Radiation imaging device and radiation imaging method

Abstract

A control unit of a radiation imaging device according to the present invention includes: an exposure time control unit for controlling a sub-frame exposure time of an image sensor such that focused fluorescence at the depth of field, generated by radiation that reaches a scintillator, and unfocused fluorescence away from the depth of field are imaged discretely, for each radiation particle, on a light receiving surface of an image sensor; a focused signal discriminating unit for discriminating between sensor signals from a light receiving pixel group corresponding to spots of the focused fluorescence included discretely in each sub-frame, and sensors signal from a light receiving pixel group corresponding to spots of unfocused fluorescence; and an image signal processing unit for generating an image by overlaying sub-frame data based on the sensor signals from the light receiving pixel group corresponding to the spots of focused fluorescence.

Description

[Technical Field]

[0001] The present invention relates to a radiation imaging apparatus, and more particularly to a new radiation imaging apparatus and method that aims to improve spatial resolution by super-resolution imaging. [Background technology]

[0002] Radiation imaging devices, particularly those using X-rays, are used as nondestructive testing tools for microfabricated products such as semiconductor devices, and as excellent observation and analysis tools in a wide range of fields, from materials science and life science to fundamental physics. Indirect X-ray imaging devices used for such applications generally include a camera optical system that converts an X-ray image into a visible light image using a scintillator and exposes it to an image sensor, and a signal processing system that performs image processing based on the image signal from the image sensor to obtain an analytical image. Meanwhile, in the field of semiconductor integrated circuit devices, the trend toward finer linewidths has been rapid in recent years, necessitating the development of high-resolution X-ray imaging devices. To meet this demand, various improvements have been made to both the camera optical system and the signal processing system to increase resolution. One solution for improving the spatial resolution of a camera optical system containing a scintillator is disclosed in Patent Document 1 below. [Prior art document] [Patent documents] [Patent Document 1] JP 2019-20336 A Problem to be Solved

[0003] Patent Document 1 proposes that a so-called indirect radiography type X-ray imaging device employs a lens-integrated scintillator, known as a solid immersion lens (SIL), in which a fluorescent film serving as a scintillator for converting incident X-rays into visible light is formed on one surface of a lens-shaped substrate made of a solid optical material with a refractive index close to that of the fluorescent film, while maintaining a uniform refractive index. However, such solid immersion lenses are technically difficult to fabricate, and there are still challenges to overcome before they can be put to practical use. Specifically, practical use of such SILs requires an extremely thin scintillator film on the order of submicrons. Furthermore, the optical quality within the scintillator is easily degraded during thin-film processing, making it difficult to achieve high-precision optical design. Furthermore, even if an X-ray imaging device using a SIL like that described in Patent Document 1 were realized, the theoretical limit of the optical system would be reached, making it impossible to achieve spatial resolution significantly exceeding this limit. This poses a problem: it is unable to meet the demand for observing increasingly minute objects on the nanometer order. General Disclosure

[0004] The basic concept of the present invention is to capture multiple images of radiation reaching a scintillator with an extremely short exposure time such that the radiation reaching the scintillator is observed as a discrete visible point image in photon or particle units, discriminate the visible point images into in-focus and out-of-focus signals, and generate a radiation observation image using only the in-focus signals. By controlling the exposure time of the image sensor to be short, the X-ray image reaching the scintillator can be obtained as a group of spatially discrete visible point image subframes in units of one X-ray photon. Furthermore, by comparing the light reception pattern of the visible point image on the image sensor with the light reception pattern (Airy disk) of a theoretical in-focus visible point image corresponding to the numerical aperture of the optical system and the fluorescent wavelength of the scintillator, the in-focus and out-of-focus signals can be discriminated. Thus, even if the penetration distance of X-ray photons reaching the phosphor screen acting as a scintillator is not constant and visible light conversion occurs at a position outside the depth of field of the objective lens, by composing an observation image by overlapping a group of subframes of only in-focus signals, excluding such out-of-focus signals, it is possible to relax the requirements for the thickness of the phosphor screen and obtain a sharp image.

[0005] Another aspect of the present invention is characterized in that ultra-high definition X-ray images are achieved by applying a photon localization method to the focus signal discriminated and extracted as described above, which further identifies the center position of the focus signal point image with an accuracy finer than the pixel size of the image sensor. The photon localization method itself is a well-known method of super-resolution processing of image signals, but by using it in combination with the image signal processing of the focus signal subframe group described above, it becomes possible to obtain clearer, high-definition images.

[0006] In another aspect of the present invention, the energy of the X-rays transmitted through the sample to be imaged is further discriminated from the number of visible photons converted from the number of visible light corresponding to one X-ray photon in the discriminated and extracted focusing signal as described above, thereby making it possible to obtain specific information about the constituent materials of the sample to be observed with high accuracy.

[0007] In short, the present invention provides a radiation imaging apparatus comprising: a radiation camera having a scintillator that receives radiation that has passed through an object to be imaged from a radiation source and emits fluorescence having a longer wavelength than the radiation; an imaging optical system that forms an image on an imaging plane through an objective lens having a depth of field matched to a specimen plane set on the scintillator; and an image sensor having a light-receiving surface made up of a plurality of pixels arranged on the imaging plane; and a control unit that generates an image based on a sensor signal for each pixel output from the image sensor of the radiation camera, The control unit is characterized by comprising: an exposure time control unit that controls the exposure time of the image sensor so that in-focus fluorescence at the depth of field of the radiation reaching the scintillator and out-of-focus fluorescence outside the depth of field are discretely imaged for each particle of the radiation on the light-receiving surface of the image sensor; a focus signal discrimination unit that discriminates the sensor signals of the light-receiving pixel groups corresponding to the in-focus fluorescent spots and the sensor signals of the light-receiving pixel groups corresponding to the out-of-focus fluorescent spots, which are discretely contained in each frame, as the spread of each fluorescent spot; and an image signal processing unit that generates the image based on the sensor signals from the light-receiving pixel groups corresponding to the in-focus fluorescent spots. The radiation referred to in this invention can be interpreted as X-rays, ultraviolet rays, or particle rays such as electrons, but in the following explanation, X-rays will be used as the most preferred example of application of the invention.

[0008] One aspect of the present invention provides a radiation imaging apparatus. The radiation imaging apparatus may include a radiation camera having a scintillator that receives radiation from a radiation source that has passed through an imaging target and emits fluorescence having a longer wavelength than the radiation, an imaging optical system that forms an image on an image plane through an objective lens whose depth of field is matched to a specimen plane set on the scintillator, and an image sensor having a light-receiving surface composed of a plurality of pixels arranged on the image plane. The radiation imaging apparatus may include a control unit that generates an image based on sensor signals for each of the plurality of pixels output from the image sensor of the radiation camera. The control unit may include an exposure time control unit that controls the sub-frame exposure time of the image sensor so that in-focus fluorescence at the depth of field generated by radiation reaching the scintillator and out-of-focus fluorescence outside the depth of field are discretely imaged for each particle of radiation on the light-receiving surface of the image sensor. The control unit may include a focus signal discriminator that discriminates between point image sensor signals of light-receiving pixel groups corresponding to in-focus fluorescence spots discretely included in each sub-frame and point image sensor signals of light-receiving pixel groups corresponding to out-of-focus fluorescence spots. The control unit may include an image signal processor that generates an image by integrating sub-frame data based on point image sensor signals from light-receiving pixel groups corresponding to in-focus fluorescence spots.

[0009] The image signal processing unit may include a point image signal detection unit that processes a point image sensor signal from a group of light-receiving pixels corresponding to a spot of focused fluorescence, calculates the center position of the point image in the group of light-receiving pixels, and generates an image based on the center position.

[0010] The image signal processing unit may include a radiation energy decomposition image generating unit that sorts the light receiving sensor signal levels of each focused fluorescence spot generated in response to radiation having different energies and discretely incident on the scintillator on a particle-by-particle basis into a plurality of level ranges in a plurality of consecutive subframes, and generates an image corresponding to the energy distribution of the radiation based on the sensor signals of the focused fluorescence spots that fall within the same level range.

[0011] The exposure time control unit may control the exposure time in accordance with the intensity of the radiation so that the radiation detection frequency in the image sensor is 0.02 photons / pixel / subframe or less, and may obtain at least 6.4 × 105 consecutive imaging signals consisting of multiple subframes to generate an image.

[0012] The objective lens may include a front lens having a solid immersion lens configuration in which a fluorescent film as a scintillator is integrally formed on the radiation incident surface side.

[0013] The objective lens may include a front lens made up of a fluorescent film as a scintillator and a lens-shaped body on which the fluorescent film is integrally formed on the radiation incident surface side, and the fluorescent film may be made of a solid optical material to which a predetermined activator is added using the solid optical material constituting the lens-shaped body as a base material.

[0014] The objective lens may have a front lens consisting of a fluorescent film as a scintillator and a lens-shaped body on the incident surface side of which the fluorescent film is integrally formed, and the refractive index difference between the fluorescent film and the lens-shaped body may be less than 0.1%.

[0015] The objective lens may include a leading lens of a solid immersion lens configuration formed from a material having scintillator functionality.

[0016] The lens-shaped substrate may be formed from undoped garnet crystal, and the fluorescent film may be formed from undoped garnet crystal with an activator added.

[0017] The lenticular substrate may be formed of undoped gadolinium aluminum gallium garnet (GAGG), and the fluorescent layer may be formed of gadolinium aluminum gallium garnet (Ce:GAGG) doped with cerium as an activator.

[0018] The radiation imaging device may have a total radiation-to-signal conversion efficiency such that fluorescence generated by one photon or particle of radiation on the fluorescent film is detected as a point image signal of 50 electrons or more by a pixel on the light-receiving surface of the image sensor.

[0019] When the radiation is X-rays, the total X-ray-to-signal conversion efficiency ηTOTAL, which indicates the efficiency of conversion into a sensor signal detected by the image sensor, may satisfy the condition [ηTOTAL=ηLY(E)×ηCE(NA)×ηTO×ηVQE>50], where ηLY(E) is the luminous efficiency of the optical system, ηCE(NA) is the light recovery efficiency of the optical system which depends on the numerical aperture NA, ηTO is the transmittance of the optical system, and ηVQE is the quantum efficiency for fluorescence of the image sensor.

