WO2014181889A1

WO2014181889A1 - Substance identification device and substance identification method employing x-ray panoramic/ct photographing

Info

Publication number: WO2014181889A1
Application number: PCT/JP2014/062631
Authority: WO
Inventors: 勉山河; 浩一尾川; 政廣辻田; 明敏勝又
Original assignee: 株式会社テレシステムズ
Priority date: 2013-05-10
Filing date: 2014-05-12
Publication date: 2014-11-13
Also published as: JPWO2014181889A1

Abstract

Provided are a device and a method for identifying (specifying or estimating) at least the type of a substance. For example, the present device and method are functionally provided by an X-ray photographing device that can perform X-ray panoramic photographing and X-ray CT photographing in combination. In this X-ray photographing device, frame data that is output from a photon-counting X-ray detector (22) is used for the reconstitution of a panoramic image. Further, the panoramic image is displayed (B3; S32), and a location of concern is designated on the displayed panoramic image in accordance with an instruction from an operator (B3; S34). Further, at least the type of one or more structures that have a thickness in an object being photographed and that are present in the projecting direction in a photographing space between an X-ray tube and the detector and corresponding to the location of concern is identified on the basis of data of said panoramic image and morphological information on said substance(s) as obtained from a CT image (B3; S35 to S40).

Description

Substance identification apparatus and substance identification method using X-ray panorama / CT imaging

The present invention relates to a substance identification apparatus and a substance identification method using X-ray panorama / CT imaging, and in particular, by combining panoramic image and CT image information obtained from X-ray panoramic imaging and CT imaging. The present invention relates to a substance identification apparatus and a substance identification method using a hybrid X imaging apparatus for identifying (specifying or estimating) at least a type of a structure (substance) at a position of interest.

One field of diagnostic apparatus using X-rays is a dental panoramic X-ray imaging apparatus.
In the field of this X-ray imaging apparatus, in recent years, tomography of a subject based on tomosynthesis has been actively performed. The principle of this tomosynthesis method has been known for a long time (see, for example, Patent Document 1). In recent years, a tomography method has also been proposed which seeks to enjoy the simplicity of image reconstruction based on the tomosynthesis method (see, for example, Patent Document 2 and Patent Document 3). Many examples are seen in dental and mammography (see, for example, Patent Document 4, Patent Document 5, and Patent Document 6).

In recent years, an X-ray panoramic imaging apparatus that uses a detector capable of collecting X-ray detection data at high speed (for example, 300 FPS) as in Patent Document 7 and loads all the detection data into a computer and executes a tomosynthesis method. Has been developed. In the case of this apparatus, the detection data is processed by the tomosynthesis method to generate a panoramic image of the tomographic plane, and the position of the tomographic plane is changed in the front-rear direction of the plane, and the panoramic image of the changed tomographic plane can be generated. .

This dental panoramic X-ray imaging device can visualize the entire oral cavity with low exposure. However, since the information of the tomographic direction cannot be obtained and the blur of the image exists, it has not yet been used for the diagnosis of the detailed examination for grasping the internal structure of the oral cavity.

In addition, when trying to image the dentition, the shadow of the cervical spine becomes an obstacle shadow, and this obstacle shadow inevitably reduces the image rendering ability. Further, when an integral type X-ray detector is used, there is a problem that accurate X-ray absorption is not always obtained.

For this reason, at present, the diagnosis of scrutiny relies exclusively on intraoral radiography and a dental X-ray CT apparatus with a film or X-ray sensor in the mouth. In particular, if a dental X-ray CT apparatus is used, three-dimensional shape information of the oral cavity can be obtained.

However, in the case of a dental X-ray CT apparatus, the X-ray tube voltage is as low as about 80 kV and the oral cavity has a lot of hard tissue. There is a problem that a stable CT value of an imaging target cannot be obtained due to the effects of beam hardening and scattered radiation.

Also, in a dental X-ray imaging apparatus, an integration type detector that integrates an X-ray transmission signal for a certain time is often used regardless of a panoramic apparatus or a CT apparatus. In such a case. Electric noise generated in the preamplifier processing circuit for the detection is mixed. In addition, due to the non-linear characteristics of the preamplifier, it is difficult to accurately reflect the absorption value from the high dose portion to the low dose portion, and so-called dynamic range becomes narrow. For this reason, if the detector is optimized in order to image hard tissue such as enamel and cancellous bone, the image of soft tissue such as gums and muscles will be out of contrast and difficult to image. On the contrary, when the soft tissue is optimized, the hard tissue is buried with noise and the contrast cannot be applied. Under these circumstances, it can be said that it is difficult to shoot both images at once.

Therefore, in recent years, as shown in Patent Document 8, a combined X-ray imaging apparatus capable of performing both panoramic imaging and CT imaging has also been proposed.

JP-A-57-203430 JP-A-6-88790 JP-A-10-295680 JP-A-4-144548 JP2008-11098A US Patent Publication US2006 / 0203959 A1 JP2007-136163A JP2011-206534

However, in the case of the conventional dual-purpose X-ray imaging apparatus, only the panoramic imaging mode and the CT imaging mode are switched and used. For this reason, the above problem has not been solved.

The present invention has been made in view of the above-described conventional situation. The features of the panoramic imaging function and the CT imaging function are complementarily fused, and normal panoramic imaging and CT imaging can be performed by one imaging (scanning). Of course, it is an object of the present invention to be able to identify (specify or estimate) at least a type of a thick structure in a region of interest to be imaged and to use the identification result at a clinical level.

In order to achieve the above-described object, as one aspect thereof, a substance identification device using X-ray panorama / CT imaging is provided. This material identification apparatus forms a two-dimensional pixel group by forming an X-ray tube for irradiating X-rays and a cell for outputting an electric pulse corresponding to the energy of the photon whenever the photon of the X-ray is detected. The detection circuit arranged in a two-dimensional manner and the number of photons detected by each cell of the detection unit are divided into two or more energy bands (energy band number M ≧ 2: M is a positive integer) A detector configured to include a measurement circuit that measures each time, and an output circuit that outputs a digital amount of an electrical signal corresponding to the output of each cell measured by the measurement circuit as frame data for each energy band; and The X-ray tube and the detector are opposed to each other with the imaging target interposed therebetween, the X-ray tube and the detector are rotated around the imaging target, and the frame data output by the detector is used as the energy band. Collect every time A scanning unit; a panoramic image generating unit that generates panoramic image data of the imaging target for each energy band based on the frame data; and a CT image generating unit that generates the CT (computed tomography) image of the imaging target; Panoramic image display means for displaying the panoramic image on a monitor; interest position designation means for designating a position of interest on the panoramic image displayed on the monitor according to an operator's instruction; and the position corresponding to the position of interest At least the kind of one or more substances having a thickness in the imaging target existing in the projection direction in the imaging space between the X-ray tube and the detector is obtained from the X-ray transmission data of the panoramic image and the CT image. A substance identifying means for identifying based on the obtained form information of the substance.

As another aspect, an X-ray tube that irradiates X-rays, and a cell that outputs an electric pulse corresponding to the energy of the photons each time the photons of the X-rays are detected include a two-dimensional pixel group. The detection circuits arranged two-dimensionally to form and the number of photons detected by each cell of the detection unit are divided into two or more energy bands (energy band number M ≧ 2: M is a positive integer). A detector configured to include a measurement circuit that measures each pixel, and an output circuit that outputs a digital amount of an electrical signal corresponding to the output of each cell measured by the measurement circuit as frame data for each energy band; The X-ray tube and the detector are opposed to each other with the imaging target interposed therebetween, the X-ray tube and the detector are rotated around the imaging target, and the frame data output by the detector is converted into the energy. Collect by band Applied material identification method to a system with a scanning unit, the for is provided. The method for identifying a substance includes generating a panoramic image data of the imaging target for each energy band based on the frame data, generating a CT (computed tomography) image of the imaging target, and the panoramic image. Between the X-ray tube and the detector corresponding to the position of interest, a step of designating a position of interest on the panoramic image displayed on the monitor according to an operator's instruction, Based on X-ray transmission data of the panoramic image and morphological information of the substance obtained from the CT image, the at least one kind of substance having a thickness in the imaging object existing in the projection direction in the imaging space. And identifying.

As described above, according to the present invention, the features of the panorama imaging function and the CT imaging function are complementarily fused, and normal panorama imaging and CT imaging can be selectively performed by one imaging (scanning). it can. Furthermore, according to the present invention, at least the types of structures having a thickness in the region of interest to be imaged can be identified (specified or estimated), and can be used at the clinical level.

In the accompanying drawings,
FIG. 1 is a perspective view showing an outline of the overall configuration of a hybrid X-ray imaging apparatus that functionally provides a substance identification apparatus of the present invention, FIG. 2 is a view for explaining the positional relationship between the imaging system of the apparatus and the jaw of the subject; FIG. 3 is a diagram for explaining the positional relationship between the rotation for scanning of the imaging system and the jaw of the subject; FIG. 4 is a plan view illustrating a detection surface (collection window) of the detector, FIG. 5 is a circuit diagram illustrating the electrical configuration of the detector. FIG. 6 is a diagram for explaining an incident pulse of X-rays and a threshold for photon counting; FIG. 7 is a diagram for explaining an energy region for X-ray energy distribution and photon counting; FIG. 8 is a block diagram showing the main part of the overall electrical configuration of the X-ray imaging apparatus; FIG. 9 is a functional block diagram showing processing executed by the controller and the data processor in order to identify the type of composition of the subject's jaw. FIG. 10 is a display diagram illustrating an example of an optimally focused 3D autofocus image; FIG. 11 is a diagram for explaining a scan range; FIG. 12 is a diagram for explaining the positional relationship between the dentition, the reference tomographic plane of the dentition, and the scan of the X-ray tube / detector; FIG. 13 is an explanatory diagram illustrating a tooth specified for substance identification and an irradiation direction of an X-ray beam that passes through the tooth; FIG. 14 is a schematic flowchart illustrating the flow of removing metal artifacts, FIG. 15 is a diagram for explaining a part of the process of removing metal artifacts. FIG. 16 is a diagram illustrating another part of the process of removing metal artifacts, FIG. 17 is a diagram illustrating another part of the process of removing metal artifacts, FIG. 18 is a diagram illustrating another part of the process of removing metal artifacts, FIG. 19 is a diagram illustrating another part of the process of removing metal artifacts. FIG. 20 is a diagram for explaining another partial process of removing metal artifacts; FIG. 21 is a diagram illustrating another part of the process of removing metal artifacts, FIG. 22 is a diagram illustrating another part of the process of removing metal artifacts, FIG. 23 is a diagram illustrating another part of the process of removing metal artifacts. FIG. 24 is a graph showing an example of a scatter diagram calculated for substance identification; FIG. 25 is a graph showing an example of a reference scatter diagram prepared in advance. FIG. 26 is a graph showing another example of a reference scatter diagram prepared in advance. FIG. 27 is a flowchart schematically showing a flow of scan and obstacle shadow removal executed in the embodiment; FIG. 28 is a diagram showing a part of the process of removing the obstruction shadow; FIG. 29 is a graph showing an example of a gain curve for reconstructing a panoramic image of a dentition, FIG. 30 is a graph showing an example of a gain curve for reconstructing a panoramic image of the cervical spine, FIG. 31 is an image showing an example of a panoramic image of a dentition (before removal of an obstacle shadow), FIG. 32 is an image showing an example of a panoramic image of the cervical spine, FIG. 33 is an image (an image of a patient different from the patient in FIG. 13) showing an example of the panoramic image of the dentition before removing the obstacle shadow, and FIG. 34 is an image showing an example after removal (reduction) of obstacle shadows in the panoramic image of the dentition to be compared with FIG.

