast image, use the internal structure information, the charge density distribution information and the equivalent atomic number distribution information as prior information required by cone beam XLCT three-dimensional reconstruction, extract three-dimensional space distribution information of a fluorescence light source of the imaging target from the internal information and the surface fluorescence information of the imaging target by utilizing a depth convolution network, and realize rapid high-resolution cone beam X-ray fluorescence imaging. Compared with the prior art, the method has at least the following advantages:
1. according to the method, the X-ray phase contrast-fluorescence imaging device is used for obtaining the prior information required by the cone beam XLCT three-dimensional reconstruction, and compared with the method that the traditional X-ray absorption contrast imaging is simply used as the prior information of the XLCT three-dimensional reconstruction, the ill-conditioned degree of the solution of the cone beam XLCT three-dimensional reconstruction equation can be effectively reduced.
2. The XLCT three-dimensional reconstruction depth convolution neural network model can more easily fuse information obtained by various modal imaging modes, can quickly and accurately reconstruct space distribution information of target fluorescence, can be conveniently migrated to observation and research experiments of other different life activities of similar biological samples by a trained model, does not need retraining, can be used for setting a specific experimental sample in the experimental process for training the network model so as to reduce the radiation dose suffered by the precious target experimental sample, and simultaneously improves the experimental efficiency.
Drawings
FIG. 1 is a flow chart of a cone beam X-ray fluorescence imaging method of an embodiment of the present application;
FIG. 2 is a schematic diagram of an X-ray phase contrast-fluorescence imaging apparatus according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a training process of an XLCT three-dimensional reconstruction deep convolutional neural network model according to an embodiment of the present application;
FIG. 4 is a schematic structural diagram of an XLCT three-dimensional reconstruction deep convolutional neural network model according to an embodiment of the present application;
FIG. 5 is a schematic structural diagram of a cone-beam X-ray fluorescence imaging system according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of a terminal according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a storage medium according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Aiming at the defects of the prior art, the cone beam X-ray fluorescence imaging method provided by the embodiment of the application acquires phase contrast projection images of an imaging target by using an X-ray phase contrast imaging method, synchronously acquires surface fluorescence information of the imaging target, extracts internal information such as the structure, composition and electron density distribution of the imaging target as prior information for XLCT reconstruction according to the phase contrast projection images, and extracts three-dimensional space distribution information of a fluorescence light source of the imaging target from the internal information and the surface fluorescence information of the imaging target by using a depth convolution network, thereby realizing the reconstruction of the cone beam XLCT fluorescence image of the imaging target.
Specifically, please refer to fig. 1, which is a flowchart of a cone beam X-ray fluorescence imaging method according to an embodiment of the present application. The cone beam X-ray fluorescence imaging method comprises the following steps:
s10: acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device, and synchronously acquiring a 360-surface fluorescence image of the imaging target;
in this step, please refer to fig. 2, which is a schematic structural diagram of an X-ray phase contrast-fluorescence imaging apparatus according to an embodiment of the present application. The X-ray phase contrast-fluorescence imaging device comprises a phase contrast imaging light path and a fluorescence imaging light path. The axes of the fluorescence imaging light path and the phase contrast imaging light path are in the same horizontal plane and are perpendicular to the phase contrast imaging light path. The phase contrast imaging light path is a Talbot-Lau imaging device and consists of a micro focal spot X-ray light source 31, a source grating 32, a phase grating 33, a sample table for placing a sample 3, an analysis grating 34 and an X-ray flat panel detector 35, wherein the source grating 32, the phase grating 33, the sample table, the analysis grating 34 and the X-ray flat panel detector 35 are all arranged on a three-axis precision motion platform, so that X, Y, Z three-dimensional electric control precision movement can be realized, and the sample table can realize 360-degree horizontal free rotation so as to meet the requirement of CT scanning; the analyzer grating 35 can realize a horizontal rotation of ± 45 degrees and a deflection in a vertical direction of ± 15 degrees, so as to meet the requirement of phase-contrast optical path adjustment and calibration.