[0020] In one aspect of the present invention, there is provided a radiation imaging method, which may include a scintillator on the radiation entrance surface side for converting incident radiation into visible light, may include a solid immersion type front lens having a lens-shaped exit surface side, and may use a radiation imaging device mainly composed of a camera unit that observes a visible light image from the solid immersion type front lens with an image sensor having a light-receiving surface consisting of a plurality of pixels, may control the exposure time for each subframe of the image sensor to collect radiation incident on the scintillator as point image signals of discrete visible light spots for each particle, may then discriminate the point image signals in each subframe into point image signals of focused light spots emitted within the depth of field of the solid immersion type front lens and point image signals of out-of-focus light spots emitted outside the depth of field, and may generate a radiation observation image by superimposing subframe data based on the point image signals of the focused light spots.

[0021] One aspect of the present invention provides a radiation imaging method. The radiation imaging method includes a radiation imaging apparatus including a scintillator that receives radiation from a radiation source that has passed through an imaging target and emits fluorescence having a longer wavelength than the radiation, an imaging optical system that forms an image on an image plane through an objective lens whose depth of field is matched to a sample plane set on the scintillator, and an image sensor having a light-receiving surface composed of a plurality of pixels arranged on the image plane. The control step may include an exposure time control step of controlling a subframe exposure time of the image sensor so that in-focus fluorescence at the depth of field generated by the radiation reaching the scintillator and out-of-focus fluorescence outside the depth of field are discretely imaged for each radiation particle on the light-receiving surface of the image sensor. The control step may include a focus signal discrimination step of discriminating between sensor signals of photo-receiving pixel groups corresponding to in-focus fluorescence spots discretely included in each subframe and sensor signals of photo-receiving pixel groups corresponding to out-of-focus fluorescence spots. The control step may include an image signal processing step for generating an image by superimposing sub-frame data based on sensor signals from groups of light-receiving pixels corresponding to spots of focused fluorescent light.

[0022] The radiation is X-rays, the radiation camera is an X-ray camera including a scintillator and an objective lens optically designed so that the conversion efficiency from X-rays to a sensor signal detected by an image sensor is 50 or more electrons per X-ray photon, and the control step may further include a step of performing signal processing to detect a point image center position of the focused fluorescent spot. The control step may further include a step of generating a super-resolution X-ray image by superimposing subframe data made up of point image center data of the detected point image center position.

[0023] According to the present invention, since an image is constructed by extracting only the focusing signal on the scintillator, it is possible to obtain a high-quality X-ray image with good contrast, and at the same time, there is an advantage in that it is not necessary to thin the scintillator. This makes it possible to provide a high-optical quality solid immersion lens at a low cost, with distortion suppressed during scintillator processing. Furthermore, by applying the photon localization method to signal processing, it is possible to achieve super-resolution X-ray imaging on the order of nanometers, which exceeds the theoretical spatial resolution limit of a solid immersion lens. [Brief explanation of the drawings]

[0024] [Figure 1] 1 shows a schematic configuration of a radiation imaging apparatus 100 according to the present embodiment. [Figure 2] 1(a) to 1(d) show a schematic configuration of the SIL 11 in this embodiment. [Figure 3] 10(a) and 10(b) are diagrams for explaining focused fluorescence F1 and unfocused fluorescence F2 in this embodiment. [Figure 4] 1(a) to 1(d) are diagrams for explaining imaging using the photon localization method. [Figure 5] 10(a) to 10(c) are diagrams for explaining the signal density of fluorescent spots. [Figure 6] FIG. 2 is a diagram showing the flow of image signal processing in the present embodiment. [Figure 7] 4 is a flowchart showing the imaging operation of the radiation imaging apparatus 100 according to the present embodiment. [Figure 8] 10A to 10C are diagrams illustrating image processing in another embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0025] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention as claimed. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention. [Schematic configuration of imaging device]

[0026] 1 shows a schematic configuration of a radiation imaging apparatus 100 according to this embodiment. The radiation imaging apparatus 100 includes a radiation camera 10 and a control unit 20. The radiation imaging apparatus 100 is an apparatus that acquires an image of a sample 900 at a predetermined spatial resolution using radiation from a radiation source 200. The radiation camera 10 includes a solid immersion lens (hereinafter also referred to as SIL) 11, a post-objective lens group 12, a band-pass filter 13, an imaging lens 14, and an image sensor 15. The solid immersion lens (SIL) 11 and the post-objective lens group 12 form an objective lens 16. The objective lens 16, the band-pass filter 13, and the imaging lens 14 form an imaging optical system 17.

[0027] The control unit 20 has a camera control unit 21 and an image signal processing unit 22. The camera control unit 21 includes an exposure time control unit 21a and an image data readout control unit 21b. The exposure time control unit 21a has a so-called camera shutter function and controls the subframe exposure time of the image sensor 15 so that radiation that has reached the SIL 11 and been converted into visible fluorescence is discretely imaged on the light-receiving surface of the image sensor 15 as fluorescent spots corresponding to single particles of radiation. The image data readout control unit 21b reads out sensor signals from the pixel array that constitutes the light-receiving surface of the image sensor 15 for each subframe and sends a group of subframe data to the image data storage unit 22b.

[0028] The image signal processing unit 22 includes a focus signal discriminator 22a, an image data storage unit 22b, a point image signal detector 22c, and an X-ray energy-resolved image generator 22d. The focus signal discriminator 22a discriminates between the sensor signals of the light-receiving pixel groups corresponding to the in-focus fluorescent spots discretely included in each subframe and the sensor signals of the light-receiving pixel groups corresponding to the out-of-focus fluorescent spots by comparing them with a theoretical in-focus spot pattern (Airy disk). The image data storage unit 22b stores the image data output from the image sensor 15.

[0029] The point image signal detector 22c and the X-ray energy-resolved image generator 22d perform additional functions for photon localization processing, which will be described later. The point image signal detector 22c processes point image sensor signals from photosensitive pixel groups corresponding to focused fluorescence spots to calculate the center positions of point images in the photosensitive pixel groups and generate subframe groups based on the calculated center positions. The X-ray energy-resolved image generator 22d sorts the photosensitive sensor signal levels of each focused fluorescence spot, which is generated in response to radiation particles with different energies that discretely enter the SIL 11, into multiple level ranges in multiple consecutive subframes, and generates an image corresponding to the radiation energy distribution based on the sensor signals of focused fluorescence spots that fall within the same level range. A display 41 and a memory unit 42 are connected to the image signal processor 22, and the captured image generated by the image signal processor 22 is displayed on the display 41.

[0030] The radiation may be gamma rays, X-rays, electron beams, or ultraviolet rays (especially, extreme ultraviolet rays, for example). When the radiation is X-rays, the radiation imaging device 100 can be used for X-ray imaging, X-ray CT, X-ray laminography, X-ray topography, X-ray imaging microscopes, and the like. Hereinafter, in this embodiment, X-rays are used as the radiation, and are indicated as X-rays R in FIG. 1.

[0031] As shown in FIG. 1, X-rays R from a radiation source 200 are intensity-controlled via a radiation-attenuating plate 400 and then irradiated onto a sample 900. The X-rays R that pass through the sample 900 or are scattered by the sample 900 are incident on a fluorescent film 11a (as a scintillator) that is integrally provided on the front surface of a lenticular substrate 11b that constitutes the front lens of an objective lens 16. The lens-integrated SIL 11, consisting of the fluorescent film 11a and the lenticular substrate 11b with a convex lens-shaped exit surface, constitutes an optical element known as a solid immersion lens. The fluorescent film 11a converts the incident X-rays R into fluorescent light F, which is visible light that can be received (sensed) by an image sensor 15. The image of the fluorescent light F from the SIL 11 is magnified (or reduced in some cases) by an imaging optical system 17 and projected onto the image sensor 15, thereby obtaining structural information about the sample 900 at a predetermined spatial resolution.

[0032] The theoretical limit of spatial resolution in such a radiation imaging device 100 is expressed by the Rayleigh resolution of the camera optical system, as in an optical microscope. In a radiation imaging device 100 with a scintillator-integrated solid immersion lens at the front, the spatial resolution δ is expressed as "δ = 0.61 × λ / NA." Here, λ represents the wavelength of the fluorescence F, which is visible light converted by the phosphor screen 11a. NA represents the numerical aperture of the camera optical system and is defined as "NA = n sin θ." n represents the refractive index of the lenticular substrate 11b itself, which serves as the medium filling the space between the phosphor screen 11a and the lenticular substrate 11b, which forms the imaging target surface. θ represents the aperture angle, which represents the maximum angle with respect to the optical axis of visible light incident from the phosphor screen 11a onto the convex surface of the lenticular substrate 11b.

[0033] For the sake of concrete explanation, a three-dimensional Cartesian coordinate system is assumed, which is made up of mutually orthogonal X-, Y-, and Z-axes. The direction of propagation of X-rays R irradiating sample 900 is parallel to the Z-axis, and therefore the direction of propagation of X-rays R transmitted through sample 900 is also parallel to the Z-axis.

[0034] A fluorescent film 11a is provided on the surface of the SIL 11 on the incident side of the X-rays R. The portion of the SIL 11 on which the fluorescent film 11a is not formed is called the lenticular substrate 11b. Therefore, the space between the lenticular substrate 11b and the fluorescent film 11a is filled with the material of the lenticular substrate 11b as a medium, and no gaps exist. The fluorescent film 11a is made of a fluorescent material that acts as a scintillator that receives X-rays R that have passed through the imaging target sample 900 and emits fluorescent light F, which is visible light.

[0035] The fluorescence F emitted from the fluorescent film 11a is typically in a wavelength band to which the image sensor 15 is sensitive. It may be light in the visible or ultraviolet region. For example, it has a wavelength in the range of 200 nm to 700 nm.

[0036] The fluorescence F generated on the fluorescent film 11a by excitation of the incident X-ray photons starts from a generation position corresponding to the depth of incidence of the X-rays R and passes through the lens-shaped body 11b and the objective rear lens group 12 to become parallel light, and is imaged on the image sensor 15 via the imaging lens 14 arranged in the rear stage.