Hereinafter, embodiments of a substance identification apparatus and a substance identification method according to the present invention will be described with reference to the accompanying drawings. The “identification” of the substance identification means that at least the kind of the substance is specified or estimated based on the X-ray information transmitted through the substance (the structure that forms the imaging target).

1 to 24, a hybrid X-ray imaging apparatus functionally equipped with the substance identification apparatus and the substance identification method will be described. In this embodiment, this X-ray imaging apparatus is implemented as a dental hybrid X-ray imaging apparatus.

Here, the hybrid type indicates that there is a function of performing both X-ray panoramic imaging (or X-ray panoramic imaging) and X-ray CT imaging (or X-ray CT imaging). By operating this hybrid X-ray imaging apparatus, a “substance identification apparatus and substance identification method using X-ray panorama / CT imaging” can be provided functionally.

FIG. 1 shows an appearance of such a hybrid X-ray imaging apparatus 1. The imaging apparatus 1 scans the subject's jaw with X-rays, and has both imaging functions of panoramic imaging and CT imaging of the jaw from the digital amount of X-ray transmission data.

An outline of the configuration of the hybrid X-ray imaging apparatus 1 (hereinafter simply referred to as an X-ray imaging apparatus) will be described. As shown in FIG. 1, the X-ray imaging apparatus 1 includes a casing 11 that collects data from a subject (patient) P, for example, while the subject P is standing or sitting in a wheelchair, and the casing 11 For controlling the collection of data to be performed, capturing the collected data to create a panoramic image, and performing post-processing of the panoramic image interactively or automatically with an operator (doctor, engineer, etc.) And a control / arithmetic unit 12 (also called a console) constituted by a computer.

The housing 11 includes a stand unit 13 and a photographing unit 14 that can move up and down with respect to the stand unit 13. The imaging unit 14 is attached to the support column of the stand unit 13 so as to be movable up and down within a predetermined range.

Here, for convenience of explanation, for the X-ray imaging apparatus 1, an XYZ orthogonal coordinate system having the longitudinal direction of the stand unit 13, that is, the vertical direction as the Z axis is set.

The photographing unit 14 includes a vertically moving unit 15 that is substantially U-shaped when viewed from the side, and a rotating unit 16 that is supported by the vertically moving unit 15 so as to be rotatable (rotatable). The vertical movement unit 15 can move up and down relative to the stand unit 13 in accordance with control from the control / arithmetic apparatus 12.

Further, the rotation unit 16 is configured by a substantially U-shaped arm as viewed from the side. The arm is suspended downwardly by the vertical movement unit 15 with the arm facing downward. The rotating unit 16 is also rotated in accordance with control from the control / arithmetic apparatus 12.

An X-ray tube 21 serving as an X-ray source and an X-ray detector 22 serving as an X-ray detection means (hereinafter simply referred to as a detector) are opposed to the two

arms

16A and 16B extending in the vertical direction of the rotary unit 16. It is arranged to do. When the rotation unit 16 rotates along the XY plane, the pair of the X-ray tube 21 and the detector 22 also rotate in a state of facing each other, and a space between them is defined as an imaging space S.

In this imaging space S, the jaw JW of the subject P is positioned using the chin rest 19 and the headrest 20.

For this reason, in this embodiment, the X-ray irradiated from the X-ray tube 21 passes through the jaw JW of the subject P and enters the detector 22 as shown in FIG. Become.

Further, the pair of the X-ray tube 21 and the detector 22 provided in the rotating unit 16 includes two circular orbits Tx.Dx having different diameters Dx and Dd from the rotation center O as shown in FIG. Rotate while maintaining the postures facing each other on Td. Dx = Dd may be sufficient.

Note that the above-mentioned “always facing each other (or facing each other)” as shown in FIG. 3 is irradiated from a dotted X-ray focal point FP of the X-ray tube 21 when viewed from the Z-axis direction, which will be described later. The center axis T of the X-ray beam shaped like a cone by the slit 23 always intersects (or intersects at 90 °) the center point C in the width direction of the X-ray detection surface 22A (described later) of the detector 22. Say. In particular, crossing at 90 ° is said to be facing here (see FIG. 3).

The X-ray tube 21 is composed of, for example, a rotating anode X-ray tube, and radiates X-rays radially from the target (anode). The focal point of the electron beam colliding with the target is as small as about 0.1 to 0.5 mm in diameter, for example, and therefore the X-ray tube 21 functions as a point X-ray source. A slit 23 is attached to a predetermined position on the front surface of the X-ray tube 21. With this slit 23, the X-rays incident on the detector 22 can be narrowed in a cone shape in accordance with the shape of a desired collection window on the detection surface.

As shown in FIG. 4, the detector 22 has an array (sensor circuit) of a plurality of detection modules B1 to Bm in which X-ray imaging elements are two-dimensionally arranged. The plurality of detection modules B1 to Bm are created as blocks independent of each other, and are mounted in a predetermined shape (for example, a rectangular shape) on a substrate (not shown) to form the entire detector 22.

A plurality of detection modules B1 to Bm are provided in two dimensions in the vertical (X-axis) and horizontal (Y-axis) directions (15 in the vertical direction, horizontal direction) while providing a certain gap between the individual modules. 8, together with arranging by further placing the 5 upper and lower ends), they are arranged obliquely inclined by an angle θ with respect to the scanning direction O _Y individual modules. This angle θ is set to about 14 °, for example. The surface of a rectangular shape (in the case of CT imaging) having a small length ratio (in the case of CT imaging) or a large ratio of length to width, that is, an elongated rectangular shape (in the case of panoramic imaging), is created by the plurality of detection modules B1 to Bm. An X-ray detection surface 22A is formed. Since the detection modules B1 to Bm are arranged obliquely, the X-ray detection surface 22A is formed so as to follow (inscribe) the inside of each detection surface of the plurality of modules B1 to Bm.

The structure of the detector 22 having this detection module arranged obliquely and the processing of the detection signal by the sub-pixel method are known, for example, from WO 2012/0866648 A1.

Note that a plurality of modules arranged in a column in the left-hand column in FIG. 4 function as panoramic shooting modules. The opening area for panoramic photography is indicated by reference numeral 22B. Further, a module group of a two-dimensional array formed by all the modules except the two modules at the upper and lower ends of the left side row and the remaining modules functions as a module for CT imaging. The aperture area of the small field for CT imaging is indicated by reference numeral 22A. As a result, the detector 22 has a collection window of “a shape having a long portion that is rectangular and further extends vertically on one end side”, in which the

opening areas

22A and 22B are combined.

Selection of the module group according to whether it is for panoramic imaging or CT imaging can be performed by controlling the opening area of the slit 23, but in this embodiment, panoramic imaging and CT imaging are performed simultaneously in one scan. It is also a feature to do. For this reason, the slit 23 narrows the fan-shaped X-ray beam so as to match the shape of the collection window in which the

opening areas

22A and 22B are combined.

In addition, reference symbol AXd in FIG. 4 is a central axis when the detector 22 itself rotates (rotates). In the present embodiment, since the detector 22 is always controlled to take a posture facing the X-ray tube 21, it is not always necessary to control this rotation operation.

Individual detection modules B1 (˜Bm) are made of a semiconductor material that directly converts X-rays into electrical pulse signals. For this reason, the detector 22 is a photon counting X-ray detector of a direct conversion method using a semiconductor.

As described above, the detector 22 is formed as an array of a plurality of detection modules B1 to Bm. Each detection module Bm includes a detection circuit Cp (see FIG. 5) for detecting X-rays and a data counting circuit 51 _n (see FIG. 5) stacked together with the detection circuit Cp, as is well known.

The detection circuit Cp includes, for each detection module, a semiconductor layer that directly converts X-rays into an electrical signal, and a charging electrode and a collecting electrode that are respectively stacked on both sides (not shown). X-rays are incident on the charged electrode. The charged electrode is a common electrode, and a high bias voltage is applied between the charged electrodes. The semiconductor layer and the collecting electrode are divided into a grid pattern, and by this division, a plurality of small regions are formed that are arranged in a two-dimensional array at a certain distance from each other. As a result, a plurality of stacked bodies of semiconductor cells C (see FIGS. 4 and 5) and collecting electrodes arranged in a two-dimensional manner on the charged electrode are formed. The plurality of stacked bodies to form a plurality of pixels S _n arranged in a two dimensional grid pattern.

As a result, a plurality of pixels S _n (n = 1 to N) occupying a predetermined area necessary for the detector 22 by the whole of the plurality of detection modules B1 to Bm (however, depending on the opening area 22A at the time of CT imaging: see FIG. 4). ) Is formed. The plurality of pixels _{S n} detection circuits (pixel group) constituting the Cp (refer to FIG. 5).

The size of each pixel S _n is, for example, 200 [mu] m × 200 [mu] m, the pixel size is set to a detectable value X-rays incident as a set of multiple photons. Each pixel S _n is responsive to incident of each photon of X-ray, and outputs an electrical pulse of amplitude corresponding to the energy possessed by the photon. That is, each pixel S _n may convert the X-rays incident on that pixel directly, into electric signals.

Therefore, the detector 22, the photon constituting the cone beam-like X-rays incident, counts for each pixel S _n constituting the detection surface of the detector 22 (acquisition window), to reflect the counted value The amount of electricity data is output at a relatively high frame rate of, for example, 75 fps. This data is also called frame data.

As the semiconductor material of the semiconductor layer, that is, the semiconductor cell C, cadmium telluride semiconductor (CdTe semiconductor), cadmium zinc telluride semiconductor (CdZnTe semiconductor (CZT semiconductor)), silicon semiconductor (Si semiconductor), thallium bromide (T1Br) Mercury iodide or the like is used. Instead of this semiconductor cell, it is composed of a cell that combines a scintillator material that is subdivided into columns and optically shielded from each column, and a photoelectric converter composed of a combination of fine avalanche photodiodes. May be.