The fluorescence imaging optical path comprises a weak signal detector 36 and a black box, and shares the same sample stage with the phase contrast imaging optical path, so that synchronous imaging can be realized. The black box is made of a material capable of shielding light waves in an infrared band and a visible light band, and covers the sample stage and the weak signal detector in the dark box 37. And a certain X-ray shielding layer is arranged on the periphery of the weak signal detector to prevent the X-ray from damaging electronic components of the weak signal detector.
After an X-ray light source of the phase contrast imaging light path is started, phase contrast imaging with multiple contrasts is carried out on an imaging target through the phase contrast imaging light path, and meanwhile, fluorescence nanoparticles in the imaging target generate excitation fluorescence under the excitation of X-rays and are received by a weak signal detector, so that surface fluorescence imaging of the imaging target is realized.
S11: carrying out information separation and extraction on the phase contrast projection image to respectively obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of the imaging target, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image;
in this step, the absorption contrast image, the refraction contrast image, and the scattering contrast image extracted from the phase contrast projection image may not only more comprehensively reflect the internal structure information such as the structure and composition of the imaging target, but also solve the charge density distribution information and the equivalent atomic number distribution information of the imaging target from the absorption contrast image and the refraction contrast image, where the charge density distribution information and the equivalent atomic number distribution information are key factors affecting the transmission property of fluorescence in the imaging target. In addition, the scattering contrast image reflects the scattering properties of the imaging target on X-rays, which also has a very strong correlation with the scattering properties of the imaging target on fluorescence. Therefore, the images with the three kinds of contrast can provide sufficient prior information for the network model, so that the ill-conditioned degree of solving the XLCT reconstruction problem is greatly reduced.
S12: inputting the three-dimensional reconstruction images of the absorption contrast image, the refraction contrast image and the scattering contrast image and the 360-surface fluorescence image into a trained XLCT three-dimensional reconstruction depth convolution neural network model, and outputting a cone beam XLCT fluorescence image of an imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
Please refer to fig. 3, which is a schematic diagram of a training process of an XLCT three-dimensional reconstruction deep convolutional neural network model according to an embodiment of the present application. The XLCT three-dimensional reconstruction deep convolutional neural network model training method comprises the following steps:
s20: collecting narrow-beam XLCT surface fluorescence images of a plurality of imaging samples, and performing three-dimensional reconstruction on the narrow-beam XLCT surface fluorescence images;
in this step, the collected narrow beam XLCT surface fluorescence image will be used as a training label of the network model.
S21: acquiring a phase contrast projection image of an imaging sample through an X-ray phase contrast-fluorescence imaging device, and synchronously acquiring a 360-surface fluorescence image of the imaging sample;
s22: separating and extracting information of the phase contrast projection image to respectively obtain an absorption contrast image, a refraction contrast image and a scattering contrast image, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image;
s23: taking the three-dimensional reconstruction images of the absorption contrast image, the refraction contrast image and the scattering contrast image and the 360-degree surface fluorescence image as training samples of a depth convolution network model, taking the three-dimensional reconstruction images of the narrow-beam XLCT surface fluorescence image as training labels of the depth convolution network model, and carrying out supervised learning training on the depth convolution network model to obtain a trained XLCT three-dimensional reconstruction depth convolution neural network model;
in this step, please refer to fig. 4, which is a schematic structural diagram of an XLCT three-dimensional reconstruction deep convolutional neural network model according to an embodiment of the present application. The model has 12 convolutional layers in total, wherein the first convolutional layer is an input layer and comprises four input ends, namely a three-dimensional absorption contrast image input end, a three-dimensional refraction contrast image input end, a three-dimensional scattering contrast image input end and a 360-surface fluorescence image input end. The last convolution layer of the network model is the output layer, which is the corresponding XLCT fluorescence image. The other 10 convolutional layers are formed by connecting multi-modal image information extraction structures in series or in parallel, wherein each multi-modal image information extraction structure is provided with three parallel convolution operation channels: the convolution kernels of two series convolution layers in the channel 1 are both 1 x 1 in size, and the channel can ensure that the resolution of a network output image is the same as that of an input image; the sizes of convolution kernels of the two series convolution layers in the channel 2 are 1 x 3 and 3 x 1 respectively, the sizes of convolution kernels of the two series convolution layers in the channel 3 are 1 x 5 and 5 x 1 respectively, and the channel 2 and the channel 3 are used for reflecting the corresponding relation possibly included in operation between the neighborhood pixels in the two different mode images. The activation functions used in the network model are all Leaky-ReLU functions.