[0037] The post-objective lens group 12 is composed of one or more lenses. The post-objective lens group 12, in cooperation with the lens-shaped body 11b of the SIL 11, converts the fluorescence F from the fluorescent film 11a into parallel light. The parallel fluorescence F passes through a bandpass filter 13, which filters only wavelengths in a specific frequency band, before entering the imaging lens 14. The imaging lens 14 forms an image of the parallel fluorescence F on the imaging surface of the image sensor 15. By optically designing the SIL 11, including the fluorescent film 11a, the post-objective lens group 12, and the imaging lens 14, various aberrations can be reduced, and spatial resolution and image quality can be improved. A configuration without the bandpass filter 13 is also possible. While the explanation of FIG. 1 above shows an example in which the lens system is configured as an infinity-corrected optical system, the SIL 11, the post-objective lens group 12, and the imaging lens 14 may also be optically designed as a single objective lens 16, which is configured as a finite-corrected optical system. In this case, the light emitted from the post-objective lens group 12 does not have to be parallel, and the band-pass filter 13 does not have to be included.

[0038] The focusing of the imaging optical system 17 is performed by arranging any of the SIL 11, the post-objective lens group 12, the imaging lens 14, and the image sensor 15 so that their positions can be adjusted, and by adjusting their positions, an image formed by the fluorescence F is formed on the image sensor 15.

[0039] It is also possible to have a lens support that supports at predetermined positions each of the lenses that make up the imaging optical system 17, including the objective lens 16 and the imaging lens 14. All of the lenses that make up the imaging optical system 17 may be supported by a single lens support, or the lens support may be composed of multiple lens supports.

[0040] Image sensor 15 is a so-called two-dimensional image sensor, such as a CCD (Charge Coupled Device) image sensor or a CMOS (Complementary Metal Oxide Semiconductor) image sensor, that has a light-receiving surface on which multiple photoelectric conversion pixels are arranged vertically and horizontally. This photoelectric conversion generates an image signal, which is an electrical signal. The image formed on image sensor 15 corresponds to the emission position and emission intensity of fluorescence F on fluorescent film 11a, and the emission position and emission intensity of fluorescence F on fluorescent film 11a depend on the incident position and incidence intensity of X-rays R on fluorescent film 11a.

[0041] The exposure time (sensitive time) of one subframe of an imaging signal representing an image (optical image) formed on the light-receiving surface is controlled to be extremely short as described below by a control signal from an exposure time control unit 21a in the camera control unit 21, and an image data signal of a visible light spot corresponding to one X-ray photon can be output in response to a signal from an image data readout control unit 21b. The image signal processing unit 22 processes the image data signal of the visible light spot in a subframe output from the image sensor 15 in a focus signal discrimination unit 22a to discriminate between in-focus and out-of-focus signals, and constructs an observation image by obtaining multiple subframes containing only in-focus signals, thereby making it possible to obtain an extremely clear, high-definition image.

[0042] The control unit 20 is connected to a stage control unit 300 external to the radiation imaging apparatus 100 and can issue commands to the stage control unit 300. Based on commands received from the control unit 20, the stage control unit 300 can move the SIL stage 11f on which the SIL 11 is mounted and the sample stage 910 on which the sample 900 is mounted to appropriate positions. Based on commands received from the control unit 20, the stage control unit 300 can also insert and remove the radiation attenuation plate 400. Based on commands received from the control unit 20, the control unit 20 is also connected to a radiation source control unit 500 and can control the intensity, timing, and irradiation time of the X-rays R emitted from the radiation source 200. Note that a stage may be mounted on any one of the SIL 11, the post-objective lens group 12, the imaging lens 14, and the image sensor 15 to perform focus adjustment, or multiple stages may be mounted on any combination of these.

[0043] The size ε of the blur caused by the diffraction of X-rays R within the sample 900 is expressed as ε = √λ √d, where λ is the wavelength of X-rays R and d is the distance between the sample 900 and the fluorescent film 11a; the larger the distance d between the sample 900 and the fluorescent film 11a, the larger the blur caused by X-ray diffraction. Therefore, to minimize the blur caused by the diffraction of X-rays R, the sample 900 is moved and positioned by the sample stage 910 so that it is close to or in close contact with the fluorescent film 11a. Furthermore, by bringing the center of the radiation source 200 and the sample 900 closer to each other, the intensity per unit area of ​​the X-rays R on the fluorescent film 11a can be increased.

[0044] 1, an X-ray imaging optical system that diffracts, refracts, or reflects X-rays R may be disposed between the sample 900 and the fluorescent film 11a, and an X-ray image of the sample 900 may be formed on the fluorescent film 11a at a magnified, reduced, or life-size magnification, thereby preventing blurring due to X-ray diffraction. At the same time, an X-ray illumination optical system may be disposed between the radiation source 200 and the sample 900. [Lens-integrated scintillator, solid immersion lens]

[0045] The following describes the lens-integrated scintillator used in this embodiment, the so-called solid immersion lens (SIL) 11. Figures 2(a) to 2(c) show a schematic configuration of the SIL 11 in this embodiment. The SIL 11, which is the premise of this embodiment, is based on the configuration disclosed in the previously referenced Patent Publication 1 and may be hemispherical as shown in Figure 2(a) or hyperhemispherical as shown in Figure 2(b). Figures 2(a) and 2(b) show a type of SIL 11 having a fluorescent film 11a and a lenticular substrate 11b as separate components, which serve as scintillators. Figure 2(c) shows a type in which the entire SIL 11 is integrally formed from a lenticular fluorescent substrate 11e, which is a solid optical material having the same scintillator function. That is, the SIL 11 in Figure 2(c) is entirely formed from a lenticular fluorescent substrate 11e that emits fluorescence. In this embodiment, the SIL 11 will be described mainly as having a fluorescent film 11a formed on the entrance surface side of a lens-shaped body 11b as shown in Fig. 2(a), and having the exit surface side processed into a convex lens shape that serves as the front lens of the objective lens 16. Note that the SIL 11 may also be in the shape of a flat plate as shown in Fig. 2(d).

[0046] SIL 11, which includes fluorescent film 11a and lenticular substrate 11b, is formed from a material transparent to fluorescence F, fluorescent film 11a absorbs radiation energy and emits fluorescence F, and lenticular substrate 11b does not emit fluorescence even when it absorbs radiation energy, or does not emit fluorescence with the same wavelength as fluorescence F, or the intensity of the fluorescence is lower than that of fluorescent film 11a. As long as this characteristic is maintained, the type of material constituting SIL 11 is arbitrary. In this embodiment, fluorescent film 11a in SIL 11 is formed from cerium-doped gadolinium aluminum gallium garnet (Ce:GAGG), and lenticular substrate 11b is formed from undoped gadolinium aluminum gallium garnet (GAGG).

[0047] Fluorescent film 11a is made of the same material as that of lenticular substrate 11b (i.e., the material of lenticular substrate 11b), to which a predetermined proportion of Ce has been added as an activator. Therefore, the difference in refractive index between fluorescent film 11a and lenticular substrate 11b is kept to 0.1% or less. Fluorescent film 11a, which functions as a scintillator, receives X-rays R and emits fluorescent light F in the visible range of, for example, 520 nm.

[0048] By optically designing the SIL 11 and the post-objective lens group 12 as a single objective lens 16 with the specimen surface on the surface of the fluorescent film 11a of the SIL 11, a solid immersion objective lens with high spatial resolution and reduced aberrations is achieved. For example, if the fluorescent film 11a is made of Ce:GAGG and the aperture angle of the objective lens 16 is 72 degrees, the numerical aperture NA = n·sinθ = 1.85 × 0.95 = 1.76. Therefore, the spatial resolution reaches δ = 0.61 × λ / NA = 0.61 × 520 nm / 1.76 = 180 nm.

[0049] The material of the SIL 11, consisting of the fluorescent film 11a and the lenticular substrate 11b, is not limited to this, and various solid optical materials can be used. The material (solid optical material) of the SIL 11 must be a transparent solid that does not exhibit birefringence at the wavelength of the fluorescent F. Examples include single crystals and polycrystals with an amorphous or cubic crystal structure, such as glass, which are optically isotropic. In the case of polycrystals, transparency is achieved by eliminating voids within the crystal and making the size of the crystal grain boundaries sufficiently smaller than the wavelength of the fluorescent F to maintain a continuity of the refractive index at the wavelength of the fluorescent F. Crystals with an optically anisotropic crystal structure may also be used, but are limited to polycrystals. In this case, transparency is achieved by eliminating voids within the crystal and making the size of not only the crystal grain boundaries but also the crystal grains sufficiently smaller than the wavelength of the fluorescent F to prevent birefringence and maintain a continuity of the refractive index at the wavelength of the fluorescent F.

[0050] As long as the above conditions are met, in this embodiment, the material may be an inorganic substance, an organic substance, or a mixture of two or more of these.

[0051] That is, when SIL11 is made of an inorganic material, a specific scintillator material having an optically isotropic cubic crystal structure is Lu3Al5O 12 (LuAG), Y3Al5O 12 (YAG), Gd3Ga5O 12 (GGG), sesquioxide crystals Lu2O3, Y2O3, Gd2O3, Sc2O3, fluorite crystals CaF2, BaF2, iodide crystals CsI, NaI. Also, Cs2HfC l6 Cubic scintillator materials include the perovskite crystals LuAlO3 (LuAP), YAlO3 (YAP), and GdAlO3 (GAP), the silicate crystals Lu2SiO5 (LSO), Y2SiO5 (YSO), LuYSiO5 (LYSO), and Gd2SiO5 (GSO), the pyrochlore crystals Lu2SiO2O7 (LPS), Y2Si2O7 (YPS), Gd2Si2O7 (GPS), and (Gd,La)2Si2O7 (La-GPS), the tungstate crystal PbWO4, the colquirete crystal LiCaAlF6, the oxysulfide crystals Lu2O2S, Y2O2S, and Gd2O2S, and other inorganic crystals such as Bi4Ge3O. 12 , Yb2SiO5, LaBr3, CeBr3, or SrI2 may also be used. Also, (Gd, Lu, Y)3(Al, Ga)5O 12 Alternatively, a crystal having a composition in which two or more elements are present at each site in the crystal in any ratio, such as (Lu,Y)2SiO5, may be used.

[0052] Fluorescent film 11a is formed by adding a predetermined proportion of an activator additive to a base material having the same composition as the constituent substance of lenticular body 11b (i.e., the material of lenticular body 11b). The activator emits fluorescence F having a wavelength that can be sensed by image sensor 15, and an element suited to the base material is selected and adjusted to a concentration and oxidation number suited to the base material so that the luminescence intensity and luminescence decay time are desired. In this embodiment, fluorescent film 11a is made of Ce:GAGG, which is GAGG doped with Ce. When exposed to X-rays R, fluorescent film 11a emits fluorescence F of, for example, 520 nm.