For this reason, when X-rays enter the semiconductor cell C, charges (electrons, holes) are generated inside the cell, and a pulse current corresponding to the amount of the charge flows. This pulse current is detected by the current collecting electrode. As a result, the amount of charge varies depending on the energy value of the X-ray photons. Therefore, the detector 22 outputs an electrical pulse signal corresponding to the energy value of the photons for respective pixels S _n.

The detector 22 further comprises respective semiconductor cell C, that the data counting circuit ₅₁ on the output side of each of the plurality of pixels _{S n} n the (n = 1 ~ N). Here, each pixel _{S n,} i.e., the path from each semiconductor cell C to the data counting circuit ₅₁ 1 (~ ₅₁ _N), optionally, referred to as acquisition channels _{CN n (n = 1 ~ N} ) (See FIG. 5).

The structure of this group of semiconductor cells C is also known from Japanese Patent Application Laid-Open Nos. 2000-69369, 2004-325183, and 2006-101926.

Meanwhile, the size (200 [mu] m × 200 [mu] m) of each pixel S _n described above is capable of detecting the number of incident photons about negligible counting loss as photons (particles) the X-rays incident in one pixel sufficiently It is a small value. In the present embodiment, the size capable of detecting X-rays as the particles is “between electric pulses responding to each incident when a plurality of radiation (for example, X-ray) particles are successively incident at or near the same position. The occurrence of the superposition phenomenon (also called pile-up) is defined as “a size that can be substantially ignored or whose amount is predictable”.

However, even with such a pixel size, it is not possible to avoid all occurrences of the superposition phenomenon. Even when two or more electrical pulses are both observed in the same pixel, the superposition phenomenon does not occur if they are separated from each other in time. On the other hand, when two or more electric pulses are difficult to separate in time in the same pixel, a superposition phenomenon occurs, and the two electric pulses overlap to be observed as one electric pulse having a high peak value. Is done.

When this superposition phenomenon occurs, X-ray particle countdown (also called pile-up count loss) occurs in the characteristic of “number of incidents versus actual count value” of X-ray particles. Therefore, the size of the pixel S _n to form the detector 22, the magnitude of which can be regarded as the counting loss does not occur or does not substantially occur, or are set to an extent counting the drop amount can be estimated.

Subsequently, a circuit electrically connected to the detector 22 will be described with reference to FIG. Each of the plurality of data counting circuits 51 _n (n = 1 to N) includes a charge amplifier 52 that receives an electrical signal of an analog amount output from each semiconductor cell C. Circuit 53, multi-stage comparator 54 _i (here i = 1 to 4), multi-stage counter 56 _i (here i = 1 to 4), multi-stage D / A converter 57 _i (here i = 1 to 4) 4) A latch circuit 58 and a serial converter 59 are provided.

Each charge amplifier 52 is connected to each collector electrode of each semiconductor cell C, charges up the charge collected in response to the incidence of X-ray particles, and outputs it as a pulse signal of electric quantity. The output terminal of the charge amplifier 52 is connected to a waveform shaping circuit 53 whose gain and offset can be adjusted. The waveform of the detected pulse signal is processed with the previously adjusted gain and offset to shape the waveform. The gain and offset of the waveform shaping circuit 53, in consideration of the variation in non-uniformity and the circuit characteristics for charge-charge characteristic for each pixel S _n of semiconductor cell C, is calibrated. As a result, it is possible to increase the output of the waveform shaping signal from which non-uniformity has been eliminated, and the relative threshold setting accuracy. As a result, corresponding to each pixel S _n, i.e., the characteristics reflecting the energy value of the X-ray particle pulse signal waveform formatted output from the waveform shaping circuit 53 for each collection channel CN _n is substantially incident Have. Therefore, the variation between the collection channels CN _n is greatly improved.

The output terminal of the waveform shaping circuit 53 is connected to the comparison input terminals of the plurality of comparators 54 ₁ to 54 ₄ . As shown in FIG. 5, analog amount threshold values (voltage values) th _i (here, i = 1 to 4) having different values are applied to the reference input terminals of the plurality of comparators 54 ₁ to 54 _4, respectively. Yes. This makes it possible to compare one pulse signal with each of the different analog amount thresholds th ₁ to th ₄ . FIG. 6 shows the magnitude relationship (th ₁ <th ₂ <th ₃ <threshold) between the peak value (representing energy) of the pulse voltage generated in response to the input of one X-ray photon and the threshold values th ₁ to th _4. th ₄ ) schematically.

The reason for this comparison is to examine which region (discrimination) the energy value of the incident X-ray particle enters among the energy regions set in advance divided into a plurality. A determination is made as to which of the analog amount threshold values th ₁ to th ₄ exceeds the peak value of the pulse signal (that is, the energy value of the incident X-ray photon). Thereby, the energy area | region discriminated differs. Note that the lowest analog amount threshold th ₁ is usually to prevent detection of disturbances, noise caused by circuits such as the semiconductor cell C and the charge amplifier 52, or low-energy radiation that is not necessary for imaging. Is set as the threshold value. Further, the number of thresholds, i.e., the number of comparators is not necessarily limited to four, three, including the amount of the analog amount threshold th _1, or may be five or more.

Analog amount threshold th 1 _~ th ₄ described above, specifically, given from the calibration computing unit 38 of the console 17 for each pixel S _n in a digital value through the interface 31, i.e., for each acquisition channels. Therefore, the reference input terminals of the comparators 54 ₁ to 54 ₄ are connected to the output terminals of the _four D / A converters 57 ₁ to 574, respectively. The D / A converter ₅₇ 1-57 ₄ is connected to the threshold receiving end _T 1 via the latch circuits 58 (~ _{T N),} the interface 31 of the threshold receiving end _{_T} 1 (~ _T _N) console 17 It is connected.

The latch circuit 58 latches the thresholds th ₁ ′ to th ₄ ′ of digital quantities given from the threshold applier 40 via the interface 31 and the threshold receiving end T ₁ (˜T _N ) at the time of shooting, and the corresponding D / are output to a converters ₅₇ 1-57 _4. Therefore, the D / A converters 57 ₁ to 57 ₄ can supply the commanded analog amount thresholds th ₁ to th ₄ to the comparators 54 ₁ to 54 ₄ as voltage amounts, respectively. Each collection channel CN _n is one from the D / A converter 57 _i (i = 1 to 4) to the counter 56 _i (i = 1 to 4) via the comparator 54 _i (i = 1 to 4). Or it is connected to a plurality of circuit systems. This circuit system is called “discrimination circuit” DS _i (i = 1 to 4).

FIG. 7 shows a setting example of the energy threshold TH _i (i = 1 to 4) corresponding to the analog amount threshold th _i (i = 1 to 4). This energy threshold TH _i (i = 1 to 4) is of course a discrete value that is set discretely and can be set to an arbitrary value by the user. FIG. 7 schematically shows an X-ray spectrum when an appropriate material is used for the anode material of the X-ray tube 21. The horizontal axis indicates X-ray energy, and the vertical axis indicates the incidence frequency of X-ray photons. This incidence frequency is a factor representative of the count value (count) or intensity of X-ray photons.

The analog amount threshold th _i is an analog voltage applied to the comparator 54 _i in each discrimination circuit DS _i , and the energy threshold TH _i is an analog value for discriminating the X-ray energy (keV) of the energy spectrum. The waveform shown in FIG. 7 shows a continuous spectrum of the energy of X-rays exposed from an X-ray tube that is normally used, for example, using tungsten as an anode material. The count value (count) on the vertical axis is an amount proportional to the photon generation frequency corresponding to the energy value on the horizontal axis, and the energy value on the horizontal axis is an amount depending on the tube voltage of the X-ray tube 21. With respect to this spectrum, the first analog quantity threshold th ₁ is set as the X-ray photon count non-counting area (the area where there is no meaningful X-ray information for counting and the circuit noise is mixed) and the lower first to set corresponding to the energy region ER ₁ and energy threshold value TH ₁ capable discrimination of. Further, the second and third analog amount threshold values th ₂ and th ₃ are set so as to sequentially provide the second and third energy threshold values TH ₂ and TH ₃ which are higher than the first energy threshold value TH ₁ . . Further, the fourth energy threshold TH ₄ is set to an energy value equal to the applied voltage to the X-ray tube, in which X photon count value = 0 if there is no superposition phenomenon in the energy spectrum. Here, the fourth energy threshold TH _4, each pixel S _n, it is one of the important features of the present application are in accordance with the energy value of the count value = 0.

As a result, appropriate discrimination points based on energy spectrum characteristics and design values are defined, and energy regions ER ₁ to ER ₄ are set.

These energy thresholds TH _i are determined so that one or more subjects as a reference are assumed and the count value for a predetermined time for each energy region is substantially constant.

Therefore, the output of the comparator ₅₄ 1-54 _3, as shown in FIG. 5, is connected to the input ends of the plurality of counters ₅₆ 1-56 _4.

Each of the counters ₅₆ 1-56 ₄ counts up every time the output of the comparator ₅₄ 1-54 ₃ (pulse) is turned on. As a result, the number of X-ray photons having energy equal to or higher than the energy value discriminated into the energy region ER ₁ (to ER ₄ ) that each counter 56 ₁ (to 56 ₄ ) takes charge of is accumulated value W ₁ ′ ( it can be counted for each pixel _{S n} as ~ _{W 4} ').

Specifically, this counting operation is determined by the relationship between the detection voltage V _dec (detected energy value of photons) input to the _four comparators 54 ₁ to 54 ₄ and the threshold values th ₁ to th ₄ . That is, when the detection voltage V _dec <th ₁ to th ₄ , the outputs of all the comparators 54 ₁ to 54 ₄ are turned off. In other words, the output = 0 of the pixel _{S n.} As a result, noise components smaller than the energy threshold TH ₁ defined as the input energy counting limit are not counted. This noise component corresponds to an energy value signal belonging to the non-countable region ERx in FIG.

However, if the detection voltage V _dec exceeds the minimum threshold th ₁ (V _dec ≧ th ₁ ), the number of photons is counted. If the relationship is V _dec ≧ th ₁ , the outputs of all the comparators 54 ₁ to 54 ₄ are turned on. That is, the count value _W 1 of all the counters _{_{_{56 1 ~ 56 4 '~ W}}} 4' is counted up.

When the relationship of V _dec ≧ th ₂ is established, the outputs of the three comparators 54 ₂ to 54 _{4 in} the second and subsequent stages are turned on. Thus, the three counters ₅₆ 2-56 ₄ counts _{_W} 2 '~ _W _4' is counted up. If the relationship of V _dec ≧ th ₃ is established, the outputs of the third-stage and fourth-stage comparators 54 ₃ and 54 ₄ are turned on. As a result, the count values W ₃ ′ and W ₄ ′ of the two counters 56 ₃ and 56 ₄ are counted up.

