Phase image capturing method and phase image capturing device using the same

Abstract

To provide a phase image capturing method and device with which it is possible to acquire the phase of an analyte having a high spatial resolution by at least one round of imaging, without requiring improvements on existing devices.SOLUTION: Provided is a phase image capturing method and device for entering a quantum beam from a radiation source to a detector via an analyte and at least one sheet of phase grating and obtaining the phase image of the analyte from a beam strength distribution in each pixel constituting the detector. A weighting process for adding weights depending on the similarity with phase value (or structure) of pixel obtained for each pixel is applied using a preliminarily given approximate phase image, and then absorption, visibility and phase are obtained, using the weights, from at least one sheet of measured image.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a phase imaging method for obtaining an image of a subject by using a phase change obtained by irradiating a quantum beam such as an X-ray, and further relates to a phase imaging apparatus using the same. [Background technology]

[0002] Unlike visible light, X-rays can penetrate matter, making it possible to observe the internal structure of an object. Today, X-rays are used in a wide range of fields, including X-ray photography in hospitals and non-destructive testing in industry. In normal X-ray photography, the intensity of X-rays changes as they pass through an object due to absorption by the object, forming a contrast image. In devices that use this type of absorption rate, the heavier the elements that make up the object, the greater the X-ray absorption, and an image with high contrast can be obtained. However, when observing weakly absorbing objects made of light elements (small atomic weights), such as soft tissues in living organisms, there is a problem in that it is difficult to achieve contrast.

[0003] In response to this, research into X-ray phase imaging methods has been underway since the 1990s, as shown in the following prior art documents. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. WO2004058070 [Patent Document 2] International Publication No. WO2010050483 [Non-patent literature]

[0005] [Non-Patent Document 1] https: / / doi.org / 10.1103 / PhysRevLett.108.111212 , Google Scholar Crossref , CAS 10. A. Bravin, P. Coan, and P. Suortti, “X-ray phase-contrast imaging: from pre-clinical applications to clinics,” Phys. I d. Biol. 58, R1 (2013).

Outdoor Tool2

Outdoor Tools3

Outdoor Tools 4

Direct Environment 5

[0006] These methods measure the change in phase that occurs when X-rays pass through an object. That is, they generate high-contrast images by taking advantage of the fact that the change in phase when X-rays pass through an object made of light elements is much larger than the change in absorption [Non-Patent Document 1]. In recent years, these methods have also been applied to radiation sources such as neutron beams [Non-Patent Document 2].

[0007] Among the many imaging techniques that have been developed, the phase imaging method, which uses a diffraction grating to exploit Talbot interference, has attracted particular attention because it is possible to use conventional X-ray tubes and create compact equipment that can be used in hospitals and other facilities.

[0008] On the other hand, in the phase imaging method, the fringe scanning method is known as a method for restoring the phase from a captured intensity distribution image [Non-Patent Documents 3, 4, Patent Document 1]. In this method, the diffraction grating is moved at least three times in the grating period direction to capture multiple images. Since the moving distance each time is almost at the submicron level, the device is required to have very high precision / stability, and strict control of temperature changes and other factors is required in the surrounding environment in which it is used. In addition, since three or more movements are required, it is difficult to apply this method to dynamic measurements of moving (or changing) objects, such as visual monitoring of cell cultures.

[0009] For this reason, the Fourier transform method has been proposed [Non-Patent Document 5, Patent Document 2]. In this method, the phase is calculated by processing the captured interference fringe image using Fourier transform, and the phase can be obtained with one capture. However, in the Fourier transform method, many high-frequency components are lost, resulting in a decrease in spatial resolution. This means that the fine structure of the subject is lost.

[0010] Other methods that have been proposed include a method using a special diffraction grating [Non-Patent Document 6] and a method using a one-dimensional sensor [Non-Patent Documents 7, 8], but all of these require new hardware.