[0053] The thickness of the fluorescent film 11a is thinner than the thickness of the lenticular substrate 11b, and is, for example, within a range of 50 nm to 2 mm. By making the thickness of the fluorescent film 11a smaller than the depth of field of the imaging optical system 17, it is possible to prevent the generation of fluorescence F that is out of focus on the image sensor 15, which is the imaging plane, and to suppress a decrease in the spatial resolution of the radiation imaging device 100. In order to achieve such high spatial resolution, it is preferable to make the fluorescent film 11a of the solid immersion lens thin. However, according to the concept of this embodiment in which an X-ray image is constructed using only signals from focused fluorescent spots generated on the fluorescent film 11a, the difficulty of producing a high-quality thin fluorescent film can be alleviated.

[0054] The fluorescent film 11a is formed with a continuum of refractive index relative to the lenticular substrate 11b. "Formed with a continuum of refractive index" means that the difference in refractive index between the fluorescent film 11a and the lenticular substrate 11b is small, they are directly bonded without any intervening foreign material, and the refractive index is substantially uniform. For example, the difference in refractive index between the lenticular substrate 11b and the fluorescent film 11a is less than 0.1% of the refractive index of the lenticular substrate 11b. This suppresses scattering, refraction, and reflection of fluorescence at the bonding interface. Regarding the bonding method, to ensure the continuum of refractive index, it is desirable that the fluorescent film 11a and the lenticular substrate 11b are directly bonded together by solid-phase diffusion bonding, surface activated bonding, or the like, and that there are no voids at the bonding surface. If a substantially uniform refractive index is achieved, the fluorescent film 11a may be formed directly on the lenticular substrate 11b by a gas-phase method such as physical vapor deposition (PVD) or chemical vapor deposition (CVD). For example, PVD includes thermal evaporation, ion beam evaporation, molecular beam epitaxy, ion plating, sputtering, and pulsed laser deposition. CVD includes thermal CVD, plasma CVD, photo CVD, and mist CVD. In particular, epitaxial growth techniques can form high-quality single-crystal films. Other methods that can be used include liquid phase epitaxial growth, solid phase epitaxial growth, and sol-gel techniques. Furthermore, an adhesive with a small refractive index difference between the fluorescent film 11a and the lenticular substrate 11b can also be used.

[0055] From the viewpoint of minimizing the difference in refractive index, it is most desirable for fluorescent film 11a to be made of a base material having the same composition as the constituent material of lenticular substrate 11b to which a predetermined proportion of an additive as an activator has been added, and it is next desirable for fluorescent film 11a to have the same crystalline structure as lenticular substrate 11b. For example, in the former case, lenticular substrate 11b may be formed of undoped lutetium aluminum garnet (LuAG), and fluorescent film 11a may be formed of praseodymium-doped lutetium aluminum garnet (Pr:LuAG) or cerium-doped lutetium aluminum garnet (Ce:LuAG). In the latter case, fluorescent film 11a may be formed of Ce:GAGG having a garnet crystalline structure, and lenticular substrate 11b may be formed of undoped lutetium aluminum garnet (LuAG) having a garnet crystalline structure. Furthermore, fluorescent film 11a may be made of a single crystal and lenticular substrate 11b of a polycrystalline material, or vice versa.

[0056] Although not shown in FIG. 1, as shown in FIG. 2, in SIL 11, an anti-reflection film 11c is arranged on the lens-shaped convex surface portion on the fluorescent light emission surface side, and a reflective film 11d is arranged on the flat surface portion on the incident surface side.

[0057] The anti-reflection film 11c may be a multi-layer film formed of, for example, two or more dielectric films.

[0058] The reflective film 11d is configured from a dielectric multilayer film or a metal reflective film, similar to the anti-reflection film 11c. A chrome coating or the like may be further disposed on the dielectric multilayer film to enhance light blocking properties. The reflective film 11d reflects the fluorescence F emitted in the direction away from the image sensor 15 toward the image sensor 15, effectively improving the transmission efficiency of the fluorescence F by nearly two times.

[0059] As described above, by using the SIL 11 in which the fluorescent film 11a as a scintillator and the lens-shaped body 11b that is the first lens of the post-objective lens group 12 are integrally formed so that the refractive index is approximately uniform, an imaging optical system 17 having a sample plane on the fluorescent film 11a and a high numerical aperture exceeding NA=1.0 can be obtained, thereby improving the spatial resolution of the radiation camera 10. On the other hand, setting the numerical aperture high reduces the depth of field, so a thinner fluorescent film 11a is required.

[0060] The depth of field d includes the front and back, tot =λ n / NA 2 where λ is the wavelength of the fluorescence F, n is the refractive index of the fluorescent film 11a for the wavelength λ, and NA is the numerical aperture of the objective lens 16. For example, if the refractive index for the fluorescent wavelength of 520 nm of Ce:GAGG is 1.85 and the numerical aperture of the objective lens 16 is 1.76, the depth of field is 311 nm. It is technically difficult to process the fluorescent film 11a to a thickness of less than 1 micron while suppressing optical degradation such as distortion. [Focused and unfocused fluorescence]

[0061] That is, when the spatial resolution is improved using an imaging optical system 17 with a large numerical aperture NA, if it is not possible to form fluorescent film 11a with a thickness approximately equal to the depth of field, fluorescent film 11a will inevitably be thicker than the depth of field of imaging optical system 17. In this case, some of the X-rays R are absorbed within the depth of field, and some of the X-rays R penetrate to and are absorbed at positions beyond the depth of field, generating fluorescence F starting from the positions where the X-rays R were absorbed. That is, the fluorescence F imaged on image sensor 15 is a mixture of fluorescence F generated within the depth of field, which is the in-focus position (hereinafter also referred to as in-focus fluorescence F1), and fluorescence F generated outside the depth of field, which is the out-of-focus position (hereinafter also referred to as out-of-focus fluorescence F2).

[0062] 3(a) and 3(b) are diagrams illustrating focused fluorescence F1 and out-of-focus fluorescence F2 in this embodiment. Fig. 3(a) shows focused, focused fluorescence F1 on the image sensor 15 of visible light emitted near the left side of the fluorescent screen 11a, which is determined to be the focal position of the objective rear lens group 12, while Fig. 3(b) shows blurred, out-of-focus fluorescence F2 on the image sensor 15 of visible light emitted near the right side of the fluorescent screen 11a, which is outside the focal depth of the objective lens 16.

[0063] As shown in the right diagram of Figure 3(a), when fluorescence F forms a focused spot on the image sensor 15, it becomes focused fluorescence F1, which has an illumination distribution shaped like an Airy disk or a spot close to it. Its diameter D is calculated by multiplying twice the Rayleigh spatial resolution δ by the optical magnification M of the imaging optical system 17, or "D = 1.22 × λM / NA." On the other hand, as shown in the right diagram of Figure 3(b), when fluorescence F does not form a focused spot on the image sensor 15, it becomes unfocused fluorescence F2, which has an illumination distribution shaped like a spot larger than an Airy disk, as indicated by the dotted circle f. An image containing an electrical signal obtained by photoelectric conversion of this unfocused fluorescence F2 will have poor contrast quality and reduced spatial resolution.

[0064] That is, even if an optical system with a maximum numerical aperture (NA) is constructed using a solid immersion lens (SIL), if the thickness of the fluorescent film 11a cannot be thinned to the depth of field, out-of-focus fluorescence F2 will be mixed in as a component of the image, reducing spatial resolution. As a result, even in an optical system with a numerical aperture (NA) maximized to the limits of manufacturing precision, it is extremely difficult to achieve diffraction-limited spatial resolution. To solve this problem, this embodiment provides a method for removing the signal from the out-of-focus fluorescence F2 and extracting only the signal from the focused fluorescence F1 in a configuration in which the fluorescent film 11a is thicker than the depth of field of the imaging optical system 17 and the fluorescence F imaged on the image sensor 15 includes both focused fluorescence F1 and out-of-focus fluorescence F2. By applying this method, it is possible to avoid a reduction in spatial resolution. [Total X-ray to signal conversion efficiency, a prerequisite for focused fluorescence discrimination]

[0065] This method utilizes a detection process in which X-ray photons are absorbed by the scintillator and converted into visible light in a radiation camera 10 that uses a scintillator, and then form a point image of a predetermined size as shown in Figures 3(a) and 3(b) on the right due to the diffraction phenomenon that occurs during transmission through the lens system. The point image is observed in units of X-ray photons, and only the signal corresponding to the focused fluorescence F1 is extracted to reconstruct an image.

[0066] To accurately observe a point image in units of X-ray photons, two components are required: a highly sensitive radiation camera 10 that can detect single X-ray photons (X-ray photon counting), and a control unit 20 that divides the signals of the X-ray photon group that makes up the specimen image into X-ray photon units.

[0067] The sensitivity of the measurement system as a camera is determined by the conversion efficiency η SE and the luminous efficiency η, which is determined by the photon energy E of the X-rays from the light source. LY (E)=η SE ×E, the light recovery efficiency η of the optical system, which depends on the numerical aperture NA CE (NA), transmittance of the optical system η TO , the quantum efficiency η of the image sensor 15 for fluorescence F VQE The overall X-ray to signal conversion efficiency η TOTAL (E,NA)=η LY (E)×η CE (NA)×η TO ×η VQE That is, when one X-ray photon with photon energy E is absorbed by the scintillator, η LY (E) visible fluorescence F is emitted in random directions, and η CE Fluorescence F is incident on the imaging optical system 17 at a transmittance of η TO , quantum efficiency η VQE On the image sensor 15, the average η TOTAL It is detected as (E,NA) electrons.

[0068] The light recovery efficiency of an optical system that depends on the numerical aperture NA is [η CE =0.5×(1+r)×(1-√(1-(NA / n s ) 2)) where r is the reflectance of the reflective film 11d on the X-ray incident surface of the scintillator, NA is the numerical aperture of the objective lens 16, and n s is the refractive index of the scintillator.