_Furthermore, if the relationship between _{V dec} ≧ th _4, the output of the comparator module 54 ₄ of the fourth stage is turned on, only the counter 56 ₄ count value _{W 4} 'of the fourth stage is counted up. In this case, the energy value of the photon related to the input is a noise component belonging to the region ER ₄ exceeding the third high energy region ER ₃ , disturbance, etc., which is not suitable for imaging or counting. On the other hand, the count value W ₄ ′ can be used as information for estimating or excluding photons that have caused a superposition phenomenon or simultaneously incident photons.

As described above, in the present embodiment, the counters 56 ₁ to 56 ₄ count the number of photons having energy exceeding the energy range ER ₁ (to ER ₄ ) to be counted by the counters 56 ₁ to 56 ₄ , respectively. Therefore, the number of X-ray photons having energy belonging to each of the first to fourth energy regions ER ₁ to ER ₄ , that is, the number of X-ray photons to be obtained for each energy region is expressed as W ₁ , W ₂ , W ₃ , W ₄ , the relationship between the count values W ₁ ′, W ₂ ′, W ₃ ′, W ₄ ′ of the counters 56 ₁ to 56 ₄ is
W ₁ = W ₁ '-W ₂ '
W ₂ = W ₂ '-W ₃ '
W ₃ = W ₃ '-W ₄ '
It becomes. Note that W ₄ = W ₄ ′ is not calculated because it is meaningless information (that is, the energy region of the X-ray photon cannot be specified) due to the superposition phenomenon.

Therefore, the count values W ₁ to W _{4 that} are to be truly obtained are obtained by subtraction processing based on the above equation by a data processor described later. Ideally, W ₄ = W ₄ ′ = 0.

Thus, in the present embodiment, X-ray photon number _W 1 to _{W 4} belonging to each of the first through fourth energy regions _ER 1 to ER _4, the actual count value _W 1 '~ _{W 4'} Is obtained by calculation (subtraction). For this reason, a circuit for deciphering which energy region ER1 to ER4 the current event, that is, the incidence of X-ray photons belongs, becomes unnecessary from the combination of turning on and off the outputs of the comparators 54 ₁ to 54 ₄ . This simplifies the circuit configuration mounted on the data counting circuit 51 _n of the detector 22.

It should be noted that the meaning of “collection” for each energy region of the number of X-ray photons according to the present application includes the meaning “obtained by calculation” from the actual count value as described above.

The counter 56 _1-56 ₄ described above start and stop signals is supplied via a start-stop terminal T2 from below to the controller of the console 17. Counting for a fixed time is managed from the outside using a reset circuit included in the counter itself.

Thus, during a certain period of time until reset by a plurality of counters 56 _1-56 _4, the number of photons of X-rays incident on the detector 22 is counted for each pixel S _n. The number of photons counted value W _k of the X-ray _'(k = 1 ~ 4), after being outputted from each of the counters 56 _1-56 ₄ in parallel as the count value of the digital quantity, serial format by serial converter 59 Is converted to The serial converter 59 ₁ is connected to the serial converter 59 2 _~ 59 _N and serial all remaining acquisition channels. Therefore, the count of all digital content is output from the last channel of the serial converter 59 _N serially sent to the console 17 via the transmitting end T3.

In the console 17, the interface 31 receives these count values and stores them in a storage unit to be described later.

As described above, the count values of the respective pixels S _n (with common pixels) for panoramic imaging and CT imaging are formatted in a predetermined order and sequentially output serially, so that a frame by panoramic imaging and CT imaging is obtained. Data is collected at every sampling time. For this reason, by specifying the address of the frame data, it is possible to obtain frame data for only panoramic imaging and frame data for CT imaging.

In the present embodiment, it is integrally constructed in CMOS by the semiconductor cell C and the data counting circuit 51 _n corresponding to N pixels S _n described above ASIC (Application Specific Integrated Circuit). Of course, the data counting circuit 51 _n may be configured as a circuit or device separate from the group of semiconductor cells C.

In the above-described embodiment, the plurality of detection modules B1 to Bm are connected to a scintillator array in which a plurality of scintillators processed into columnar shapes are bundled, and receives light incident from the scintillator. A plurality of avalanche photodiodes are mounted on the light receiving surface, and the avalanche photodiodes belonging to the region are electrically connected by a quenching element for each rectangular region having a predetermined size corresponding to the cell on the light receiving surface. And a silicon photomultiplier.

The material of the scintillator is LFS (lutetium silicate), GAGG: Ce (gadolinium aluminum gallium garnet), LuAG: Pr (praseodymium-added lutetium aluminum garnet), or the same decay time and light emission amount as the LuAG: Pr. It may be a material having a specific gravity.

As shown in FIG. 8, the console 17 includes an interface (I / F) 31 that performs input and output of signals, a controller 33 that is communicably connected to the interface 31 via a bus 32, and a first storage unit 34, a data processor 35, a display unit 36, an input unit 37, a calibration calculator 38, a second storage unit 39, ROMs 40A to 40D, and a threshold value assigner 41.

The controller 33 controls driving of the X-ray imaging apparatus 1 in accordance with a program given in advance to the ROM 40A. This control includes sending a command value to the high voltage generator 42 that supplies a high voltage to the X-ray tube 21 and a drive command to the calibration calculator 38. The first storage unit 34 stores frame data and image data that are count values sent from the detector 22 via the interface 31.

The data processor 35 operates based on a program given in advance to the ROM 40B under the control of the controller 33. By the operation, the data processor 35 processes the frame data stored in the first storage unit 34 by a desired CT reconstruction method to perform a CT image reconstruction process. Further, the data processor 35 performs a tomosynthesis method based on a known calculation method called “shift and add” on the frame data stored in the first storage unit 34 by the operation. Thereby, a CT image and a panoramic image of the jaw JW of the subject P are obtained. The display unit 36 is responsible for displaying an image to be created, information indicating the operation status of the apparatus, and operator operation information given via the input unit 37. The input device 37 is used for an operator to give information necessary for photographing to the apparatus.

Further, the calibration computing unit 38, under the control of the controller 33, operating under program built in advance in ROM40C, giving for each energy discriminator circuit for each pixel S _n in the data counting circuit, X-rays energy Calibrate the digital quantity threshold for discrimination.

Under the control of the controller 33, the threshold value assigner 41 calls the digital amount threshold value stored in the second storage unit 39 at the time of photographing for each pixel and for each discrimination circuit, and uses the threshold value as a command value as an interface. 31 to the detector 22. In order to execute this process, the threshold value assigner 41 executes a program stored in advance in the ROM 40D.

The controller 33, the data processor 35, the calibration calculator 38, and the threshold value assigner 41 are all provided with a CPU (central processing unit) that operates according to a given program. Those programs are stored in advance in each of the ROMs 40A to 40D.

In this embodiment, the data processor 35 reads the count value stored in the first storage unit 34 in response to an operator command from the input device 37, and uses this count value for image processing and substance identification. The commanded process such as the above process and the measurement process is executed. The image processing includes, for example, generation of a panoramic image of a cross section of a dentition based on a mosynthesis method for “panoramic imaging” and generation of a tomographic image based on a desired reconstruction method for “CT imaging”.

Also, the concept of substance identification includes identification (specification) of the types of structures (substances) constituting the jaw and the state of the structures using beam hardening information. This substance identification process is one of the features of the present application.

Here, referring to FIG. 9, a scan (imaging) for substance identification, preparation for substance identification, and substance identification executed by the dental X-ray imaging apparatus 1 according to the present embodiment. An outline of the processing will be described. The substance identification referred to in the present embodiment identifies at least the type of substance at the position of interest (part) designated by a user such as a dentist in the jaw of the subject P scanned by X-rays. The substance is a substance having a thickness in the X-ray irradiation direction (tongue, cancellous bone, enamel, cortical bone, metal (stuffing, covering), etc.).

[Scan]
First, scanning (imaging) performed by the X-ray imaging apparatus 1 will be described (see block B1).

According to this scan, a scan for panoramic imaging can be performed by rotating both the X-ray tube 21 and the detector 22 along a circular orbit, and a scan for CT imaging may be a half scan. The range in which the pair of the X-ray tube 21 and the detector 22 is actually rotated is a range of an angle α from the initial position as shown in FIG. 10, for example, and this angle α is an angle range of a half scan for CT imaging. This is a range determined by the angle range necessary for panoramic photography (for example, 210 °).

In addition, these two panoramic and CT scans are integrated while the X-ray tube 21 and the detector 22 are rotated only once around the jaw JW of the subject P over the angular range α. To be executed. That is, since panoramic imaging and CT imaging can be performed simultaneously, it is possible to shorten the scanning time, reduce the X-ray exposure amount, reduce the operator's labor, and the like.

Of course, CT scanning may be full scan instead of half scan. Further, panoramic imaging and CT imaging may be performed as separate scans.

When the jaw portion JW of the subject P is positioned at a predetermined position in the imaging space IS, the controller 33 rotates the X-ray tube 21 and the detector 22 that rotate the rotating unit 16 around the jaw portion JW. At this time, as shown in FIG. 3, there are two circular orbits Tx and Td having different radii Dx and Dd from the rotation center O. During this rotation, for example, continuous X-rays are exposed radially from the X-ray focal point FP of the X-ray tube 21. This X-ray is narrowed by the slit 23 so as to follow the shape of the collection window “22A + 22B” (see FIG. 4) of the detector 22. The narrowed X-rays pass through various substances present in the jaw JW and enter the collection window of the detector 22. The detector 22 samples incident X-rays at a constant frame rate (for example, 75 fps). As a result, digital frame data corresponding to the number of X-ray photons incident on each pixel is output from the rotating detector at regular intervals for each energy region ER ₁ (˜ER ₃ ). These frame data are sequentially or collectively transferred to the first storage unit 34 of the console 17.

[Advance preparation]
Next, the preparations required for performing the identification process shown in block B2 of FIG. 9 will be described. For this preparation,
Creation of correction data for correcting a panoramic image using tough bones (block 21):
Calibration for analyzing the geometric relationship for each projection angle between the imaging space IS, the X-ray tube 21, the detector 22, and their rotation center (or preparation of the theoretical value) (block B22):
In this embodiment, as known from International Publication No. WO2011 / 142343 (International Application No. PCT / JP2011 / 060731), the structure of the imaging space IS is analyzed using a phantom, and the collection channel CN _{n of the} detector 22 is analyzed. Is calibrated. This calibration is executed at an appropriate timing such as before photographing or during maintenance inspection.
-Creation of a reference scatter graph database (block B23) to be described later,
There is.

[Substance identification process]
Next, a substance identification process executed interactively with the operator by the data processor 35 shown by the block B3 in FIG. 9 will be described.