Based on the network structure, charge density distribution information and equivalent atomic number distribution information of an imaging sample are solved from an absorption contrast image and a refraction contrast image, internal structure information, charge density distribution information and equivalent atomic number distribution information are used as priori knowledge, and an XLCT three-dimensional reconstruction depth convolution neural network model learns a physical mapping relation between a surface fluorescence image obtained by a weak signal detector and three-dimensional space distribution information of a target fluorescence light source, so that the ill-condition degree of solving an XLCT reconstruction problem is greatly reduced. The network model after learning and training can obtain the three-dimensional space distribution information of the corresponding target fluorescent light source from the surface fluorescent image of the new imaging target, thereby realizing the reconstruction of the cone beam XLCT fluorescent image of the imaging target.
Based on the above, in the cone beam X-ray fluorescence imaging method according to the embodiment of the present application, the surface fluorescence image, the absorption contrast image, the refraction contrast image, and the scattering contrast image of the imaging target are obtained by using the X-ray phase contrast-fluorescence imaging apparatus, the internal structure information of the imaging target is obtained by using the obtained internal structure information, the charge density distribution information and the equivalent atomic number distribution information of the imaging sample are solved from the absorption contrast image and the refraction contrast image, the internal structure information, the charge density distribution information, and the equivalent atomic number distribution information are used as prior information required for three-dimensional cone beam XLCT reconstruction, and then the three-dimensional spatial distribution information of the fluorescence source of the imaging target is extracted from the internal information and the surface fluorescence information of the imaging target by using the depth convolution network, so as to realize fast and high-resolution cone beam X-ray fluorescence imaging. Compared with the prior art, the method has at least the following advantages:
1. according to the method, the X-ray phase contrast-fluorescence imaging device is used for obtaining the prior information required by the cone beam XLCT three-dimensional reconstruction, and compared with the method that the traditional X-ray absorption contrast imaging is simply used as the prior information of the XLCT three-dimensional reconstruction, the ill-conditioned degree of the solution of the cone beam XLCT three-dimensional reconstruction equation can be effectively reduced.
2. The XLCT three-dimensional reconstruction depth convolution neural network model can more easily fuse information obtained by various modal imaging modes, can quickly and accurately reconstruct space distribution information of target fluorescence, can be conveniently migrated to observation and research experiments of other different life activities of similar biological samples by a trained model, does not need retraining, can be used for setting a specific experimental sample in the experimental process for training the network model so as to reduce the radiation dose suffered by the precious target experimental sample, and simultaneously improves the experimental efficiency.
Please refer to fig. 5, which is a schematic structural diagram of a cone beam X-ray fluorescence imaging system according to an embodiment of the present application. The cone beam X-ray fluorescence imaging system 40 of the embodiment of the present application includes:
the image acquisition module 41: the system comprises an X-ray phase contrast-fluorescence imaging device, a phase contrast projection image acquisition unit, a data acquisition unit and a data acquisition unit, wherein the X-ray phase contrast-fluorescence imaging device is used for acquiring a phase contrast projection image of an imaging target and synchronously acquiring a 360-surface fluorescence image of the imaging target;
phase contrast image processing module 42: the device is used for separating and extracting information of the phase contrast projection image, respectively obtaining an absorption contrast image, a refraction contrast image and a scattering contrast image of an imaging target, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image;
fluorescence imaging module 43: the X-ray imaging system is used for inputting the three-dimensional reconstruction images of the absorption contrast images, the refraction contrast images and the scattering contrast images and the 360-surface fluorescence images into a trained XLCT three-dimensional reconstruction depth convolution neural network model and outputting cone beam XLCT fluorescence images of an imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
Please refer to fig. 6, which is a schematic diagram of a terminal structure according to an embodiment of the present application. The terminal 50 comprises a processor 51, a memory 52 coupled to the processor 51.