[0069] For example, in the conventional configuration disclosed in JP 2016-45183 A, the refractive index n s When 10 keV X-rays are detected using a radiation camera consisting of a flat Ce:LuAG scintillator with a luminous efficiency of [160 visible photons / X-ray photon] for 10 keV X-rays and a reflecting film with a reflectance of r=0.95 coated on the X-ray incident surface of the scintillator, an imaging optical system with a dry objective lens with a numerical aperture NA=0.95 and a transmittance of 0.9 for the wavelength of fluorescence F, and an image sensor with a quantum efficiency of 0.9 for the wavelength of fluorescence F, [η LY (10 keV) = 160, η CE (0.95NA)=0.14, η TO =0.9, η VQE =0.9], and the overall X-ray signal conversion efficiency of the entire radiographic camera is [η TOTAL (10 keV, 0.95 NA) = ~18 electrons / X-ray photon]. In other words, the fluorescence F generated when one X-ray photon is absorbed by the fluorescent screen of a conventional radiation camera is detected as an average of 18 electrons on the image sensor.

[0070] To observe the irradiation distribution of fluorescence F, it is necessary to measure signals at multiple pixels that are smaller than the irradiation distribution of fluorescence F. In an X-ray camera with a dry objective lens configuration such as the one described above, for example, if one X-ray photon is incident on the radiation camera and a total of 18 electrons are distributed across nine pixels on the image sensor 15, the signal will be divided into nine and detected according to the irradiation distribution. In other words, there will be many pixels whose signals are equal to or lower than the noise value (1 to 2 electrons) that the image sensor has for each pixel. This not only results in an extremely insufficient number of statistics for observing the irradiation distribution, but also makes it difficult to distinguish between the signal of a single X-ray photon and the noise signal (X-ray photon counting). In other words, it is not possible to classify the fluorescence F generated by a single X-ray photon into focused fluorescence F1 and unfocused fluorescence F2.

[0071] On the other hand, in the embodiment of the present invention shown in FIG. 1, the refractive index n s The imaging optical system 17 has a numerical aperture increased to NA=1.85×0.95=1.76 by using a scintillator made of a Ce:GAGG phosphor film 11a having a luminous efficiency of [600 visible photons / X-ray photon] for 10 keV X-rays R and a hemispherical or hyper-hemispherical convex lens shape on the exit surface side, and is integrally formed on the entrance surface side of the lens-shaped substrate 11b. At this time, various parameters are set to [η LY (10 keV) = 600, η CE (1.76NA)=0.69, η TO =0.9, η VQE =0.9], and the overall X-ray signal conversion efficiency of the radiation camera 10 is [η TOTAL (10 keV, 1,76 NA) = 335 electrons / X-ray photon. In other words, the fluorescence F generated when one photon of an X-ray is absorbed by the fluorescent screen 11a of the radiation camera 10 is detected as an average of 335 electrons on the image sensor 15. [Discrimination between focused and unfocused fluorescent point images]

[0072] Thus, the radiation camera 10 of this embodiment, which uses the scintillator-integrated SIL 11 described above, has a light recovery efficiency η of about 5 times that of the conventional one. CE and 4 times the luminous efficiency of conventional products LY As a result, the overall X-ray to signal conversion efficiency η TOTALThe signal density is improved to 335 electrons / X-ray photon, nearly 20 times that of the conventional configuration. Even though 335 electrons are detected across nine pixels, the signal value per pixel easily exceeds the noise value (1–2 electrons) of the image sensor 15. The increased signal value significantly reduces the effect of the optical shot noise of fluorescence F, allowing the fluorescence F irradiance distribution for each X-ray photon to be observed with sufficient statistical accuracy. Therefore, if the fluorescence F generated per X-ray photon is imaged on the image sensor 15 with an irradiance distribution spot shape similar to or equal to the Airy disk f, it can be distinguished as in-focus fluorescence F1. If the irradiance distribution spot shape is larger than the Airy disk f, it can be distinguished as out-of-focus fluorescence F2. Here, the Airy disk f indicates the smallest image spot at the focal position, which is theoretically determined by the diffraction phenomenon of light. Its irradiance distribution is determined depending on the wavelength λ of fluorescence F, the numerical aperture NA of the objective lens 16, and the optical magnification M of the imaging optical system 17. Therefore, the Airy disk f is an appropriate standard for the allowable illumination distribution to distinguish in-focus images. Under ideal conditions, i.e., in a perfect optical system without aberrations, the point image formed would be the Airy disk f if it were generated from fluorescence originating at the center of the depth of field. However, because optical systems actually have a certain degree of aberration, the point image of in-focus fluorescence F1 detected within the depth of field will also be somewhat larger than the Airy disk f. Therefore, the reference value multiplied by the tolerance factor is used to classify the image into in-focus fluorescence F1 and out-of-focus fluorescence F2. For example, if the tolerance factor is set to 1.5, a measured point image with a diameter less than 1.5 times the Airy disk f is classified as in-focus fluorescence F1, and one with a diameter greater than 1.5 times the Airy disk f is classified as out-of-focus fluorescence F2. The tolerance factor can be adjusted arbitrarily depending on the target spatial resolution, target detection efficiency, and the magnitude of the optical system's aberrations. Although the discrimination criterion used here is the diameter of the Airy disk f, this does not limit the reference target. Other features such as the aspect ratio of the diameter or the similarity to the shape of the Airy disk f may also be used, or multiple criteria may be used simultaneously.

[0073] However, if the objective lens 16 of the X-ray camera is simply made to have a solid immersion structure with an integrated scintillator, the light recovery efficiency η CE is only improved by about 5 times, so the overall X-ray to signal conversion efficiency [η TOTAL89 electrons / X-ray photons], and in the dry objective lens configuration in which the material of the phosphor film 11a as the scintillator is a Ce:GAGG crystal, the luminous efficiency η LY is only improved by about 4 times, so the overall X-ray to signal conversion efficiency [η TOTAL The resulting noise level is 68 electrons / X-ray photon, which reduces the accuracy of discriminating between focused and unfocused fluorescence F1 and F2. Therefore, to implement the concept of constructing an observation image by discriminating only focused images from unfocused images, it is desirable to design the camera's front optical system, including the selection of optical materials, so that at least 100 electrons are generated on the image sensor 15 per X-ray photon. In this regard, as described below, the most preferable configuration is one in which a solid immersion objective lens is placed at the front of the X-ray camera, with a Ce-doped GAGG fluorescent film 11a integrally formed with a lenticular substrate 11b made of GAGG. However, in the focused fluorescence discrimination method of this embodiment, as long as the number of electrons detected by photoelectrically converting the fluorescence incident on the light-receiving surface of the image sensor 15 per X-ray photon is 50 or more, using an image sensor 15 with low noise (less than 1 electron) can achieve the ability to discriminate between focused fluorescence F1 and unfocused fluorescence F2, thereby improving resolution. The number of photoelectrically converted electrons corresponds to the level of the electrical signal obtained from each pixel.

[0074] That is, the above is an example in which an X-ray with a photon energy of 10 keV is used as the light source, but when an X-ray with a higher photon energy E is used, η LY (E) increases in proportion to the photon energy E, so η CE (NA), η TO , η VQE That is, if the photon energy is high, a configuration in which the NA is relaxed can be adopted by adopting a flat substrate and a dry objective lens optical system as shown in FIG. 2(d) instead of a solid immersion objective lens optical system with an integrated scintillator. Alternatively, the conversion efficiency η SE Conversely, if the photon energy E is low, η LY (E) becomes smaller, so η CE (NA), η TO , η VQEand the scintillator conversion efficiency η SE The demand for high-NA imaging optics17 and high-conversion-efficiency scintillators is increasing. Regardless of the photon energy E, the overall X-ray-to-signal conversion efficiency [η TOTAL =η LY (E)×η CE (NA)×η TO ×η VQE >50] is an index for efficiently obtaining the effect of this method. In other words, it is important to maintain the total conversion efficiency of 50 or more, from the time when visible fluorescence is generated by one X-ray photon that reaches the scintillator until it is received by the image sensor 15 and converted into electrons. To classify focused fluorescence F1 and unfocused fluorescence F2 with even higher accuracy, η TOTAL It is desirable that (E,NA)>300.

[0075] As described above, by utilizing a solid immersion objective lens optical system using SIL11 and a scintillator with high conversion efficiency, such as Ce:GAGG scintillator, we have achieved an X-ray to signal overall conversion efficiency of η TOTAL By sufficiently increasing the signal intensity, it becomes possible to discriminate between the in-focus fluorescence F1 and the out-of-focus fluorescence F2 on an X-ray photon basis, or the accuracy of this discrimination can be improved. As a result, even when the fluorescent film 11a is thicker than the depth of field of the imaging optical system 17, a reduction in spatial resolution can be avoided by reconstructing a specimen image using only the signal of the in-focus fluorescence F1. Furthermore, even when the entire SIL 11 uses a lenticular fluorescent substrate 11e having a scintillator function, as shown in FIG. 2(c), an image can be constructed by isolating only the in-focus fluorescence F1, thereby avoiding a reduction in spatial resolution.

[0076] By constructing an image using only the signal from the focused fluorescence F1, when forming the fluorescent film 11a as shown in Figures 2(a) and 2(b), there is no need to strictly control the thickness of the fluorescent film 11a, which makes it possible to suppress deterioration in optical quality, such as distortion, that occurs during thin-film processing. This allows for high-precision optical design of the imaging optical system 17, including the SIL 11, and ultimately improves spatial resolution. When using a lenticular fluorescent substrate 11e in which the entire SIL 11 is made of a fluorescent material, as shown in Figure 2(c), there is no need for a thin-film formation process, which further suppresses deterioration in optical quality. At the same time, the reduction in processing steps also improves manufacturing costs.

[0077] Furthermore, since this method integrates the signal of the focused fluorescence F1 in units of X-ray photons, i.e., it applies photon counting, signals other than those originating from X-ray photons, such as noise from the image sensor 15 and flare and ghosts generated by scattered light inside the lens that do not contribute to imaging, are filtered out during the counting process, allowing for the acquisition of images with higher contrast.

[0078] The in-focus signal discriminator 22a of the image signal processor 22, which is an image processing system that divides the signals of X-ray photons that make up the specimen image into individual photons, performs discrimination processing in which it generates an image using only the electrical signal (hereinafter also referred to as the in-focus signal) obtained by photoelectrically converting the in-focus fluorescence F1, while not using the electrical signal (hereinafter also referred to as the out-of-focus signal) obtained by photoelectrically converting the out-of-focus fluorescence F2. Details of the process of dividing the signals of the X-ray photon group into X-ray photon units to discriminate between the in-focus fluorescence F1 and the out-of-focus fluorescence F2, and details of the process of generating an image using only the in-focus signal, will be explained in detail in the following paragraphs and the flowchart in Figure 7.