For this substance identification process,
Creation of a panoramic image (step S31),
Creating a panorama 3D image (step S32);
Removal / reduction of obstacle shadow (step S33),
X-ray direction setting (step S34),
CT image reconstruction (step S35),
3D display of CT image (step S36),
Material thickness measurement (step S37),
Calculating the absorption coefficient of the substance (step S38);
Creating a scatter diagram and identifying the type of substance (step S39), and
Identification result display (step S40),
Is included.

Note that the step group from one step S31 to S34 may be executed in this order, and the other step group from step S35 to S36 may be executed in this order. Either group may be processed first. The data processor 35 executes the processes of steps S37 to S40 when the processing of both groups is completed.

Hereinafter, the processing of these steps will be described separately.
[Panorama image creation (step S31)]
The data processor 35 reads out panoramic image frame data for each X-ray irradiation angle (projection angle) stored in the first storage unit 34, applies a tomosynthesis method to the frame data, and applies the 3D reference tomographic plane SS. Is reconstructed for each energy region ER _n (n = 1 to 3). Note that the frame data read out here are frames collected at regular intervals (for example, 75 fps) from the module Bm that forms the collection window 22B for the left vertical column shown in FIG.

In this embodiment, a panoramic image is created for each of the _three energy regions ER ₁ to ER ₃ shown in FIG. 7, but the panoramic image may be reconstructed for two or more energy regions ER _n. .

The tomosynthesis method accompanying this reconstruction, that is, shift-and-add processing, is executed in accordance with International Publication No. WO2012 / 008492. Since the calibrated gain of the imaging space IS is held in advance, the actual position of the dentition is reflected by shifting the frame data by the shift amount along this gain and adding the pixel values to each other. A pseudo three-dimensional panoramic image PI _focus is automatically generated (see FIG. 10). This panoramic image PI _focus is called a 3D autofocus image.

[Creation of Panorama 3D Image (Step S32)]
Next, the data processor 35 reads correction data based on the prepared tough bone from the first storage unit 34 for each energy region ERn, and corrects the panoramic image PI _focus for each energy region ER _n using this correction data. . As a result, standardized panoramic 3D images (3D autofocus images) are generated for the three energy regions ER _n (see FIGS. 10 and 11).

[Removal / Reduction of Obstacle Shadow (Step S33)]
The data processor 35 executes the removal / reduction processing of the obstacle shadow on the panorama 3D image of each of the three energy regions ER _n generated in this way. As a result, the obtained panorama 3D image is displayed on the display 36.

In this process, as schematically shown in FIG. 12, if the dentition TR in the jaw JW is an object to be imaged, the cervical vertebra CS is an object that obstructs imaging. In the case of panoramic photography, the shadow of the cervical spine is reflected in the panoramic image. This reflection is inevitable in a specific X-ray irradiation angle range. Therefore, when this dentition (gum) is to be imaged, it is desirable to remove / reduce the obstruction shadow caused by the cervical spine from the created panoramic 3D image. This removal / reduction method will be described later in detail as a separate item.

[X-ray direction setting (step S34)]
Next, the data processor 35 designates the position of interest P _int on the currently displayed panoramic 3D image through interactive information exchange with an operator (such as a dentist). For example, it is assumed that the operator designates the position P _int while observing a panoramic 3D image and wants to know the type of the substance (structure of the jaw). This designation is also performed, for example, when it is desired to know the type of metal object used as a tooth filling.

In response to this, the data processor 35 designates “L−1” positions P _add in the vicinity of the position of interest P _int . Thereby, a total of L positions P _int and P _add are set. These L-number of points are required to solve the communicating columns equation according to _L C _N-number of combinations to be described later required when calculating the X-ray absorption coefficient of a substance having a thickness. Here, N is the number of substances having a thickness that exist in the X X-ray irradiation directions corresponding to “L−1” positions P _add in the vicinity of the position of interest P _int . Here, “near” means within a range of a substance considered to be the same as the substance indicated by the position of interest.

For this reason, the operator only needs to observe in advance a CT image, which will be described later, and know how many kinds of substances have a thickness along the L X-ray irradiation directions in the three-dimensional CT image. Since the number of types = N, in order to obtain a solution from this equation, L positions satisfying L ≧ N are set automatically or interactively.

For example, as shown in FIG. 13, if there are substances having three types of thicknesses in the X-ray direction (virtual X-ray beam direction) projected to the designated position of interest _Pint , N = 3. For this reason, for example, the data processor 35 adds two or more points separated by 0.5 mm (default value) at equal distances around the position of interest P _int and adds a total of three or more points (= L). It may be set. Further, the data processor 35 may rely on an instruction from the operator for the position and number of the additional points.

Since it is necessary to set the L positions so as not to be separated from each other and to pass through different substances, it is also possible to add processing for confirming with the operator.

When the L points are designated in this way, the data processor 35 is processed in the block B22, and the geometric structure in the imaging space IS stored in advance in the first storage unit 34 is obtained. The analysis data shown is read out, and based on this analysis data, the three-dimensional X-ray directions (virtual three-dimensional X-ray beam directions) projected onto the total L positions P _int and P _add are calculated. To do. The three-dimensional X-ray direction includes a direction in which a direction sandwiching each of the positions P _int and P _add on the panorama 3D image from the X-ray focal point FP of the X-ray tube 21 is projected on the XY plane, and the XY plane in that direction Is calculated based on the inclination angle with respect to. Information on the direction and angle is acquired from the analysis data and used for calculation.

The data indicating the three-dimensional direction of the X-rays projected on the L positions (points) thus obtained is temporarily stored in the first storage unit 34.

[Reconstruction of CT Image (Step S35)]
On the other hand, the data processor 35 reads the CT image frame data stored in the first storage unit 34. This frame data is a frame collected at regular time intervals (for example, 75 fps) from the module Bm forming the collection window 22A shown in FIG. The data processor 35 adds the frame data of the three energy regions ERn collected for each X-ray irradiation angle, that is, each projection angle, to each other to generate projection data. Further, a plurality of sets of projection data generated for a plurality of projection angles for these half scans are subjected to CT reconstruction based on, for example, a successive approximation method. Thereby, a three-dimensional CT image of the jaw JW is generated.

Here, if there is a metal padding on the teeth, that portion will be blurred due to the influence of metal artifacts, making interpretation and measurement described later difficult. For this reason, when a metal artifact has occurred, it is desirable to re-perform CT reconstruction with a procedure for removing the artifact. Hereinafter, a CT reconstruction method with removal of this metal artifact will be described.

<Removal of metal artifact>
Here, it is assumed that a metal part exists in one tomographic plane, and a process for removing the artifact of the part will be described. An example of this processing is shown in FIG. The processing from steps S1 to S11 shown in FIG.

Step S1 (creation of sinogram):
In step S1, a sinogram is created based on the value of projection data P _init from 0 to 360 degrees (see FIG. 15). The horizontal axis of the sinogram is the detector position, and the vertical axis is the projection angle. The sinogram is a logarithmic operation by dividing between the number of X-rays actually emitted and the value measured by the detector.

Step S2 (detection of metal part):
Next, in step S2, the value of the projection data _Pinit is scanned in the direction of the horizontal axis of the sinogram, and the value of the portion that greatly changes from the value of the nearby projection data is detected (see FIG. 16). At this time, the standard deviation with respect to the average value of the projection data may be used as a scale for determining whether or not the metal portion is greatly attenuated, that is, whether it is a metal part, or analysis of variance from a one-dimensional ROI set in the vicinity. Thus, information on the sudden rise and fall of the projection data of the portion approaching the metal portion may be used.

In this embodiment, assuming that the position of the projection data of the metal part can be completely specified in the process of step S2, the processes after step S3 are executed. If the metal portion cannot be identified from the one-dimensional ROI, a process corresponding to step S2 may be performed using a sinogram (two-dimensional). Even in this case, if the metal part cannot be specified in all projection data, the area on the projection data of the metal part becomes a sine curve, and the part that failed to extract the metal part is used. It is also possible to estimate. Furthermore, even if not all metal parts can be necessarily extracted, it is possible to estimate in the processing after step S3.

Step S3 (image reconstruction):
In step S3, the projection data of the metal part extracted in step S2 is set to 0, and 1 is assigned to other projection data areas, and this operation is performed for all projection angles. This projection data is reconstructed into an image by simple back projection calculation (see FIG. 17). Here, if the metal portion of the projection data cannot be specified in step S2, back projection calculation is performed using only the projection data excluding the angle. In the image created in this way, only the metal portion is 0, and the other images have values other than 0. In this way, an appropriate value, for example, 1 (1 / cm) is substituted for the 0 pixel of this simple backprojection image, and the other pixel values are set to 0. This is the basic image I _metal of the metal part.

Since the basic image I _metal corresponds to the metal part, as described above, even if the metal part of the projection data is not completely detected in the process of step S2, the data of the basic image base I _metal is used as the metal part. Function as. Using this data, it is clear from which position of the specific projection data to which position is the metal part.

Step S4 (calculation of projection data):
Next, in step S4, projection data P _{metal of} only the metal portion of the basic image I _metal created in step S3 is calculated (see FIG. 18). Further, to the projection data other than the metal portion (P _metal ≠ 0) of the projection data P _init is called a P _org (see FIG. 19).

Step S5 (image reconstruction):
In step S5, the projection data P _org other than the _metal part is reconstructed by the successive approximation method (ML-EM, OS-EM, etc.) for the part where P _metal ≠ 0 is P _org = 0. (See FIG. 20). Among the reconstructed images, an image in a region where I _metal ≠ 0 is referred to as I _recon .

Step S6 (image addition):
In step S6, the pixel value of the region where I _metal = 1 in the image I _recon is set to I _recon = 0, and then both the images I _recon and I _metal are added to each other for each pixel. As a result, I _all is created as an added image (see FIG. 21).

Step S7 (projection processing):
Further, in step S7, the added image I _all is subjected to projection processing, and projection data P _all is created (see FIG. 22).

Step S8 (difference calculation and correction):
Furthermore, in step S8, it is compared for each pixel of the projection data P _org except projection data P _all the metal parts of the sum image I _all. That is, the difference between the two is calculated, and the difference data is corrected using a successive approximation reconstruction method (for example, an arithmetic image reconstruction method such as ART) (see FIG. 22). At this time, only the reconstruction area of P _metal = 0 is corrected, and projection data and the like in the metal portion are not processed. This is processed for all projection angles, and pixel values other than the metal portion are corrected. In this way, an image I _{recon (n)} is created. Here, n = 1.

Step S9 (image addition, projection calculation):
In step S9, the value of the region of I _metal = 1 in the image I _{recon (1)} is set to I _{recon (1)} = 0, and both images I _{recon (1)} and I _metal are added to each other for each image. Is done. Thereby, the addition image I _{all (1)} is created. This added image I _{all (1)} is subjected to projection processing to create projection data P _{all (1)} (see FIG. 23).