[0011] Therefore, the present invention has been made in consideration of the above-mentioned conventional technology, and more specifically, an object of the present invention is to provide a phase imaging method that can acquire the phase of a subject with high spatial resolution by at least one imaging session without requiring any improvements to existing devices or equipment (particularly, without any improvements / changes to hardware), and a phase imaging device utilizing the same. [Means for solving the problem]

[0012] In order to achieve the above object, according to the present invention, first, a phase image acquisition method is proposed in which a quantum beam from a line source is incident on a detector via an object and at least one phase grating, and a phase image of the object is obtained from the intensity distribution of the beam at each pixel constituting the detector, in which the intensity distribution of the beam contains information on absorption (a0), visibility (V), and phase (φ), and it is assumed that the absorption (a0), visibility (V), and phase (φ) at at least three adjacent pixels are approximately identical by variable approximation of the image, and the absorption, visibility, and phase are obtained from at least one measurement image.

[0013] Furthermore, in order to achieve the above-mentioned object, the present invention proposes a phase imaging method in which a quantum beam from a radiation source is incident on a detector via a subject and at least one phase grating, and a phase image of the subject is obtained from the intensity distribution of the beam at each pixel constituting the detector, in which a weighting process is performed using a pre-given approximate phase image to assign a weight to each pixel according to the similarity with the phase value (or structure) of the pixel to be obtained, and then the absorption, visibility, and phase are obtained from at least one measurement image using the weights.

[0014] In addition, according to the present invention, there is proposed a phase image capturing device comprising a radiation source which generates a quantum beam, a means for holding an object to be irradiated with the beam from the radiation source, and a detector which receives the beam from the object via at least one phase grating, and further comprising a processing unit which obtains a phase image of the object from the intensity distribution of the beam at each pixel constituting the detector, wherein the processing unit is the phase image capturing method described above. Effect of the Invention

[0015] According to the present invention described above, there is provided a phase imaging method and a phase imaging apparatus that can acquire the phase of a subject with high spatial resolution by at least one imaging operation without requiring any improvement to existing devices or facilities (particularly, without any improvement / change to hardware), which has a practically excellent effect. [Brief description of the drawings]

[0016] [Figure 1] FIG. 2 is a diagram for explaining the principle of a method A according to an embodiment (Example 1) of the present invention. [Diagram 2] FIG. 13 is a diagram for explaining the principle of the above-mentioned method A. [Diagram 3] FIG. 13 is a diagram for explaining a modified example of the above method A. [Figure 4] FIG. 2 is a diagram for explaining a convex function minimization problem for three variables in the above method A. [Diagram 5] FIG. 1 is a diagram showing the configuration of a Talbot interferometer, which is a facility or device to which the above-mentioned method A can be applied. [Figure 6] FIG. 1 is a diagram showing the configuration of a Talbot-Lau interferometer, which is a facility or device to which the above-mentioned method A can be applied. [Figure 7] FIG. 2 is a diagram showing the configuration of a Lau interferometer, which is a facility or device to which the above-mentioned method A can be applied. [Figure 8] FIG. 11 is a diagram illustrating weights in a method B according to another embodiment (Example 2) of the present invention. [Figure 9]FIG. 13 is a diagram for explaining the calculation of weights in the above method B. [Figure 10] FIG. 13 shows an example of phase restoration using a numerical phantom according to methods A and B of the present invention, in comparison with the prior art. [Figure 11] FIG. 1 shows phase images of polymer spheres obtained by methods A and B of the present invention in comparison with the prior art. [Figure 12] FIG. 13 shows phase images obtained by methods A and B of the present invention using actual data photographed by a Talbot interferometer for a polymer, in comparison with the prior art. [Figure 13] FIG. 1 is a diagram showing an example of the configuration of a high-speed phase CT imaging device using a Talbot interferometer to which the phase imaging method and phase imaging device of the present invention can be applied. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings.