[0079] As described above, by generating an image using only the focusing signal, the contrast quality of the generated image can be improved, and even when using an X-ray camera having a scintillator with a fluorescent film 11a thicker than the depth of field, the spatial resolution can reach the diffraction limit. Furthermore, in this embodiment, in addition to generating an image using only the focusing signal, imaging is performed using a photon localization method, which is one of the super-resolution techniques, thereby achieving a resolution that exceeds the diffraction limit, which is the theoretical spatial resolution. Imaging using the photon localization method will be described below. [Photon localization method]

[0080] Figures 4(a) to 4(d) are diagrams illustrating an ultra-high-definition imaging method using photon localization as a function of the point image signal detector 22c in the system configuration of Figure 1. In an X-ray camera using a scintillator, when the X-ray photons constituting a specimen image are observed in units of X-ray photons, if they are absorbed by the scintillator at the focal position, they are converted into visible light by the scintillator, and then become a blurred bright spot (point image) in the shape of an Airy disk f due to diffraction occurring during transmission through the lens system, which is then measured by the image sensor 15. In other words, the image received by the image sensor 15 is constructed by integrating this point image, which is the smallest signal unit generated by one X-ray photon, according to the incident position of the X-ray photon and the number of X-ray photons. As mentioned in paragraph 0017, the spatial resolution δ is defined as the radius of the point image of the Airy disk f, [δ = 0.61 × λ / NA].

[0081] The photon localization method is a technique for identifying the incident position of an X-ray photon with high precision by detecting the center position of the measured X-ray photon point image. The precision of the localization depends on the number N of photons of fluorescence F that compose the point image, and is expressed as [δ] using the radius δ of the point image of the Airy disk f. SUPER =δ / √N]. In other words, the greater the number N of photons of fluorescence F detected on the image sensor 15, the more accurate the localization. This formula is an approximation when the noise value of the image sensor 15 is extremely small and the effect of pixel size is ignored.

[0082] When localization processing is applied, the radius size of the point image, which is the minimum signal unit, is substantially reduced from the radius δ of the point image of the Airy disk f to the detection error δ of the center position. SUPER After reducing the size of the point image, if the signals of the detected X-ray photons are integrated again to form an image, the spatial resolution will be reduced to δ SUPER This is an improvement of [1 / √N] times compared to the original image before processing.

[0083] To implement this method, it is necessary to measure and localize each X-ray photon individually as a point image. In a typical imaging process, in which the signals of incoming X-ray photons are collectively exposed in a single frame, the resulting image is an accumulated image of point images, making it impossible to measure and evaluate each point image individually. Therefore, by dividing a frame into multiple subframes (frame-by-frame images) with short exposure times and capturing them continuously, the signal density of the X-ray photon group per subframe can be reduced. If the density can be reduced to the point where X-ray photons can be observed discretely within a subframe, as shown in Figure 4(a), the illumination distribution of fluorescence F can be measured for each point image, and localization can be applied as shown in Figure 4(c).

[0084] If the X-ray photon signals cannot be sufficiently discretized within a subframe even when the exposure time of the image sensor 15 is set to the minimum, the X-ray output from the radiation source 200 may be reduced or a radiation attenuation plate 400 may be inserted between the radiation source 200 and the sample 900 to attenuate the X-ray intensity, thereby further reducing the density of the X-ray photon group reaching the sample 900.

[0085] The signal density of the fluorescent spot corresponding to the X-ray photon group detected within the index subframe is explained in Figures 5(a) to 5(c). For simplicity, the focused fluorescent light (focused signal) F1 is arranged at equal intervals in the figure, but in reality, the image includes out-of-focus fluorescent light (out-of-focus signal) F2, which is larger than the focused fluorescent light F1, and the detection position of each is random. In addition, the effective pixel size A on the detection surface is set to 1 / 2 to 1 / 3 of the radius of the focused fluorescent light F1 (Airy disk radius f). Figure 5(a) shows the detection density I, where an average of 16 X-ray photons are detected per 400 pixels. DET(=16 photons / 400 pixels = 0.04 photons / pixel), and Fig. 5(b) shows the detection density I where an average of 9 X-ray photons are detected per 400 pixels. DET (=9 photons / 400 pixels = 0.023 photons / pixel), and Fig. 5(c) shows the detection density I where an average of 4 X-ray photons are detected per 400 pixels. DET (=4 photons / 400 pixels=0.01 photons / pixel) is shown below.

[0086] Under the conditions shown in Figures 5(a) and (b), the density of the light-receiving pixels for the point images is high. Therefore, in actual measurements where point images are detected at random positions, spatial overlap occurs frequently, making it difficult to evaluate each point image individually. In contrast, in Figure 5(c), point images are frequently detected separately, making it possible to evaluate each point image individually. Therefore, the detection density index at which X-ray photons can be discretely observed within a subframe is 0.01 photons / pixel, and it is recommended to set a value close to or smaller than this. However, even at around 0.02 photons / pixel, overlapping point image signals can be detected from the spread of the light-receiving pixels. Therefore, by performing processing such as excluding such point images from the data, localization processing can be performed.

[0087] The minimum exposure time of the image sensor 15 is limited by the characteristics of the sensor device itself, the limitations of the signal circuit, and the decay time τ, which is the time it takes for the scintillator's fluorescence F to emit and then extinguish. For example, if the subframe exposure time is not sufficiently longer than this value, the signal of a single X-ray photon will arrive at the image sensor 15 at the boundary between subframes, increasing the frequency of the signal splitting into two consecutive subframes. To sufficiently reduce the frequency of splitting, it is desirable to set the exposure time to approximately 10 times the decay time τ. Although it is impossible to completely eliminate the frequency of signal splitting, if the number of detected fluorescence F photons N is less than a predetermined value and there are point image signals at the same position in two consecutive subframes, data quality can be maintained by removing these signals or by combining them to restore a single X-ray photon signal.

[0088] For example, when Ce:GAGG is used for the fluorescent film 11a, the fluorescent afterimage time or decay time τ of Ce:GAGG is about 100 ns, so it is recommended that the exposure time of the image sensor 15 be 10 times longer, 1000 ns or more, i.e., the maximum frame rate be 1 MHz or less.

[0089] An example of applying this embodiment to a synchrotron radiation facility called SPring-8 owned by the applicant will be shown below. For example, when a quasi-monochromatic X-ray with a photon energy of 10 keV is used as the radiation source 200, a maximum of [I PH =1×10 14 photons / second / mm 2 X-rays R are incident with an intensity of . In a configuration in which GAGG is used as the solid immersion lens optical system and solid immersion lens material, the numerical aperture is set to NA=1.76 and the spatial resolution is set to δ=180 nm. The effective pixel size A on the detection surface is set to approximately 1 / 2 to 1 / 3 of the spatial resolution δ, which is set to 60 nm in this case. In addition, the thickness of Ce:GAGG at which focused fluorescence F1 is obtained is 311 nm, and the probability that X-rays R are absorbed by the scintillator (X-ray quantum efficiency) η XQE is η when the X-ray energy is 10 keV XQE When the image sensor 15 performs exposure at a frame rate of 1 MHz, the detection density I of X-ray photons in a subframe is DET is [I DET =I PH ×η XQE ×T = 1×10 14 (photons / sec / mm 2 )×3.2(%)×1000(ns)=3.2×10 6 (photons / mm 2 )=1.15×10 ―2 In other words, it is possible to obtain a value close to one X-ray photon per 100 pixels (=0.01 photon / pixel), which is an index of discretization of a visible point image.

[0090] As mentioned above, once the light source intensity is determined, the number of pixels per 100, which is the index for discretizing the visible point image, is 1 X-ray photon or less (I DET ≦0.01 photons / pixel), the minimum frame rate is determined. For example, in the above embodiment,10 (photons / sec / mm 2 When X-rays R are incident on the detection surface at an intensity of 100 Hz or more, a frame rate of the image sensor 15 of about 100 Hz or more is suitable.

[0091] Figure 4(a) shows a group of subframes acquired by irradiating a sample 900 with X-rays R and capturing images continuously at a predetermined time interval. By shortening the exposure time and creating subframes that are sufficiently fragmented in time, the X-ray signals in the image can be observed discretely in units of X-ray photons. Note that Figure 4(a) uses an image generated using only the focused signal shown in Figure 3(a). Figure 4(b) shows image data (shown as an example by a blurred "R") obtained by integrating the group of subframes shown in Figure 4(a).

[0092] As shown in Fig. 4(a), even when only the focusing signal is selected, the focusing signal is somewhat blurred because it is a signal within a predetermined size range in Fig. 3(a), that is, a signal having a spot shape of the Airy disk f or a similar shape, and accumulating these signals results in somewhat blurred image data such as that shown in Fig. 4(b). Note that each point (circle) in the frame in Fig. 4(a) represents an image of fluorescence F generated in response to one photon of X-rays R emitted from radiation source 200 and absorbed by fluorescent film 11a.

[0093] Figure 4(c) shows the subframes after localization processing of the subframe group shown in Figure 4(a). Specifically, the center position of the fluorescence F generated corresponding to one X-ray photon in Figure 4(a) is calculated by image processing, and the fluorescence is represented by a single point (point image center) at that center position. That is, the signal of the focused fluorescence spot detected across multiple pixels exhibits a distribution close to a Gaussian distribution in the Airy disk f, so the center position coordinates (x, y) of the spot can be determined by comparing the signal levels of each pixel. As shown in Figure 4(c), the subframe image after localization processing is represented by a single point at the point image center per X-ray photon, resulting in data with highly accurate X-ray incident position. Therefore, as shown in Figure 4(d), integrating the subframe group of Figure 4(c) produces a clear image (shown as a clear "R" in the figure as an example).

[0094] As described above, by selecting only the in-focus signal (see FIG. 3) and then performing imaging using the photon localization method (see FIG. 4), the spatial resolution of the radiation imaging device 100 can be increased to a value exceeding the diffraction limit. In this embodiment, the radiation camera 10, which uses Ce:GAGG as the material for the fluorescent film 11a of the SIL 11, has a spatial resolution of δ=180 nm, an average photon number N of the fluorescence F, and an overall X-ray-to-signal conversion efficiency η TOTAL coincides with η TOTAL = 335 electrons / X-ray photons, so by applying the photon localization method, the super-resolution spatial resolution [δ SUPER =δ / √η TOTAL = 180 nm / √335 = 9.8 nm (resolving 4.9 nm line & space). Therefore, compared to the theoretical limit of 96 nm (resolving 48 nm line & space) of conventional technology, spatial resolution can be significantly improved.