Step S10 (difference calculation and correction):
Next, in step S10, as in step S8, both projection data P _{all (1)} and P _org are compared (differed) for each pixel and corrected. That is, the image I _{recon (1)} is corrected to create I _{recon (n + 1)}

Step S11 (convergence determination):
Next, in step S11, it is determined whether or not the error between the projection P _{all (n)} (where P _metal = 0) and the projection data P _org has become a certain value or less, and the convergence of the calculation is determined. . If this convergence determination condition is satisfied, the process is returned to the main process. When the convergence determination condition is satisfied, the processes after step S9 are repeatedly executed.

As described above, in CT imaging, a metal part is recognized with projection data, and the recognized projection data of the metal part is reconstructed by filling a fixed value to remove or suppress metal artifacts from the metal part. CT images obtained are obtained. Thereby, the three-dimensional shape of the metal part can be measured more accurately.

Metal artifacts are reduced in images created with the convergence criteria (in some cases exceeding the maximum number of comparisons), and a preset line attenuation coefficient is given to the metal part. The portion is given a value estimated from correctly measured projection data.

The above-described metal artifact reduction operation may be performed in a successive approximation image reconstruction simultaneously with the beam hardening removal process of CT images and scattered radiation removal.

[Three-dimensional display of CT image (step S36)]
Next, the data processor 35 three-dimensionally displays the generated three-dimensional CT image of the jaw JW on the display 36. As a result, an operator (such as a dentist) can observe the form of the jaw JW three-dimensionally, how many substances have thickness in which X-ray irradiation direction, which substance and which substance. It is possible to estimate the state of the same type.

[Measurement of material thickness (step S37)]
Here, the data processor 35 automatically or interactively interacts with the operator along the direction of the substance existing along each of the total L X-ray directions including the total L points P _int and P _add . Thicknesses t _i1 to t _iN (i = 1 to M: M is the number of energy regions ER _n : N is the number of thick materials) are measured. This thickness is measured, for example, by an operator specifying a bridge position of each substance in each X-ray direction on the 3D CT image with a point ROI and calculating this specified position from the analysis data (block B22). As a result, the thicknesses of N (L ≧ N) substances in each of a plurality of X-ray directions of interest are measured.

[Calculate the absorption coefficient of the substance (step S38)]
When the preparation is completed in this manner, the data processor 35 converts the 3D panoramic image data from which the obstacle shadow has been removed (or reduced) in step S33 and the material thicknesses t _i1 to t _iN obtained in step S37. Based on the X-ray absorption coefficient μ _1j , μ _2j ,..., Μ _Mj (M is the number of energy regions: j = 1 to N: N is the number of materials) of _L C _{N for} each energy region ER _n. Calculate from the equation.

As an example, as shown in FIG. 13, the total number of observed materials in the X-ray direction is 3, and their thicknesses are t ₁ , t ₂ , t ₃ , and their X When the linear absorption coefficient is μ _1j , μ _2j , μ _3j (j = 1 to 3),
Logarithmic value of pixel value (measured value) at a position on the 3D panoramic image in the energy region ER ₁ = μ ₁₁ * t ₁ + μ ₁₂ * t ₂ + μ ₁₃ * t ₃ + k ₁ (k ₁ : constant),
Logarithmic value of a pixel value (measured value) at a position on the 3D panoramic image in the energy region ER ₂ = μ ₂₁ * t ₁ + μ ₂₂ * t ₂ + μ ₂₃ * t ₃ + k ₂ (k ₂ : constant),
Logarithmic value of a pixel value (measured value) at a position on the 3D panoramic image in the energy region ER ₃ = μ ₃₁ * t ₁ + μ ₃₂ * t ₂ + μ ₃₃ * t ₃ + k ₃ (k ₃ : constant),
You can solve these simultaneous equations. These equations hold for the total L of the position P _int to be observed and the position P _{add in} the vicinity thereof. In this example, L = 3. For this reason, since nine equations are established for nine variables μ _1j , μ _2j , μ _3j (j = 1 to 3 (N = 3)), an arithmetic method such as singular value decomposition of the simultaneous equations is used. A total of nine X-ray absorption coefficients μ _Mj are obtained for the solution, that is, three for each energy region. In the case of L> N is, _L C _N-number of only melts simultaneous equations combined, _L C inclusive solution the statistical error of the _N kinds are obtained, the type of the structure of the jaw (substances with a thickness) Used for identification.

[Creation of scatter diagram and identification of substance type (step S39)]
Next, the data processor 35 creates a two-dimensional scatter diagram for each substance from the solution of the X-ray absorption coefficient μ _Mj obtained in step S38. This two-dimensional scatter diagram includes a relative attenuation index RAI (Relative Attenuation Index) shown on the vertical axis forming one dimension and a quality change index SDI (Spectrum Deformation Index) shown on the horizontal axis forming another dimension. Of course, this scatter diagram only needs to include the relative attenuation index RAI and the quality change index SDI, and may be a three-dimensional or four-dimensional or more scatter diagram combining other indexes.

Among these, the relative attenuation index RAI is an index corresponding to the CT value of each pixel forming the three-dimensional volume space, and for each substance, the relative attenuation index RAI for each pixel is equal to μ _1j + μ _2j + _,. _Mj (M: number of energy regions, j = 1 to N, N is the number of substances). On the other hand, the line quality change index SDI is an index indicating a linear absorption coefficient specific to a substance, and the line quality change index SDI is, for example, μ _3j − for each substance and for each pixel in the three-dimensional volume space. Defined as μ _1j . It may be defined as SDI = μ _2j −μ _1j .

An example of this two-dimensional scatter diagram is shown in FIG.

After this, the data processor 35 reads a reference scatter diagram (see FIG. 25) from the database of the first storage unit 34 prepared in advance (block B23), and compares it with the scatter diagram generated in step S39. . Thereby, it is possible to identify the type of substance present at the position of interest P _int designated on the panorama 3D image by the operator. That is, depending on which plot group on the scatter diagram shown in FIG. 24 matches or is closest to one of the plot groups in the reference scatter diagram shown in FIG. The type of a certain substance can be determined. The scatter plot of FIG. 24 is identified as plot group A = alveolar bone and plot group B = metal, assuming the identification of substances along the X-ray direction passing through the implant illustrated in FIG. .

Although not shown in particular, when there are different types of metals, the map position on the scatter diagram differs depending on the types of the metals, and thus different types of metals can be distinguished and identified. It should be noted that in the discrimination between metals, there may be a difference in discrimination accuracy depending on the X-ray transmission characteristics of the metals.

[Display of identification result (step S40)]
The result thus identified is displayed on the display 36 by the data processor 35, for example.

FIG. 26 shows another scatter diagram for reference. According to this, it is also possible to provide information for diagnosing the presence or absence of the influence of a medical condition such as whether a substance such as bone has osteoporosis or periodontal disease.

As described above, if a database is created in advance in order to perform a specific distribution depending on the observation site, the site-specific distribution such as the type of metal, diagnosis of osteoporosis, and cancellous bone that has been solved by periodontal disease can be obtained. Draw and identify substances. There are only _L _CN combinations for the number of data of each part, but even one point can be estimated. In CT for medical use, the substance is identified only by the RAI on the vertical axis based on water (so-called CT value), but in the photon counting detector that can add a SDI axis and make a two-dimensional scatter plot, The great advantage is that the differences can be emphasized and expressed. Thereby, identification accuracy improves.

In the example of the scatter diagram of FIG. 25, it is unnatural that gold (prosthesis) is distributed on the left side of the cortical bone on the SDI axis. However, in the case of gold, the energy band is the lowest. This is considered to be because the count of the region is extremely reduced and μ in this energy band is calculated to be large. However, although there is such a unique phenomenon, from the viewpoint of substance identification, since it is distributed at a specific position, the database is not affected by creating a database.

<How to remove obstacle shadows>
Here, the principle of the method of removing the obstacle shadow from the panoramic image executed in the present embodiment will be described.

The tomosynthesis image g (i) (i represents a position perpendicular to the X-ray beam) is an object f _i (i) (j = 1, 2) divided into n layers in the X-ray beam direction.
..., n) and the degradation function h _i (i) (j = 1,2) generated by the overlay operation at that position
..., n) is expressed in the form of a sum.

”Here, all functions are considered to be logarithmic. That is, the tomosynthesis image g (i) is expressed as follows. * Is a symbol indicating convolution.

g (i) = f ₁ (i) * h ₁ (i) + f ₂ (i) * h ₂ (i) +… + f _j (i) * h _j (i) +… + f _n (i) * h _n (i)
In the tomosynthesis process, a specific layer g _j (i) is imaged by making h _j (i) of a specific layer a delta function and making the others as uniform functions as possible.

Now, the model is simplified, and a three-layer model composed of two specific layers f ₁ (i), f ₂ (i) and the other f _r (i) is considered as follows.

g (i) = f ₁ (i) * h ₁ (i) + f ₂ (i) * h ₂ (i) + f _r (i) * h _r (i)
Here, when focusing on the second layer, the following equation is established.

h ₂ (i) = δ (i)
g ₂ (i) = f ₁ (i) * h ₁ (i) + f ₂ (i) * δ (i) + f _r (i) * h _r (i)
= f ₁ (i) * h ₁ (i) + f ₂ (i) + f _r (i) * h _r (i)

… (1)
Further, when focusing on the first layer, the following formula is established.

h ₁ (i) = δ (i)
g ₁ (i) = f ₁ (i) * δ (i) + f ₂ (i) * h ₂ (i) + f _r (i) * h _r (i)
= f ₁ (i) + f ₂ (i) * h ₂ (i) + f _r (i) * h _r (i)

… (2)
If the component of f _r (i) * h _r (i) is very small, the following equation is derived from equations (1) and (2).

g ₂ (i) ＝ f ₁ (i) * h ₁ (i) + f ₂ (i)… (3)
g ₁ (i) = f ₁ (i) + f ₂ (i) * h ₂ (i)… (4)
Here, g ₁ (i) is an image focused on the dentition, and g ₂ (i) is an image focused on the cervical vertebra, and the true dentition image f ₁ (i) after removing the shadow is f ₁ ( i) = (g ₁ (i) -g ₂ (i) * h ₂ (i)) / (1-h ₁ (i) * h ₂ (i))… (5)
As given.

The meaning of equation (5) is
(Image focused on dentition)-
(Image focused on cervical spine * blurring function at dentition position)
This means that an image of only the dentition can be obtained by dividing by the superposition integration of the blur functions corresponding to the positions.

Even if the components of f _r (i) * h _r (i) are not small, the above calculation is valid if the bias component is subtracted from g ₁ (i) and g ₂ (i) as certain parameter values. Therefore, it is possible to remove or reduce the obstacle shadow.