The memory 52 stores program instructions for implementing the cone-beam X-ray fluorescence imaging method described above.
The processor 51 may also be referred to as a CPU (Central Processing Unit). The processor 51 may be an integrated circuit chip having signal processing capabilities. The processor 51 may also be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Fig. 7 is a schematic structural diagram of a storage medium according to an embodiment of the present application. The storage medium of the embodiment of the present application stores a program file 61 capable of implementing all the methods described above, where the program file 61 may be stored in the storage medium in the form of a software product, and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a mobile hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, or terminal devices, such as a computer, a server, a mobile phone, and a tablet.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A cone-beam X-ray fluorescence imaging method, comprising:
acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device, and synchronously acquiring a surface fluorescence image of the imaging target;
carrying out information separation and extraction on the phase contrast projection image to obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of the imaging target, and respectively carrying out three-dimensional reconstruction on the absorption contrast image, the refraction contrast image and the scattering contrast image;
inputting the three-dimensional reconstruction images of the absorption contrast image, the refraction contrast image and the scattering contrast image and the surface fluorescence image into a trained XLCT three-dimensional reconstruction depth convolution neural network model, and outputting a cone beam XLCT fluorescence image of an imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
2. The cone-beam X-ray fluorescence imaging method according to claim 1, wherein the X-ray phase contrast-fluorescence imaging apparatus comprises a phase contrast imaging optical path and a fluorescence imaging optical path, the phase contrast imaging optical path is a Talbot-Lau imaging apparatus, and is configured to perform phase contrast imaging with three contrasts of absorption, refraction and scattering on the imaging target after an X-ray light source is turned on, and meanwhile, fluorescent nanoparticles in the imaging target generate excited fluorescence under excitation of the X-ray light source and are received by the fluorescence imaging optical path, so as to generate a surface fluorescence image of the imaging target.
3. The cone beam X-ray fluorescence imaging method according to claim 2, wherein the phase contrast imaging optical path comprises an X-ray light source, a source grating, a phase grating, a sample stage, an analyzer grating and an X-ray flat panel detector, the source grating, the phase grating, the sample stage, the analyzer grating and the X-ray flat panel detector are all mounted on a three-axis precision motion platform, and the sample stage can horizontally freely rotate by 360 degrees; the analytical grating can horizontally rotate at +/-45 degrees and deflect at +/-15 degrees in the vertical direction;
the fluorescence imaging optical path comprises a weak signal detector and a black box, and shares the same sample stage with the phase contrast imaging optical path; the black box is made of materials capable of shielding light waves in an infrared band and a visible light band, the sample stage and the weak signal detector are covered in the dark box, and an X-ray shielding layer is further arranged on the periphery of the weak signal detector.
4. The cone beam X-ray fluorescence imaging method according to claim 1, wherein the XLCT three-dimensional reconstruction depth convolution neural network model comprises 12 convolution layers, the first convolution layer is an input layer, the last convolution layer is an output layer, and the other 10 convolution layers are formed by connecting multi-modal image information extraction structures in series or in parallel;
the activating function of the XLCT three-dimensional reconstruction depth convolution neural network model is a Leaky-ReLU function.
5. The cone-beam X-ray fluorescence imaging method of claim 4, wherein the input layer comprises four inputs, respectively a three-dimensional absorption contrast image input, a three-dimensional refraction contrast image input, a three-dimensional scattering contrast image input, and a surface fluorescence image input.
6. The cone beam X-ray fluorescence imaging method of claim 4, wherein each of the multi-modal image information extraction structures has three parallel convolution operation channels: the size of convolution kernels of two series convolution layers in a channel 1 is 1 x 1, and the channel 1 is used for ensuring that the resolution of a network output image is the same as that of an input image; the sizes of convolution kernels of two series convolution layers in the channel 2 are 1 x 3 and 3 x 1 respectively, the sizes of convolution kernels of two series convolution layers in the channel 3 are 1 x 5 and 5 x 1 respectively, and the channel 2 and the channel 3 are used for reflecting the corresponding relation possibly included in the operation between the neighborhood pixels in the two different mode images.