[0095] Note that the position of the fluorescence F imaged on the image sensor 15 may fluctuate due to movement of the sample 900 caused by excessively strong intensity of the X-rays R, or due to shaking of the equipment for fixing and adjusting the position of the sample 900 or the radiation camera 10. In this embodiment, drift correction is performed to correct for this fluctuation in the position of the fluorescence F. In this embodiment, the drift correction involves determining the x-coordinate and y-coordinate of the point image center of each fluorescence F at time t, and correcting the predetermined values ​​Δx and Δy. Note that whether or not to perform drift correction is optional, and drift correction does not necessarily have to be performed.

[0096] FIG. 6 is a diagram illustrating the flow of image signal processing in this embodiment, along with the system configuration. When data for a subframe group read from the image sensor 15 of the radiation camera 10 is input to the image signal processing unit 22, it is temporarily stored in a frame memory corresponding to the image data storage unit 22b (S11). Subsequently, all point image signals within that subframe group are classified as in-focus or out-of-focus signals (S12). Subsequently, the center positions of point images within that subframe group that are deemed to be in-focus signals are detected (S13), and detection data is generated for each point image, resulting in a detection data group (S14). The detection data consists of the point image center position (x, y), the number of photons N(E) of the fluorescence F forming the point image, the standard deviation (σ) of the point image signal distribution, and the time (t) at which the point image was detected. Note that if all pixel signals from the image sensor 15 are measured at the same time, the time t may be used as a number identifying the subframe.

[0097] The amount of spatial drift at time t when each point image is detected is detected (S15), and drift correction of the point image center position is performed based on the detected amount of drift (S16), resulting in a corrected detection data group shown in S17. The corrected detection data group obtained in S17 is stored in memory unit 42 (S20). Meanwhile, an output frame of the reconstructed image is generated (S18), and the detection data group obtained in S17 is integrated and plotted in the output frame to obtain a high-resolution frame that is essentially an integration of subframes that have been subjected to photon localization processing (S19). The high-resolution frame is stored in memory unit 42 (S21).

[0098] In addition, x in Figure 6 i indicates the x coordinate of the center position of the point image of X-ray photon i, and y i indicates the y coordinate of the center position of the point image of X-ray photon i, and Δx i indicates the drift of X-ray photon i in the x direction, and Δy i indicates the drift of X-ray photon i in the y direction, and X i indicates the x-coordinate of the center position of the point image of X-ray photon i after drift correction, and Y i indicates the y-coordinate of the center position of the point image of X-ray photon i after drift correction, and E i indicates the signal value of the X-ray photon i (a value equivalent to the number N of photons of the fluorescence F detected by the image sensor 15), and σ i indicates the standard deviation of the point image signal distribution of X-ray photon i (the standard deviation of the irradiation distribution of fluorescence F detected by the image sensor 15), and t i indicates the time of detection (or a number specifying the subframe). The subscript n in S14 and S17 indicates that n focus signals are detected from the subframe group.

[0099] In step S19, when plotting the detection data obtained in S17, low-quality detection data may be excluded. Low-quality data refers to, for example, data with a large standard deviation σ or a small signal value. The threshold is determined based on the numerical aperture and magnification of the optical system, the target spatial resolution, the theoretical visible point image (Airy disk f) shape corresponding to the wavelength of the scintillator's fluorescence F, and the overall X-ray-to-signal conversion efficiency η TOTAL It is set to a predetermined value taking into consideration the above.

[0100] FIG. 7 is a flowchart showing the imaging operation of the radiation imaging apparatus 100 according to this embodiment. When the imaging operation of the radiation imaging apparatus 100 begins, in step S01, the control unit 20 sets the photon energy, output value, energy spectrum, and other parameters of the X-rays R from the radiation source 200 used in measurement. Next, the focus of the imaging optical system 17 is adjusted so that an image of the vicinity of the surface of the fluorescent film 11a onto which the X-rays R are incident is formed on the image sensor 15. Furthermore, imaging conditions such as an exposure time and intensity of the X-rays R suitable for spatially distinguishing each X-ray photon are set. Specifically, the exposure time control unit 21a shortens the exposure time of the image sensor 15 or reduces the intensity of the X-rays R to lower the signal density of the X-rays R per subframe, thereby setting imaging conditions that allow spatially distinguishing each X-ray photon, as shown in FIG. 5(c).

[0101] Here, when using high-energy X-rays R with strong penetrating power, when configuring the front optical system using an SIL 11 made entirely of phosphor material as shown in Figure 2(c), or when configuring the radiation camera 10 using both, the proportion of detected out-of-focus signals can become extremely high. In this case, even if the signal density of X-rays R per frame is reduced, if the diameter of the blur spot created by the out-of-focus signals is large, the overlap of multiple out-of-focus signals can increase the background noise, and the in-focus signal point image can become buried in the noise and become indistinguishable. In this case, it can be addressed by further shortening the exposure time or reducing the intensity of X-rays R so that only one X-ray photon is detected per subframe, and separating the in-focus and out-of-focus signals into separate subframes.

[0102] Next, under the imaging conditions established in step S01, the expected spatial resolution δ SUPER A frame for output of the reconstructed image is generated according to the image size and field of view.

[0103] Subsequently, in step S02, the camera control unit 21 continuously captures images of the sample 900 under the set imaging conditions, and acquires a group of subframes of the sample 900.

[0104] Next, in step S03, the image signal processing unit 22 processes all the bright spots in the subframe group obtained in step S02 to obtain a profile for each bright spot. Next, in step S04, the image signal processing unit 22 evaluates the profile for each bright spot to determine whether the signal is an in-focus signal or an out-of-focus signal. Here, an in-focus signal is detected as a signal that falls within a predetermined range on the image sensor 15, i.e., as an airy disk f or a spot shape similar thereto, as shown in FIG. 3(a). Meanwhile, an out-of-focus signal is detected as a signal that exceeds the predetermined range on the image sensor 15, i.e., as a spot shape larger than the airy disk f, as shown in FIG. 3(b). Therefore, by detecting and evaluating the distribution of the signal, it is possible to determine whether the signal is an in-focus signal or an out-of-focus signal.

[0105] Next, in step S05, the image signal processing unit 22 detects the point image centers of the bright points classified as in-focus signals in step S04 by image processing. In step S06, the image signal processing unit 22 corrects the drift of the point image center positions based on the detected drift amount. In step S07, the image signal processing unit 22 plots each detected point image center in an output frame to construct a high-resolution frame in which the entire image is reconstructed.

[0106] As described above, according to the radiation imaging apparatus 100 of this embodiment, an image is constructed using only the focusing signal, and further, data processing is performed using the photon localization method, thereby improving the contrast quality of the image and increasing the spatial resolution of the radiation imaging apparatus 100 by δ SUPER =9.8nm (=resolving 4.9nm line & space).

[0107] The number of subframes required is the detection density I DET , the effective pixel size A on the detection surface and the pixel size A of the final output high-resolution frame SUPERand the dynamic range D (maximum number of X-ray photons in one pixel) of the final output high-resolution frame, and N FRAME =D / (I DET ×(A SUPER / A) 2 )

[0108] When performing super-resolution processing such as photon localization, the detection density value that serves as the index is I DET = 0.01 photons / pixel, the spatial resolution is often improved to about 10 times the original image, and the maximum gradation that the human eye can see is about 8 bits (= 256). FRAME =256 / (0.01×(1 / 10) 2 )=2.56×10 6 In reality, the number of subframes is less than this value, for example, D=6 bits and N FRAME =6.4×10 5 Even with only a few sheets of paper, the photon localization processing can be applied to obtain a high-resolution image with improved spatial resolution.

[0109] Using the parameters of this example, I DET =0.01 photons / pixel, A=60nm, pixel size A SUPER is the spatial resolution δ SUPER About 1 / 3 of A SUPER =δ SUPER / 3=9.8nm / 3=3.3nm, Dynamic Range D=8bit=2 8 If we set the number of photons to 256, the number of subframes required is N FRAME =256 / (0.01×(3.3 / 60) 2 )=8.5×10 6 In the case of a configuration that can acquire subframes at a rate of 1 MHz, the total imaging time is T TOTAL =8.5×10 6 / 1MHz=8.5 seconds. [Generation of energy-resolved images]

[0110] The above is an example in which all X-ray photons have monochromatic or quasi-monochromatic energy, as in a synchrotron radiation facility. On the other hand, Figures 8(a) to 8(c) are diagrams for explaining image processing by the X-ray energy-resolved image generation unit 22d (Figure 1) when an X-ray source with mixed energies is used. The focused fluorescence F1 acquired by the continuous imaging in step S02 of Figure 7 in the above embodiment and discriminated and extracted in step S04 has different incident X-ray photon energies, and is therefore detected on the image sensor 15 as a point image composed of different photon numbers N (Figure 8(a)). The sum of the pixels constituting the point image (detection data E) is proportional to the photon energy, so the magnitude of the X-ray energy can be measured for each focused fluorescence F1. The detection error is a Poisson error that depends on the photon number N of the fluorescence F forming the point image and is expressed as 1 / √N.

[0111] For example, if fluorescence F is detected on the image sensor 15 with a photon number N=335, it has a detection error of 1 / √335=5.4%. This means that 10 keV X-rays can be measured with an energy resolution of 0.54 keV. In this way, the larger the photon number N of the detected fluorescence F, the smaller the detection error and therefore the improved energy resolution. In other words, if a solid immersion lens optical system or a scintillator with high conversion efficiency such as Ce:GAGG is used, the overall X-ray-to-signal conversion efficiency η TOTAL Increasing the energy resolution improves.

[0112] Next, multiple X-ray energy ranges are set, and the focused fluorescence F1 is classified into each range. Figure 8(b) shows an example in which three X-ray energy ranges are set, and each subframe is divided and integrated for each energy range. In actual data processing, for each point image detection data, an output frame is selected and plotted according to the number of photons N of the fluorescence F that forms the point image. This makes it possible to generate an image for each energy range (Figure 8(c)). Note that in Figures 8(b) and (c), dark hatching (left) indicates fluorescence F with a large number of photons N, medium-thickness hatching (center) indicates fluorescence F with an intermediate photon number N, and light hatching (right) indicates fluorescence F with a small number of photons N.