[Specific flow of removal of obstacle shadow]
Next, an example of processing for removing and reducing obstacle shadows executed in accordance with the above-described principle in the X-ray imaging apparatus 1 will be specifically described with reference to FIGS. Here, it is assumed that the obstacle shadow is a shadow of the cervical vertebra reflected in the panoramic image of the dentition of the subject.

First, frame data relating to panoramic imaging of the jaw of the subject P is read (FIG. 27, step S51). This frame data is stored in the first storage unit 34. For example, as shown in FIG. 28A, this data collection is performed so that the reference tomographic plane (cross section) preset in the dentition TR is focused. Symbol CS shown in FIG. 28A indicates the cervical spine.

Next, the data processor 35 reconstructs a panoramic image focused on the reference tomographic plane (see FIG. 28A) of the dentition TR using the frame data stored in the first storage unit 34. Displayed (step S52). An example of this panoramic image is shown in FIG. The reconstructed panoramic image data is stored in the first storage unit 34.

In step S52, specifically, the image of the optimum focus on the reference tomographic plane is reconstructed by the tomosynthesis method, that is, shift-and-add processing. The second ROM 40B stores gain curve information for optimally focusing a reference tomographic plane of a predetermined dentition TR. An example of this gain curve is shown in FIG. The gain curve is a curve that indicates an amount (differential value of the curve) by which each piece of strip-shaped frame data (frame image) is shifted (shifted) with respect to each other in the shift-and-add process.

The shift amount in this gain curve corresponds to the blur amount when creating a blur function at the dentition position. Therefore, the data processor 35 uses the shape of the blur as a Gaussian function, matches its standard deviation to this shift amount, and convolves with the frame data as a shift variant blur function that changes for each angular position. Thereafter, the data processor 35 adds frame data to each other at each position (mutual addition of pixel values) to create a panoramic image of the reference tomographic plane along the tooth row TR.

Next, as described above, the data processor 35 uses the frame data stored in the first storage unit 34 to generate a panoramic image focused on a tomographic plane (see FIG. 28A) passing through the cervical vertebra CS. Reconstructed and displayed (step S53). An example of this panoramic image is shown in FIG. The reconstructed panoramic image data is also stored in the first storage unit 34.

The cervical vertebrae CS focused on the image creates a gain curve assuming a trajectory T _CS folded back trajectory T _TR focus on teeth surfaces. At this time, a gain curve is created depending on the position of the wire (lead phantom) of the calibration phantom. In this case, the gain curve is created in the form of inverting the above-described gain curve (see FIG. 30). Based on the amount of deviation indicated by the gain curve (that is, the differential value of the curve), a shift-and-add operation is performed. In this case, generally, in the reconstruction of the dentition TR, the shift amount in the vicinity of the front teeth is small, but in the reconstruction of the cervical vertebra CS, the collected frame data is moved greatly to perform the addition operation.

The frame data used for panoramic image reconstruction in the above steps S52 and S53 may be frame data having energy belonging to all energy regions ER ₁ to ER ₃ , or may be included in a part of the energy region. It may be frame data of the energy to which it belongs. Even when using frame data of energy over all energy regions, in addition to simple averaging, weighting can be performed for each region.

Since the detector 22 is a photon counting type, the frame data can be detected in a state where energy is discriminated for each pixel. This makes it possible to use transmission data (frame data) that sensitively reflects parameters such as a change in radiation quality when X-ray photons pass through the jaw material. For this reason, a panoramic image in which a specific substance in the jaw is emphasized can be acquired, and the obstacle shadow removal method can be performed based on the panoramic image.

Next, the data processor 35 performs a scaling process (step S54). Figure 31 shows the best focus image I _TR along the reference tomographic plane of the dentition TR with the vertical line. This vertical line shows the position of the specific angle measured with the calibration phantom. Further, FIG. 32 shows the best focus image I _CS along the fault plane cervical CS with the vertical line. This vertical line also indicates the position of a specific angle measured with the calibration phantom.

In order to remove the obstacle shadow such as the cervical spine, FIGS. 31 and 32 must be scaled to have the same image size. The distance between the vertical lines in FIG. 31 and the distance between the vertical lines in FIG. 32 are different for each position, and the size of FIG. 32 cannot be matched to the size of FIG. 31 using one reduction ratio. Therefore, the scales of the two images are adjusted using the frame defined by the vertical lines. In order to easily match, it is only necessary to match each area surrounded by two vertical lines so as to have the same size. Further, in order to match with higher accuracy, it is desirable to match the corresponding positions for each strip-shaped frame data.

Here, the plurality of rectangular regions defined by vertical lines 15 represent components of the optimum focus image I _TR dentition TR, 15 present those regions by the vertical line of best focus image I _CS cervical CS Each area of the image _ICS is reduced so as to match. Thus, the size of both images _{I TR} and _{I CS} are the same.

Next, the data processor 35 executes a blurring process of the optimally focused image I _CS of the cervical vertebra CS (step S55). In order to determine the blur function in this operation, a gain curve is used. More specifically, the full width at half maximum of the gain value at each angular position of the gain curve: create a Gaussian function with a (FWHM full width at half maximum) , which is convolution optimally focused image I _CS cervical CS.

In the cervical vertebra blurred image I _CS ′ (not shown) thus formed, the dentition is greatly blurred, whereas the cervical vertebra is in contrast to the cervical vertebra in the image focused on the dentition surface. The blur is about the same.

Note that this blurring process, that is, the process of step S5 may be omitted depending on the situation.

After this blurring process, the data processor 35 performs, for each pixel, between the optimal focus image I _TR of the dentition _TR and the optimal focus image I _CS ′ of the cervical vertebra CS reduced and blurred as described above. The pixel value is subtracted or divided to perform a process of taking the image difference (step S56). When pixel values are expressed in natural logarithm, subtraction is performed, and “I _TR -I _CS ′” is performed between pixel values. On the other hand, when the natural logarithm of the pixel value is not taken, the division of “I _TR / I _CS ′” is performed between the pixel values. Further, since the density unevenness depending on the blur function occurs in the pixel value, the blur function on the cervical vertebra focal plane and the blur function on the dentition surface are superimposed and integrated, and a value obtained by subtracting this from 1 is used. By dividing each angular position, it is possible to obtain an image in which the influence of the cervical spine is reduced as much as possible and the dentition surface is focused. It is desirable to add this process to step S56.

Since the difference image by the difference or the division becomes the optimum focus image I _{TR_REV} of the dentition TR after removing (or reducing) the shadow of the cervical spine, this is displayed on the display 36 and, for example, the first image (Step S57).

Figure 33 shows another example of the best focus image I _TR dentition TR before cervical removal. In this image _ITR , a white shadow (disturbance shadow) of the cervical vertebra CS is reflected in the central portion thereof, the vicinity of the front teeth is blurred, and the drawing ability is low. On the other hand, as shown in FIG. 34, in the optimally focused image I _{TR_REV} of the dentition TR after cervical vertebra removal, the cervical vertebra image is almost removed, and the anterior teeth are more clearly depicted.

Thus, a relatively simple correction after data acquisition, shading cervical CS that interfere shadow is reliably removed or alleviated from the panoramic image I _TR dentition TR. For this reason, the drawing ability of the dentition TR is improved.

As described above, the hybrid X-ray imaging apparatus 1 according to this embodiment can function as an all-in-one X-ray diagnostic apparatus with a dental X-ray imaging apparatus capable of material identification. Substance identification is possible even with CT imaging in a small field of view. Even when there is a metal prosthesis, the substance can be identified stably. This is because the thickness of the metal part can be measured more accurately from a CT image in which metal artifacts due to the metal part are removed or reduced. Since panoramic imaging and CT imaging can be performed simultaneously in a circular orbit in one scan, the amount of X-ray exposure to the patient is small.

In this embodiment, the photon counting technique is used for the detector, and images from soft tissue to hard tissue can be visualized. A successive approximation CT reconstruction technique is employed, and a CT image with a clear tissue boundary can be obtained with a small field detector. Moreover, since the obstacle shadow of the panoramic image is deleted or reduced, the substance identification can be performed in combination with the panoramic image.

Furthermore, since panoramic imaging and CT imaging can be performed simultaneously in one scan, it is not necessary to align the reconstructed panoramic image and CT image.

From the above, one hybrid X-ray imaging apparatus according to the present invention can play two roles of a conventional panoramic imaging apparatus and CT apparatus. That is, highly reliable diagnosis can be performed using only the hybrid X-ray imaging apparatus 1, and treatment time can be shortened, treatment accuracy can be improved, and throughput can be improved. In addition, the peri-implantitis is observed, the positional relationship between the soft tissue tumor and the hard tissue is observed, and the clinical application range is dramatically expanded.

In FIG. 9 showing functional blocks, block B1 corresponds to the scanning means, step S31 of block B3 corresponds to the panoramic image generating means, step S35 corresponds to the CT image generating means, and step S32 corresponds to the panoramic image display. Step S34 corresponds to the position of interest designation means, and steps S36 to S40 correspond to the substance identification means.

[Other embodiments]
In the above-described embodiment, panoramic imaging and CT imaging are completed by one scan at the same time, but the X-ray diagnostic apparatus according to the present invention is not necessarily limited to this. For example, panoramic imaging and CT imaging can be separately executed twice, the reconstructed panoramic image and CT image can be aligned, and the same processing as described above can be performed. In this case, the detector used in the CT imaging apparatus may be a conventional integral X-ray detector, and it is not always necessary to use a photon counting X-ray detector.

The X-ray imaging apparatus 1 described above may be configured to take an image in a state where the patient is lying on his / her back on the dental chair (supposed position). What is necessary is just to have an imaging system that rotates with the jaw of the subject positioned between the X-ray tube and the detector. It may be a device configuration that receives an image in a sitting position or standing position. Further, such a photographing system may be attached to a fixed structure such as a house or a vehicle wall or ceiling. Furthermore, such an imaging system may be configured as a portable unit, and may be configured to perform imaging by placing on a patient's shoulder or behind a general chair.

Further, the X-ray imaging apparatus 1 described above includes an imaging system in which the X-ray tube 21 and the detector 22 can be driven to rotate independently of each other while varying the distance between them around the same rotation center. Good.

Furthermore, in the embodiment of the present application, the detector 22 is a photon counting type detector. However, the scintillator and the photoelectric element are combined to accumulate electrical signals for a predetermined time and output frame data. A so-called integral type detector may be used.

Furthermore, the obstacle shadow removal method of the present application can be applied to an X-ray imaging apparatus using a detector using a scintillator and a CCD (charge coupled device), to which the tomosynthesis method cannot be applied. In that case, an image taken under the rotational trajectory of the X-ray tube and detector focused on the cross-section passing through the dentition, and the rotational trajectory of the X-ray tube and detector focused on the cross-section passing through the cervical spine It is sufficient to prepare images taken under the image separately and perform the above-described obstacle shadow removal method between these images.