7. The cone beam X-ray fluorescence imaging method according to any one of claims 1 to 6, wherein the cone beam XLCT fluorescence image of the imaging target output by the XLCT three-dimensional reconstruction depth convolution neural network model is specifically:
acquiring the internal structure information of the imaging target from the absorption contrast image, the refraction contrast image and the scattering contrast image, solving the charge density distribution information and the equivalent atomic number distribution information of the imaging target from the absorption contrast image and the refraction contrast image, taking the internal structure information, the charge density distribution information and the equivalent atomic number distribution information as priori knowledge, acquiring the three-dimensional space distribution information of the corresponding target fluorescent light source from the surface fluorescent image of the imaging target, and performing cone beam XLCT fluorescent image reconstruction on the imaging target.
8. A cone-beam X-ray fluorescence imaging system, comprising:
an image acquisition module: the device comprises a phase contrast imaging device, a fluorescence imaging device and a data processing device, wherein the phase contrast imaging device is used for acquiring a phase contrast projection image of an imaging target through an X-ray phase contrast-fluorescence imaging device and synchronously acquiring a surface fluorescence image of the imaging target;
a phase contrast image processing module: the phase contrast projection image is used for carrying out information separation and extraction on the phase contrast projection image to obtain an absorption contrast image, a refraction contrast image and a scattering contrast image of the imaging target, and the absorption contrast image, the refraction contrast image and the scattering contrast image are respectively subjected to three-dimensional reconstruction;
a fluorescence imaging module: and the XLCT three-dimensional reconstruction depth convolution neural network model is used for inputting the three-dimensional reconstruction images and the surface fluorescence images of the absorption contrast image, the refraction contrast image and the scattering contrast image into the trained XLCT three-dimensional reconstruction depth convolution neural network model and outputting the cone beam XLCT fluorescence image of the imaging target through the XLCT three-dimensional reconstruction depth convolution neural network model.
9. A terminal, comprising a processor, a memory coupled to the processor, wherein,
the memory stores program instructions for implementing the cone-beam X-ray fluorescence imaging method of any of claims 1-7;
the processor is configured to execute the program instructions stored by the memory to control cone beam X-ray fluorescence imaging.
10. A storage medium having stored thereon program instructions executable by a processor to perform the method of cone beam X-ray fluorescence imaging according to any one of claims 1 to 7.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110576925.0A CN113367717B (en) | 2021-05-26 | 2021-05-26 | Cone beam X-ray fluorescence imaging method, system, terminal and storage medium |
| PCT/CN2021/138057 WO2022247234A1 (en) | 2021-05-26 | 2021-12-14 | Cone-beam x-ray fluorescence imaging method and system, and terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110576925.0A CN113367717B (en) | 2021-05-26 | 2021-05-26 | Cone beam X-ray fluorescence imaging method, system, terminal and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113367717A true CN113367717A (en) | 2021-09-10 |
| CN113367717B CN113367717B (en) | 2022-11-22 |
Family
ID=77571977
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110576925.0A Active CN113367717B (en) | 2021-05-26 | 2021-05-26 | Cone beam X-ray fluorescence imaging method, system, terminal and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113367717B (en) |