[0018] <Example 1> <Principle of the present invention: Method A> In devices that form interference fringes using a radiation source that generates so-called quantum beams, including X-rays, neutrons, and visible light, what is directly measured by a detector (two-dimensional) is not the phase but the intensity distribution of the Moire interference fringes (or the self-image due to the Talbot effect) that are deformed by the specimen. The intensity distribution measured by this detector can generally be written as follows:

[0019]

number

[0020] where a0 is absorption, V is visibility, φ is differential phase (or differential phase, sometimes simply written as "phase"), and f is a waveform that represents the sine wave of the Moiré interference fringes. In phase imaging, to find the phase, an image of the background is taken when there is no subject present and when there is, and here f is found from the image of the background. When it is not possible to photograph the background, it is also possible to estimate f using a local part of a single image where the change in the object's phase is small (i.e. where the deformation of the Moiré interference fringes is small).

[0021] Thus, in equation (1), there are three variables for each pixel (or "picture element") of the detector, and at least three different data are required to find the phase. As mentioned above, in the conventional fringe scanning method, the diffraction grating is moved M (M≧3) times to obtain multiple captured images. Then, the phase is found using the multiple captured data obtained, for example, by the least squares method. This method can find the phase with high spatial resolution, but as mentioned above, high accuracy is required in moving the diffraction grating. In the fringe scanning method, the distance the diffraction grating moves each time is uniform and is 1 / M of the period of the diffraction grating. The period of the diffraction grating is generally several μm, so the movement is performed at almost the submicron level. Therefore, strict stability is required for the device / equipment.

[0022] The present invention provides a method for determining three variables, namely, absorption (a0), visibility (V), and phase (φ), with at least one image capture. As explained above, at least three different data are required to determine the phase, but the present invention solves this problem by approximating the variables of the image.

[0023] Specifically, as shown in FIG. 1, for three adjacent pixels of a detector (two-dimensional), the above equation (1) assumes that the absorption (a0), visibility (V), and phase (φ) are almost identical, as follows:

[0024]

number

[0025]

number

[0026] Note that, as in the following explanation, only three pixels are used for the explanation. However, it is of course possible to use more pixels for the calculation. is also possible.

[0027] Then, from the above formula (1), it can be written as follows. This means that in pixel (m, n), one piece of data I -1 This means that (m,n) has increased.

[0028]

number

[0029] Furthermore, according to the same approximation, at least three pieces of data can be obtained, as shown below.

[0030]

number

[0031] As described above, according to the variable approximation of the image, since three pieces of data exist for three variables, it becomes possible to obtain the phase (φ).

[0032] <Example> According to the variable approximation of the image, that is, the approximation of the above equations (2) and (3), the data corresponding to one pixel increases from one to three, as shown in FIG.

[0033] Here, the above-mentioned image variable approximation is not limited to the vertical direction, but can be performed in various directions such as horizontal or diagonal, as shown in Fig. 3(a) to (c). For the purpose of explanation, it is assumed that the absorption (a0), visibility (V), and phase (φ) are almost the same in three adjacent pixels as a specific example. Naturally, as shown in Fig. 3(d) and (e), it is possible to obtain the absorption, visibility, and phase by the method of the present invention by assuming that the absorption (a0), visibility (V), and phase (φ) are almost the same in more surrounding pixels. Naturally, it is also possible to perform variable approximation with different numbers of pixels, such as assuming that the absorption and visibility are almost the same in five adjacent pixels and that the phase is almost the same in three adjacent pixels.

[0034] In this way, the phase can be found by solving a minimization problem from the data obtained by approximating the variables of the image. Specifically, the difference between the data obtained by the above formula (5) and the theoretical value of the above formula (1) is minimized (the difference between the measured value and the theoretical value, L p It can be found by solving the problem of minimizing the (p≧0) norm.

[0035]

number

[0036] However, I k (k=0,±1) are multiple data obtained by approximating the variables of the same image as in equation (5). If p=2, this becomes a least squares problem.

[0037] The equation (6) of this problem can be solved by various existing iterative methods (e.g., the steepest descent method) by providing an initial value in the following equation (7).

[0038]

number

[0039] The optimal solution {a0,V,φ} obtained then becomes the desired absorption (a0), visibility (V), and phase (φ), respectively.