According to the present invention, the position, distance, and angle parameters of the X-ray tube, 3D reference tomographic plane, and detector that define the structure of the imaging space can be easily and accurately analyzed by measurement using a phantom. Can be calibrated and prepared for shooting. Therefore, it is possible to provide an imaging apparatus using radiation that can accurately image an object three-dimensionally.

1 Dental X-ray radiography system (functionally forms a substance identification system)
16 Rotating unit 17 Console 21 X-ray tube 22 Detector 23 Slit 33 Controller (one of the elements for functionally realizing various means)
34 1st memory | storage part 35 Data processor (one of the elements which implement | achieves various means functionally)
36 Display 37 Input device 40A to 40D ROM
C semiconductor cells Cp detecting circuit _{S n} pixels

Claims

An X-ray tube that emits X-rays;
A detection circuit in which cells each outputting an electric pulse corresponding to the energy of the photon each time the X-ray photon is detected are two-dimensionally arranged so as to form a two-dimensional pixel group; and A measurement circuit that measures the number of photons detected by each cell into two or more energy bands (number of energy bands M ≧ 2: M is a positive integer) for each pixel, and each cell measured by the measurement circuit A detector configured to include an output circuit that outputs a digital amount of an electric signal corresponding to the output of each of the energy bands as frame data;
The X-ray tube and the detector are opposed to each other with the imaging target interposed therebetween, the X-ray tube and the detector are rotated around the imaging target, and the frame data output by the detector is converted into the energy. Scanning means for collecting each band;
Panorama image generating means for generating data of the panorama image to be photographed for each energy band based on the frame data;
CT image generation means for generating a CT (computed tomography) image of the imaging target;
Panoramic image display means for displaying the panoramic image on a monitor;
Interest position specifying means for specifying a position of interest on the panoramic image displayed on the monitor in accordance with an operator's instruction;
X-rays of the panoramic image represent at least one type of one or a plurality of substances having a thickness in the imaging target that exists in the projection direction in the imaging space between the X-ray tube and the detector corresponding to the position of interest. A substance identifying means for identifying based on transmission data and morphological information of the substance obtained from the CT image;
A substance identification apparatus using X-ray panorama / CT imaging characterized by comprising:
The substance identification means includes
CT image presentation means for presenting the CT image generated by the CT image generation means to an operator;
From the CT image, there are N substances (N is a positive integer of 1 or more) as one or more substances having the thickness along the projection direction interactively with the operator. Determining means for determining
The total L projected from the positions of L points (L ≧ N) including the position in the vicinity of the position of interest including the projection direction in the shooting space corresponding to the position of interest along the projection direction to the shooting space. A projection direction identifying means for identifying projection directions in the vicinity of each other;
The thicknesses t i1 , t i2 ,..., T iN (i = 1 to M;) of the N substances to be imaged that exist along the passing directions of the L projection directions in the CT image data. N is the thickness calculation means for calculating the number of substances) from the CT image data as the morphology information;
When the absorption coefficients of the N substances for the X-rays for each energy band are μ 1j , μ 2j ,..., Μ Mj (j = 1 to N), the thicknesses of the substances are known and L C N-number of the absorption coefficient by solving the simultaneous equations mu 1j having an absorption coefficient as a variable, μ 2j, ..., and the absorption coefficient calculating means for calculating a value of mu Mj,
For each of the N substances, the vertical axis represents μ 1j + μ 2j ,..., + Μ Mj (j = 1 to N) as a relative attenuation index: RAI, and the horizontal axis represents μ aj −μ bj (a, a positive integer satisfying b ≦ M, where a> b) is a quality change index: a scatter diagram creating means for creating a two-dimensional scatter diagram taken as SDI;
The scatter diagram for each substance created by the scatter diagram creation means is compared with the reference scatter diagram showing the relative attenuation index and the quality change index, which are preset for each substance, and the interest. Substance type determination means for determining the type of each of the one or more substances present in the projection direction according to the position;
The substance identification apparatus according to claim 1, comprising:
The display device according to claim 2, further comprising a presentation unit that presents to the user the types of the one or more substances existing in the projection direction according to the position of interest determined by the substance type determination unit. The substance identification apparatus described.
4. The substance identification apparatus according to claim 1, wherein the two or more energy bands are three or more energy bands.
The CT image generation means includes
Projection data generation means for generating projection data by adding frame data for each projection angle collected for each energy band;
Reconstructing means for reconstructing three-dimensional data of the CT image to be imaged based on projection data for each projection angle;
The substance identification apparatus according to any one of claims 1 to 4, further comprising:
The projection data generating means is a half-scan type that uses frame data collected during a rotation of the X-ray tube rotated about the object to be imaged by the scanning means and about a half turn of the detector. The substance identification apparatus according to claim 5, wherein the substance identification apparatus is configured to generate the projection data below.
The scanning means is configured to rotate on the same circular orbit with the X-ray tube and the detector always facing each other,
The CT image generation means is configured to generate the CT image based on the frame data output from the detector.
The panoramic image generation means and the CT image generation means use the frame data output from the detector when the X-ray tube and the detector are simultaneously rotated once by the scanning means. Configured to generate an image and a CT image respectively;
The substance identification device according to any one of claims 1 to 6.
The detection circuit includes a semiconductor layer formed of a semiconductor material that directly converts the X-rays into the electric pulse and divided for each pixel, a charged electrode stacked on one surface of the semiconductor layer, and the semiconductor A current collecting electrode stacked on the other surface of the layer and divided for each pixel,
The measurement circuit and the output circuit are built as an ASIC (Application Specific Integrated Circuit) layer as a layer integrated with the detection circuit,
The semiconductor material is CdTe, CZT, or TlBr.
The X substance identification device according to any one of claims 1 to 7, characterized in that:
The subject is the subject's jaw,
The panoramic image generation means automatically focuses on the position of each tooth of the dentition existing in the jaw based on the frame data and reflects the actual position and shape for each energy band. 9. The substance identification apparatus according to claim 1, wherein the substance identification apparatus is configured to create a pseudo three-dimensional panoramic image obtained by three-dimensionally developing a three-dimensional panoramic image.
The panoramic image generating means
First panoramic image generation means for generating a first panoramic image focused on the dentition from the frame data for each energy band, the dentition being an object;
A second panoramic image generation means for generating a second panoramic image focusing on the cervical vertebra for each energy band from the cervical vertebra existing on the back side of the jaw as an obstacle;
Based on the first and second panoramic images, obstacle shadow removal / reduction means for removing or reducing the reflection of the shadow of the cervical vertebra from the first panoramic image;
The substance identification device according to any one of claims 1 to 9, further comprising:
A phantom that is arranged in the imaging space and has a marker that is positioned on a predetermined tomographic plane in the imaging space and can image known position information with the X-ray is arranged in the imaging space. An image creating means for creating an image based on the frame data collected according to the X-rays irradiated from the X-ray tube in a state;
Based on the known position information of the marker and the position information of the marker obtained from the image, distance information between the X-ray tube and the detector and height information of the X-ray tube with respect to the detector First computing means for computing
Second computing means for computing parameters defining the positional relationship of the X-ray tube, the detector, and the tomographic plane in the imaging space based on the computation result of the first computing means and the data;
Storage means for storing parameters calculated by the second calculation means as calibration data,
The projection direction specifying means is configured to specify a total of M projection directions that are projected in the imaging space, using the calibration data.
The substance identification device according to any one of claims 2 to 10.
The CT image generation means includes
An artifact mitigation means for mitigating metal artifacts of the metal when including metal as part of the one or more substances in the projection direction corresponding to the position of interest;
Means for generating the CT image again in a state where the artifact is reduced by the artifact reduction means;
The substance identification device according to any one of claims 1 to 11, further comprising:
An X-ray tube that emits X-rays;
A detection circuit in which cells each outputting an electric pulse corresponding to the energy of the photon each time the X-ray photon is detected are two-dimensionally arranged so as to form a two-dimensional pixel group; and A measurement circuit that measures the number of photons detected by each cell into two or more energy bands (number of energy bands M ≧ 2: M is a positive integer) for each pixel, and each cell measured by the measurement circuit A detector configured to include an output circuit that outputs a digital amount of an electric signal corresponding to the output of each of the energy bands as frame data;
The X-ray tube and the detector are opposed to each other with the imaging target interposed therebetween, the X-ray tube and the detector are rotated around the imaging target, and the frame data output by the detector is converted into the energy. A substance identification method applied to a system comprising a scanning means for collecting each band,
Generating panoramic image data to be imaged for each energy band based on the frame data;
Generating a CT (computed tomography) image of the imaging target;
Displaying the panoramic image on a monitor;
Designating a position of interest on the panoramic image displayed on the monitor according to an operator's instruction;
X-rays of the panoramic image represent at least one type of one or a plurality of substances having a thickness in the imaging target that exists in the projection direction in the imaging space between the X-ray tube and the detector corresponding to the position of interest. Identifying based on transmission data and morphological information of the substance obtained from the CT image;
The substance identification method characterized by having.
The identification step includes
Presenting the generated CT image to an operator;
From the CT image, there are N substances (N is a positive integer of 1 or more) as one or more substances having the thickness along the projection direction interactively with the operator. To determine that
The total L projected from the positions of L points (L ≧ N) including the position in the vicinity of the position of interest including the projection direction in the shooting space corresponding to the position of interest along the projection direction to the shooting space. Identify the directions of projections near each other,
The thicknesses t i1 , t i2 ,..., T iN (i = 1 to M;) of the N substances to be imaged that exist along the passing directions of the L projection directions in the CT image data. N is the number of substances) calculated from the CT image data as the morphological information,
When the absorption coefficients of the N substances for the X-rays for each energy band are μ 1j , μ 2j ,..., Μ Mj (j = 1 to N), the thicknesses of the substances are known and L C N-number of the absorption coefficient by solving the simultaneous equations mu 1j having an absorption coefficient as a variable, μ 2j, ..., and calculates the value of mu Mj,
For each of the N substances, the vertical axis represents μ 1j + μ 2j ,..., + Μ Mj (j = 1 to N) as a relative attenuation index: RAI, and the horizontal axis represents μ aj −μ bj (a, A positive integer satisfying b ≦ M, where a> b) is taken as a quality change index: SDI, and a two-dimensional scatter diagram is created.
The projection according to the position of interest is compared with the reference scatter diagram showing the relative attenuation index and the quality change index that are set in advance as a reference for each of the substances set in advance. Determining the type of each of the one or more substances present in the direction;
The method of identifying a substance according to claim 13.
15. The substance identification method according to claim 14, further comprising the step of presenting the determined type of the one or more substances present in the projection direction according to the position of interest to the user.