| WO (1) | WO2022247234A1 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022247234A1 (en) * | 2021-05-26 | 2022-12-01 | 中国科学院深圳先进技术研究院 | Cone-beam x-ray fluorescence imaging method and system, and terminal and storage medium |
| CN114137002B (en) * | 2021-11-18 | 2023-07-14 | 北京航空航天大学 | Low-dose X-ray differential phase contrast imaging method based on contrast enhancement |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101304684A (en) * | 2005-11-10 | 2008-11-12 | 皇家飞利浦电子股份有限公司 | Absorption and scattering map reconstruction for optical fluorescence tomography |
| US20090238334A1 (en) * | 2008-03-19 | 2009-09-24 | C-Rad Innovation Ab | Phase-contrast x-ray imaging |
| CN106137129A (en) * | 2016-06-28 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Fluorescent scattering optical tomography systems and method |
| CN108680589A (en) * | 2018-05-31 | 2018-10-19 | 北京航空航天大学 | X-ray grating differential phase contrast 3 D cone-beam computer tomography method based on transversion malposition grating and device |
| JP2018179795A (en) * | 2017-04-14 | 2018-11-15 | コニカミノルタ株式会社 | X-ray phase contrast imaging device |
| CN109147049A (en) * | 2018-07-28 | 2019-01-04 | 天津大学 | Image reconstruction method for X-ray optical dynamic therapy |
| US10634890B1 (en) * | 2018-10-26 | 2020-04-28 | General Electric Company | Miniaturized microscope for phase contrast and multicolor fluorescence imaging |
| CN111089870A (en) * | 2019-12-12 | 2020-05-01 | 中国科学院苏州生物医学工程技术研究所 | X-ray grating phase contrast imaging method and system based on two-time imaging, storage medium and equipment |
| CN111103275A (en) * | 2019-12-24 | 2020-05-05 | 电子科技大学 | PAT prior information assisted dynamic FMT reconstruction method based on CNN and adaptive EKF |
| CN112568923A (en) * | 2020-12-10 | 2021-03-30 | 中国科学院深圳先进技术研究院 | X-ray phase contrast image extraction method, device, terminal and storage medium |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102221565B (en) * | 2010-04-19 | 2013-06-12 | 清华大学 | X-ray source grating stepping imaging system and imaging method |
| DE102016204046B4 (en) * | 2016-03-11 | 2018-05-24 | Siemens Healthcare Gmbh | Method and X-ray device for interferometric X-ray 2D imaging |
| US11154196B2 (en) * | 2017-06-20 | 2021-10-26 | Siemens Healthcare Gmbh | Deep-learnt tissue deformation for medical imaging |
| JP7069670B2 (en) * | 2017-12-04 | 2022-05-18 | コニカミノルタ株式会社 | X-ray system |
| CN108169257B (en) * | 2017-12-08 | 2019-08-13 | 中国科学院高能物理研究所 | There are three types of the nanometer CT three-D imaging methods of quantitative imaging mechanism for tool |
| CN111915733B (en) * | 2020-08-11 | 2022-10-25 | 天津大学 | LeNet network-based three-dimensional cone-beam X-ray luminescence tomography method |
| CN113367717B (en) * | 2021-05-26 | 2022-11-22 | 中国科学院深圳先进技术研究院 | Cone beam X-ray fluorescence imaging method, system, terminal and storage medium |
-
2021
- 2021-05-26 CN CN202110576925.0A patent/CN113367717B/en active Active
- 2021-12-14 WO PCT/CN2021/138057 patent/WO2022247234A1/en unknown
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101304684A (en) * | 2005-11-10 | 2008-11-12 | 皇家飞利浦电子股份有限公司 | Absorption and scattering map reconstruction for optical fluorescence tomography |
| US20090238334A1 (en) * | 2008-03-19 | 2009-09-24 | C-Rad Innovation Ab | Phase-contrast x-ray imaging |
| CN106137129A (en) * | 2016-06-28 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Fluorescent scattering optical tomography systems and method |
| JP2018179795A (en) * | 2017-04-14 | 2018-11-15 | コニカミノルタ株式会社 | X-ray phase contrast imaging device |
| CN108680589A (en) * | 2018-05-31 | 2018-10-19 | 北京航空航天大学 | X-ray grating differential phase contrast 3 D cone-beam computer tomography method based on transversion malposition grating and device |
| CN109147049A (en) * | 2018-07-28 | 2019-01-04 | 天津大学 | Image reconstruction method for X-ray optical dynamic therapy |
| US10634890B1 (en) * | 2018-10-26 | 2020-04-28 | General Electric Company | Miniaturized microscope for phase contrast and multicolor fluorescence imaging |
| CN111089870A (en) * | 2019-12-12 | 2020-05-01 | 中国科学院苏州生物医学工程技术研究所 | X-ray grating phase contrast imaging method and system based on two-time imaging, storage medium and equipment |