[0113] As described above, this embodiment can classify each X-ray photon into multiple X-ray energy ranges and generate an image for each X-ray energy range. In other words, this image processing acts as an energy filter. This makes it possible to observe the energy dependence of the X-ray image with a single imaging session, allowing the internal structure of the sample 900 as well as the constituent materials to be identified. This method allows the energy to be selected for use as data by the detector, making X-ray imaging using quasi-monochromatic energy possible even with industrial X-ray generators that have a wide range of energies.

[0114] It is also possible to apply photon localization processing to the point images in each subframe for each energy range shown in Figure 8(b) after energy resolution, thereby realizing a radiation imaging device 100 that simultaneously has high energy resolution and high spatial resolution. [Other variations]

[0115] 6 and 7 can be performed not only by a CPU in a computer but also by hardware calculations using integrated circuits such as FPGA, ASIC, and GPU in the electronic circuit of the image data read control unit 21b, which is the drive and data read unit of the image sensor 15, thereby enabling high-speed and real-time data processing. Alternatively, a configuration in which circuits equivalent to these functions are implemented on the image sensor 15 is also possible.

[0116] This embodiment can be applied to not only refractive objective lenses but also reflective objective lenses (Schwarzschild type). In this case, it is possible to design an objective lens with a high numerical aperture in the short wavelength UV region, further improving spatial resolution. For example, Pr:LuAG, which generates fluorescent light F, which is ultraviolet light with a wavelength of 305 nm, can be used for the fluorescent film 11a.

[0117] The imaging optical system 17 may further include a cylindrical lens. By using a cylindrical lens with an appropriate focal length to match the focal depth of the post-objective lens group 12, it is possible to evaluate not only the size of the point image profile but also the shape of the profile, such as its aspect ratio, as shown in Figures 3(a) and 3(b), thereby improving the accuracy of discriminating between in-focus and out-of-focus signals and enabling high-resolution observation of the sample 900.

[0118] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be apparent to those skilled in the art that various modifications and improvements can be made to the above embodiments. It is clear from the claims that such modifications and improvements can also be included within the technical scope of the present invention.

[0119] It should be noted that the order of execution of each process, such as operations, procedures, steps, and stages, in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not specifically stated as "before," "prior to," etc., and that the processes can be performed in any order unless the output of a previous process is used in a subsequent process. Even if the operational flow in the claims, specifications, and drawings is described using "first," "next," etc. for convenience, this does not mean that the processes must be performed in this order. [Explanation of symbols]

[0120] 10 radiation camera, 11 SIL, 11a fluorescent film, 11b lenticular substrate, 11e lenticular fluorescent substrate, 11f SIL stage, 12 objective post-lens group, 13 band-pass filter, 14 imaging lens, 15 image sensor, 16 objective lens, 17 imaging optical system, 20 control unit, 21 camera control unit, 21a exposure time control unit, 21b image data readout control unit, 22 image signal processing unit, 22a focusing signal discrimination unit, 22b image data storage unit, 22c point image signal detection unit, 22d X-ray energy resolved image generation unit, 100 radiation imaging device, 200 radiation source, 300 stage control unit, 400 radiation attenuation plate, 500 radiation source control unit, 900 sample, 910 sample stage, RX-ray, F fluorescence, F1 focused fluorescence, F2 unfocused fluorescence, f Airy disk

Claims

1. a scintillator that receives radiation from a radiation source that has passed through an object to be imaged and emits fluorescence having a wavelength longer than that of the radiation; an imaging optical system that forms an image on an imaging plane through an objective lens whose depth of field is matched to the specimen plane set on the scintillator; an image sensor having a light receiving surface made up of a plurality of pixels arranged on the imaging surface; a radiation camera having a control unit that generates an image based on the sensor signals for each of the plurality of pixels output from the image sensor of the radiation camera, the control unit comprises: an exposure time control unit that controls sub-frame exposure times of the image sensor so that in-focus fluorescence at the depth of field generated by radiation reaching the scintillator and out-of-focus fluorescence outside the depth of field are discretely imaged for each particle of the radiation on the light-receiving surface of the image sensor; a focus signal discriminator that discriminates between point image sensor signals of light-receiving pixel groups corresponding to the in-focus fluorescence spots discretely included in each sub-frame and point image sensor signals of light-receiving pixel groups corresponding to the out-of-focus fluorescence spots; and an image signal processing unit that generates the image by integrating sub-frame data based on the point image sensor signals from the light-receiving pixel groups corresponding to the in-focus fluorescence spots.

2. 2. The radiation imaging apparatus according to claim 1, wherein the image signal processing unit includes a point image signal detection unit that processes point image sensor signals from a group of light-receiving pixels corresponding to the focused fluorescent spot, calculates a center position of a point image in the group of light-receiving pixels, and generates the image based on the center position.

3. 2. The radiation imaging apparatus according to claim 1, wherein the image signal processing unit includes a radiation energy decomposition image generating unit that sorts, into a plurality of level ranges in a plurality of consecutive subframes, a light receiving sensor signal level for each of the spots of the focused fluorescence generated in response to radiation having different energies discretely incident on the scintillator on a particle-by-particle basis, and generates the image corresponding to the energy distribution of the radiation based on sensor signals of the spots of the focused fluorescence that fall within the same level range.

4. 2. The radiation imaging device according to claim 1, wherein the exposure time control unit controls the exposure time in accordance with the intensity of the radiation so that the radiation detection frequency in the image sensor is 0.02 photons / pixel / subframe or less, and generates the image by obtaining at least 6.4×105 consecutive image pickup signals consisting of a plurality of subframes.

5. 2. The radiation imaging apparatus according to claim 1, wherein the objective lens includes a front lens having a solid immersion lens configuration in which a fluorescent film as a scintillator is integrally formed on the radiation incident surface side.

6. 2. The radiation imaging device according to claim 1, wherein said objective lens includes a front lens comprising a fluorescent film as said scintillator and a lenticular body on the radiation incident surface side thereof integrally formed with said fluorescent film, and said fluorescent film is made of a solid optical material having a solid optical material constituting said lenticular body as a base material and having a predetermined activator added thereto.

7. 2. The radiation imaging device according to claim 1, wherein the objective lens has a front lens composed of a fluorescent film as the scintillator and a lenticular body on the incident surface side of which the fluorescent film is integrally formed, and the difference in refractive index between the fluorescent film and the lenticular body is less than 0.1%.

8. 2. The radiation imaging device according to claim 1, wherein the objective lens includes a leading lens of a solid immersion lens configuration formed of a material having scintillator properties.

9. 7. The radiation imaging device according to claim 6, wherein said lens-shaped substrate is formed of undoped garnet crystal, and said fluorescent film is formed of garnet crystal obtained by adding an activator to said undoped garnet crystal.

10. 7. The radiation imaging device of claim 6, wherein the lenticular body is formed of undoped gadolinium aluminum gallium garnet (GAGG), and the fluorescent film is formed of gadolinium aluminum gallium garnet (Ce:GAGG) doped with cerium as an activator.

11. 2. The radiation imaging device according to claim 1, wherein the radiation imaging device has a total radiation-to-signal conversion efficiency such that fluorescence generated by one photon or one particle of radiation on a fluorescent film is detected as a point image signal of 50 electrons or more by a pixel on a light-receiving surface of an image sensor.

12. 2. The radiation imaging device of claim 1, wherein the radiation is X-rays, and an X-ray-to-signal total conversion efficiency ηTOTAL, which indicates the efficiency of conversion into a sensor signal detected by the image sensor, satisfies the condition [ηTOTAL=ηLY(E)×ηCE(NA)×ηTO×ηVQE>50], where ηLY(E) is the light emission efficiency of the optical system, ηCE(NA) is the light recovery efficiency of the optical system which depends on the numerical aperture NA, ηTO is the transmittance of the optical system, and ηVQE is the quantum efficiency of fluorescence of the image sensor.

13. a radiation imaging device including a camera unit that is primarily configured to observe a visible light image from said solid immersion type front lens using an image sensor having a light-receiving surface made up of a plurality of pixels, and that is provided with a scintillator on the radiation entrance surface side for converting incident radiation into visible light, and that includes a solid immersion type front lens having a lens-shaped exit surface side; the radiation imaging device controls the exposure time for each subframe of the image sensor to collect radiation incident on said scintillator as point image signals of visible light spots that are discrete for each particle; the point image signals in each subframe are then discriminated into point image signals of focused light spots emitted within the depth of field of said solid immersion type front lens and point image signals of out-of-focus light spots emitted outside the depth of field; and the radiation observation image is generated by superimposing subframe data based on the point image signals of the focused light spots.

14. a scintillator that receives radiation from a radiation source that has passed through an object to be imaged and emits fluorescence having a wavelength longer than that of the radiation; an imaging optical system that forms an image on an imaging plane through an objective lens whose depth of field is matched to the specimen plane set on the scintillator; an image sensor having a light receiving surface made up of a plurality of pixels arranged on the imaging surface; A radiation imaging device comprising a radiation camera having: a control step of generating an image based on the sensor signals for each of the plurality of pixels output from the image sensor; the control step includes: an exposure time control step of controlling subframe exposure times of the image sensor so that focused fluorescence at the depth of field generated by radiation reaching the scintillator and out-of-focus fluorescence outside the depth of field are discretely imaged for each particle of the radiation on the light-receiving surface of the image sensor; a focus signal discrimination step of discriminating between sensor signals of light-receiving pixel groups corresponding to the focused fluorescence spots discretely included in each subframe and sensor signals of light-receiving pixel groups corresponding to the out-of-focus fluorescence spots; and an image signal processing step of generating the image by superimposing subframe data based on sensor signals from light-receiving pixel groups corresponding to the focused fluorescence spots.

15. the radiation is an X-ray, and the radiation camera is an X-ray camera including a scintillator and an objective lens optically designed so that the conversion efficiency from the X-ray to a sensor signal detected by the image sensor is 50 or more electrons per X-ray photon, 15. The radiation imaging method according to claim 14, further comprising: a step of performing signal processing to detect point image center positions of the focused fluorescent spots; and a step of generating a super-resolution X-ray image by superimposing subframe data consisting of point image center data of the detected point image center positions.

Images