[0040] However, because equation (6) contains a trigonometric function of the variable φ, it is a minimization problem of a non-convex function. Therefore, depending on how the initial value is chosen, it may converge to an incorrect solution (for example, a local minimum value). Therefore, in this invention, for the case of p≧1, we propose the following new analysis method.

[0041] First, the following variable transformation is performed on (a0, V, φ):

[0042]

number

[0043] With this transformation, equation (6) in question can be rewritten as follows:

[0044]

number

[0045] Equation (9) is a minimization problem of a convex function with respect to three variables (a0, bc, bs). Therefore, as shown in Figure 4, no matter what the initial value, it is guaranteed to converge to the correct optimal solution. After finding the optimal solution a0, bc, bs, the absorption (a0), phase (φ), and visibility (V) are found by the following variable transformation.

[0046]

number

[0047] Therefore, according to the phase imaging method (method A) of the present invention described above in detail, it is possible to obtain the phase of the subject with high spatial resolution by at least one imaging using existing devices and equipment, without requiring significant improvements to the hardware. That is, compared to the conventional fringe scanning method: A. The absence of movement of the diffraction grating eliminates the requirement for high stability in the fringe scanning method. B. Operation of the device is simplified since only one exposure is required. Furthermore, compared to the traditional Fourier transform method: C. By maintaining a higher spatial resolution, it is possible to provide images that reveal more fine internal structures (see Figures 10(b) and 11(c) below). D. In addition, because the analytical method converts the problem into a minimization problem of a convex function and solves it, it provides an optimal solution regardless of the specific initial values ​​given. This is particularly important in practical applications.

[0048] <Facilities and devices to which the method of the present invention can be applied> The above-mentioned phase imaging method (Method A) can be applied to existing devices and facilities, and can be applied to, for example, a Talbot interferometer (FIG. 5), a Talbot-Lau interferometer (FIG. 6), or a Lau interferometer (FIG. 7), which form interference fringes using an X-ray source (or a neutron source), as shown in outline in FIGS. 5 to 7.

[0049] As shown in FIG. 5, the X-ray Talbot interferometer uses an X-ray source 10 that generates X-rays, and is composed of a phase diffraction grating 21 that generates X-ray interference, an absorption diffraction grating 22 that converts the interference intensity distribution generated in the specimen S into a moire intensity distribution, and an X-ray detector 30. In this device, a periodic image (self-image) similar to a phase grating is generated at a certain distance from the grating due to interference, and this phenomenon is called the Talbot effect. A second diffraction grating is placed at the position where the self-image is generated to form moire interference fringes. Therefore, what is directly measured by the Talbot interferometer is not the phase but the intensity distribution of the moire interference fringes deformed by the specimen, and the intensity distribution I(m,n) measured by the image detector is sent to an image data processing unit including, for example, a CPU as a computing element and a RAM or ROM as a storage means, and is output as absorption (a0), visibility (V), and phase (φ) by a predetermined calculation process. In these interferometers, the two-dimensional image detector 30 that detects the interference fringes is configured by arranging a large number of pixels, which are semiconductor X-ray detection elements, in a plane, and may employ, for example, the electromagnetic wave detector and image detector disclosed in Japanese Patent Application Laid-Open No. 2002-26300, etc. Each pixel outputs the above-mentioned X-ray intensity (I).

[0050] Moreover, the Talbot-Lau interferometer shown in FIG. 6 further includes a diffraction grating 23 in front of the X-ray source 10, and the Lau interferometer shown in FIG. 7 includes a phase grating 24 between the X-ray source and the subject. In these interferometers, as in the above, the intensity distribution I(m,n) measured by the image detector 30 is sent to an image data processing unit 100 including, for example, a CPU as an arithmetic element and RAM, ROM, etc. as a storage means, and is output as absorption (a0), visibility (V), and phase (φ) by a predetermined arithmetic processing.

[0051] In this way, the phase imaging method of the present invention can be easily applied to existing devices and equipment, and by performing a predetermined calculation by the image data processing unit using the imaging data (intensity signal (I)) obtained by the device or equipment, it becomes possible to obtain the phase of the subject with high spatial resolution from at least one imaging.

[0052] <Example 2> <Method B> Next, the present invention provides a method for obtaining a solution with higher accuracy than the above-mentioned method A by performing at least one imaging operation when incomplete phase information is known in advance. Here, the above-mentioned known phase information can be obtained from a phase image obtained by the conventional Fourier transform method or the above-mentioned method A of the present invention.

[0053] In the following explanation of the present invention's method B, this known phase information will be referred to as an "approximate phase image."

[0054] In this method, as in invention method A, the phase is found by solving a minimization problem. The difference from invention method A is that a weight function W(m,n) is used to solve the minimization problem of a weighted energy function. Here, the method is explained by fixing one arbitrary pixel (m,n). Other locations can be calculated in the same way. This weight is determined by using an approximate phase image to search for points in pixel (m,n) with similar phase values ​​(or structures) over a wide range, and the more similar they are, the greater the weight is given. Then, the following weighted minimization problem is solved.

[0055]

number

[0056] However, I l,k (l, k = 0, ±1, …) are multiple data obtained by an approximation formula (such as formula (5) above), and b C and b S is a variable defined in equation (8). Weight W l,k is defined as follows:

[0057]

number

[0058] Here, W(m+l,n+k) (l,k=0,±1,...) represents the similarity in the phase values ​​between the pixel of interest (m,n) for which the phase is to be calculated and pixel (m+l,n+k) (l,k=0,±1,...), and the calculation method thereof will be described later.

[0059] Here, we will explain the effect of weights. For example, assume that pixel (m,n) is the boundary between objects, as shown in Figure 8, and there is a large difference in phase value between it and the adjacent pixel (m,n-1). In this case, the assumption in equation (3) above is incorrect. Therefore, the phase image obtained using the data for pixel (m,n-1) also has a large error.

[0060] In the method B of the present invention, this problem is alleviated by using a weight Wl,k. Specifically, first, as shown in FIG. 9, the weight is calculated using an approximate phase image. Since the difference between the phase values ​​of pixel (m,n) and pixel (m,n-1) is large and they are not similar to each other, the weight is small. Therefore, in the above formula (12), the weight W 0,-1 would be as follows:

[0061]

number

[0062] Then, by using this weight in the above equation (11), the data of pixel (m,n-1) is excluded from the calculation. This means that the approximation of the above equation (3) is transformed into a more accurate one. As a result, a more accurate phase can be obtained even in the presence of discontinuous boundaries.

[0063] Next, a method for calculating the weight function W(m,n) will be described. Various weights can be used as a method for calculating the weights. As an example, a specific weight calculation method using non-local means will be described below.

[0064] For example, suppose there is an approximate phase image as shown in FIG. 9. Consider a situation in which the phase value of a pixel of interest (A5) is to be calculated. In this case, the weight of each of the surrounding pixels is determined by the phase value (or structure) of the approximate phase image. For example, as shown in FIG. 9, the weight of each pixel (B5) is determined as follows by using a target block centered on the pixel of interest (A5) and a reference block centered on (B5).

[0065]

number

[0066] Here, C is a normalization constant for making the sum of the weights of all the pixels affected be 1.

[0067] Here, h1>0 is a parameter determined in advance. Equation (14) is a weight according to the similarity between blocks, and the more similar the phase values ​​(or structures) are, the larger the weight is given.

[0068] <Experimental Results> Next, in order to confirm the effects of the above-mentioned phase imaging method and phase imaging device of the present invention, a numerical phantom and an X-ray differential phase image were processed by methods A and B of the present invention. The method of creating a contrast image is explained below. In this experiment, first, a single piece of interference fringe data is generated by formula (1) using the assumed phase (φ), absorption (a0), and visibility (V) of the object. Using this interference fringe image, phase images obtained by the Fourier transform method and the present invention (method A or method B) are shown.

[0069] Fig. 10 shows an example of phase restoration using a numerical phantom simulating a CT image inside the human body, where Fig. 10(a) is a phase image obtained by the conventional Fourier transform method, Fig. 10(b) is a phase image obtained by the method A of the present invention, and Fig. 10(c) is a phase image obtained by the method B of the present invention. As is clear from these phase images, the phase images obtained by the method A or B of the present invention have clearly higher spatial resolution than the phase images obtained by the conventional Fourier transform method, and further, it can be seen that the boundaries and details are obtained more accurately in the phase images obtained by the method B compared to the phase images obtained by the method A.

[0070] FIG. 11 shows the results of a simulation using differential phase images of a polymer sphere (diameter 5 / 16 inch), and shows (a) a phase image of the subject (true phase image), (b) a phase image obtained by the Fourier transform method, (c) a phase image by method A of the present invention, and (d) a phase image obtained by method B of the present invention. As is clear from this result, the resolution is significantly reduced in the conventional method (Fourier transform method) of FIG. 11(b). On the other hand, in method A of the present invention of FIG. 11(c), the resolution is improved overall, and furthermore, in method B of the present invention of FIG. 11(d), the image is almost correctly restored up to the boundary.

[0071] FIG. 12 shows the results of phase reconstruction using actual data of a polymer photographed by a Talbot interferometer. The photographs were taken at Spring-8 (a large synchrotron radiation facility). The size of the photographed image is 2048*1024 pixels, and one period of the moiré fringes is 22 pixels. FIG. 12(a) shows a phase image obtained by the conventional Fourier transform method, FIG. 12(b) shows a phase image obtained by the method A of the present invention, and FIG. 12(c) shows a phase image obtained by the method B of the present invention. As is clear from these phase images, the phase images obtained by the method A or B of the present invention are less affected by noise compared to the phase images obtained by the conventional Fourier transform method, and the spatial resolution is clearly improved. Furthermore, it can be seen that the boundaries and details are more accurately obtained in the phase images obtained by the method B compared to the phase images obtained by the method A.

[0072] As described above in detail, the phase imaging method of the present invention makes it possible to acquire the phase of the subject with high accuracy in at least one imaging operation.

[0073] Compared with the conventional fringe scanning method, A: The requirement for high stability in the fringe scanning method is eliminated because no moving diffraction grating is required. B: Because phase images can be obtained in a single exposure, it becomes easier to operate the equipment and facilities. C: Because a phase image can be obtained in a single shot, dynamic measurements of moving or changing objects can be made more accurate.

[0074] In addition, compared to the traditional Fourier transform method, D: By maintaining a higher spatial resolution, it is possible to provide images that reveal more detailed internal structures. E: The accuracy of the method of the present invention depends on the period of the interference fringes to be photographed, but it is possible to obtain a phase image whose accuracy does not decrease significantly even if the period becomes larger (for example, several tens of pixels per period) than that normally used in the Fourier transform method (about 5 or 6 pixels per period). In X-ray phase imaging using a diffraction grating, visibility increases significantly as the period of the moiré fringes becomes larger, so this method can obtain images of better quality than conventional methods.

[0075] In the method B of the present invention, specific weight calculations are performed using non-local means, but of course, various other methods can be applied, such as filters such as bilateral filters, or other methods other than filters that require larger weights as images become more similar.

[0076] In the above-mentioned method B of the present invention, the weight is calculated using an approximate phase image. This approximate phase image is not limited to the image of the present method A or the Fourier transform method, and it is natural that a phase image obtained by any other method can be used for the weight calculation. For example, an image of the same object taken by another device or hardware, previous measurement data (if it exists), or an image of another object known to have a similar structure can be used. Furthermore, even if it is not an approximate phase image, it is possible as long as there is position information of the discontinuous place of the object. One example is using an absorption photographed image of the same object.

[0077] <Other equipment and devices to which the present invention can be applied> From the above, the phase imaging method and the phase imaging device according to the present invention are particularly suitable for use in a high-speed phase CT imaging device using a Talbot interferometer, as shown in Fig. 12, by taking advantage of the simplicity of operation (no movement of the diffraction grating). Note that fringe scanning is not performed here.

[0078] The present invention can be applied to technologies other than those mentioned above, and can also be applied to, for example, phase-shift X-ray CT, which generates an image of phase shift distribution from line integral data of phase shift distribution when irradiated with X-rays, PET (positron emission CT) and SPECT (single photon emission CT), which are nuclear medicine imaging devices that generate an image of the distribution of radioactive drugs administered inside the body, CT using waves such as ultrasound, microwaves, sound waves, and seismic waves, electron beam CT, and MRI (magnetic resonance imaging) using image reconstruction from projection data. That is, in the present invention, "object" or "image" refers to the spatial distribution of the physical quantity to be imaged.

[0079] In addition, the phase shift, the phase shift of the quantum beam, the diffraction, or the numerical value of the image including the diffraction can be extracted from the multiple sets obtained by adding or changing the position of an optical element, and the phase value can be restored using the extracted numerical value of the phase shift of the quantum beam, the diffraction, or the image including the diffraction.

[0080] The present invention provides a phase imaging method for obtaining an image of a subject by using a phase change obtained by irradiating a quantum beam such as an X-ray, and further provides a phase imaging device using the same. [Explanation of symbols]

[0081] 10: radiation source, 21-24: diffraction (phase) grating, S: specimen, 100: image data processing unit

Claims

1. A phase image capturing method for capturing a phase image of a subject, comprising: a quantum beam from a radiation source is incident on a detector via a subject and at least one phase grating; and a phase image of the subject is obtained from an intensity distribution of the beam at each pixel constituting the detector, the method comprising: Using a previously given approximate phase image, a weighting process is performed to give weights to each pixel according to the similarity between the phase value (or structure) of the pixel to be determined, and then A phase image capturing method, comprising the steps of: determining absorption, visibility, and phase from at least one measurement image using the weights.

2. In the phase image capturing method according to claim 1, the absorption in three adjacent pixels in a range in which the phase values ​​(or structures) are similar is ..., a 0 ',a 0 , a 0 "..., the visibility corresponding to the absorption in the three pixels is..., V', V, V"..., and the phase corresponding to the absorption in the three pixels is..., φ', φ, φ"..., then, [0010] A phase image capturing method comprising determining absorption, visibility, and phase from at least one measurement image.

3. In the phase image capturing method according to claim 2, the weighting of the difference between the measured value and the theoretically calculated value is L p A phase imaging method characterized in that absorption, visibility, and phase are obtained using (p≧0) norm minimization.

4. 4. The phase imaging method according to claim 3, wherein in the weighted Lp norm minimization, for p≧1, a non-convex minimization problem is converted into a convex function minimization problem by variable transformation to obtain absorption, visibility, and phase.

5. 2. The phase image photographing method according to claim 1, wherein the predetermined approximate phase image is a phase image obtained by a Fourier transform method.

6. 2. The phase imaging method according to claim 1, wherein the predetermined approximate phase image is an image of the same object captured by a different device, previous measurement data, or an image of another object known to have a similar structure.

7. 2. The phase image photographing method according to claim 1, wherein the weighting is performed by calculation using a non-local means, or by a filter such as a bilateral filter, or by a method in which the more similar the images are, the larger the weight is obtained.

8. 2. The phase image photographing method according to claim 1, wherein an absorption photographed image of a discontinuous portion of the subject is used instead of the approximate phase image.

9. A source that generates a quantum beam; a means for holding a subject to be irradiated with a beam from the radiation source; a detector that receives the beam from the subject through at least one phase grating, a processor for obtaining a phase image of the subject from a beam intensity distribution at each pixel constituting the detector, A phase imaging apparatus, characterized in that the processing unit executes the phase imaging method according to any one of claims 1 to 8 to obtain absorption, visibility, and phase.

Images