| CN111103275A (en) * | 2019-12-24 | 2020-05-05 | 电子科技大学 | PAT prior information assisted dynamic FMT reconstruction method based on CNN and adaptive EKF |
| CN112568923A (en) * | 2020-12-10 | 2021-03-30 | 中国科学院深圳先进技术研究院 | X-ray phase contrast image extraction method, device, terminal and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| JIN WU，ET AL: "Fusion imaging of fluorescent and phase-contrast X-ray computed tomography using synchrotron radiation in medical biology", 《PROCEEDINGS OF SPIE》, vol. 6318, 31 December 2006 (2006-12-31), pages 631828 - 1 * |
| POUYAN MOHAJERANI，ET AL: "FMT-PCCT:Hybrid Fluorescence Molecular Tomography—X-Ray Phase-Contrast CT Imaging of Mouse Models", 《IEEE TRANSACTIONSON MEDICALI MAGING》, vol. 33, no. 7, 31 July 2014 (2014-07-31), pages 1434 - 1446, XP011552353, DOI: 10.1109/TMI.2014.2313405 * |
| 欧鸽: "基于多尺度分解的图像融合方法研究", 《中国优秀博硕士学位论文全文数据库（硕士） 信息科技辑》, no. 2, 15 February 2021 (2021-02-15), pages 138 - 1354 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022247234A1 (en) * | 2021-05-26 | 2022-12-01 | 中国科学院深圳先进技术研究院 | Cone-beam x-ray fluorescence imaging method and system, and terminal and storage medium |
| CN114137002B (en) * | 2021-11-18 | 2023-07-14 | 北京航空航天大学 | Low-dose X-ray differential phase contrast imaging method based on contrast enhancement |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022247234A1 (en) | 2022-12-01 |
| CN113367717B (en) | 2022-11-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11583237B2 (en) | Method and measuring apparatus for an X-ray fluorescence measurement | |
| US10285659B2 (en) | Stored luminescence computed tomography | |
| CN113367717B (en) | Cone beam X-ray fluorescence imaging method, system, terminal and storage medium | |
| Banhart | Advanced tomographic methods in materials research and engineering | |
| Zhang et al. | Quantitative imaging of Gd nanoparticles in mice using benchtop cone-beam X-ray fluorescence computed tomography system | |
| US9230319B2 (en) | Method of reconstructing a biological tissue image, and method and apparatus for acquiring a biological tissue image | |
| Chernomyrdin et al. | In vitro terahertz spectroscopy of gelatin-embedded human brain tumors: a pilot study | |
| US11145431B2 (en) | System and method for nanoscale X-ray imaging of biological specimen | |
| JP7424289B2 (en) | Information processing device, information processing method, information processing system, and program | |
| Lakshmanan et al. | Design and implementation of coded aperture coherent scatter spectral imaging of cancerous and healthy breast tissue samples | |
| Yu et al. | Medipix-based spectral Micro-CT | |
| US8565376B2 (en) | Method and apparatus for measuring properties of a compound | |
| Lun et al. | Focused x-ray luminescence imaging system for small animals based on a rotary gantry | |
| Michelet et al. | Quantitative reconstruction of PIXE-tomography data for thin samples using GUPIX X-ray emission yields | |
| Hu et al. | Full-angle optical imaging of near-infrared fluorescent probes implanted in small animals | |
| Meloni et al. | Cardiovascular Applications of Photon-Counting CT Technology: A Revolutionary New Diagnostic Step | |
| Yao et al. | Leveraging deep generative model for direct energy-resolving CT imaging via existing energy-integrating CT images | |
| Lun et al. | X-ray luminescence imaging for small animals | |
| CN109709127A (en) | Low scattered x-ray fluorescence CT imaging method | |
| CN110292392A (en) | A kind of fluorescence CT sparse angular projection Gamma spectrum analysis method | |
| US11437218B2 (en) | Apparatus and method for nanoscale X-ray imaging | |
| Morris et al. | Validation of coded aperture coherent scatter spectral imaging for normal and neoplastic breast tissues via surgical pathology | |
| Lun et al. | Fast three-dimensional focused x-ray luminescence computed tomography | |
| Gelli et al. | The new INFN-CHNet neutron imaging facility | |
| Arslan et al. | Efficient 3D reconstruction of whole slide images in melanoma |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |