X-ray detectors CHAPTER 09 1 Si photodiodes 2 Si photodiode arrays 2-1 2-2 2-3 2-4 Structure Features Applications New approaches 3 CCD area image sensors 3-1 3-2 3-3 3-4 Direct CCD area image sensors CCD area image sensors with scintillator How to use Applications 4 CMOS area image sensors 4-1 Features and structure 4-2 How to use 4-3 Applications 5 Flat panel sensors 5-1 5-2 5-3 5-4 5-5 5-6 Features Structure Operating principle Characteristics How to use Applications 1 X-ray detectors X-rays were first discovered by Dr. W. Roentgen in Germany in 1895 and have currently been utilized in a wide range of fields including physics, industry, and medical diagnosis. Detectors for X-ray applications span a broad range including a-Si detectors, single crystal detectors, and compound detectors. There are many kinds of detectors made especially of Si single crystals. For X-ray detectors, Hamamatsu offers Si photodiodes, Si APDs, CCD area image sensors, and CMOS area image sensors, flat panel sensors, etc. Applications of our X-ray detectors include dental X-ray imaging and X-ray CT (computer tomography) in medical equipment fields, as well as non-destructive inspection of luggage, foods, and industrial products; physics experiments; and the like. In the low energy X-ray region called the soft X-ray region from a few hundred eV to about 20 keV, direct detectors such as Si PIN photodiodes, Si APDs, and CCD area image sensors are utilized. These detectors provide high detection efficiency and high energy resolution, and so are used in X-ray analysis, X-ray astronomical observation, physics experiments, etc. The hard X-ray region with energy higher than soft X-rays is utilized in industrial and medical equipment because of high penetration efficiency through objects. Scintillator detectors are widely used in these applications. These detectors use scintillators to convert X-rays into light and detect this light to detect X-rays indirectly. Especially in the medical field, the digital X-ray method, which uses X-ray detectors with large photosensitive area, is becoming mainstream, replacing the conventional film-based method. In non-destructive inspection, dual energy imaging, which allows image capturing with deep tones by simultaneously detecting high- and low-energy X-rays, is becoming popular. Example of detectable photon energy and spectral response range Wavelength [nm] = Si photodiode for X-ray (with scintillator) 1240 Photon energy [eV] NMOS/CMOS image sensor Si photodiode for X-ray (without scintillator) Back-thinned CCD Back-thinned CCD (windowless type) X-ray imaging CCD/CMOS image sensor (with FOS) Front-illuminated CCD (windowless type) Frontil uminated CCD Flat panel sensor 1 MeV 100 keV 10 keV 1 keV 100 eV 10 eV 1 eV 0.1 eV Photon energy Wavelength 0.01 nm 0.1 nm 1 nm 10 nm 100 nm 1 μm 10 μm KMPDC0178EB Hamamatsu X-ray detectors Type Si photodiode Si photodiode array CCD area image sensor CMOS area image sensor Image sensor Flat panel sensor Photodiode array with amplifier 2 Features combined with CsI(Tl) or ceramic scintillator are available. •Products •Back-illuminated CSP photodiodes that can be tiled (two-dimensional array) are available. long, narrow image sensor can be configured by arranging multiple arrays in a row. •ASupports dual energy imaging • of FOS to FFT-CCD (CCD with scintillator) •Coupling •Front-illuminated CCD for direct X-ray detection are available. •Coupling of FOS to CMOS image sensor large-area two-dimensional imaging •For •Captures distortion-free, high-detail digital images in real time Allows configuring a long, narrow image sensor by use of multiple arrays •(See chapter 5, “Image Sensors.”) 1. [Figure 1-2] Examples of Si photodiodes combined with scintillator Si photodiodes (a) Front-illuminated Si photodiode When used for X-ray detection, Si photodiodes are typically used with scintillators to form detectors for scintillator coupling. Hamamatsu offers two types of Si photodiodes for X-ray detection: Si photodiodes with scintillators and Si photodiodes without scintillators (which assume that users will bond the appropriate scintillators). In either case, Si photodiodes have a spectral response matching the emission band of scintillators. In the case of Si photodiodes with scintillators, CsI(Tl) scintillators or GOS ceramic scintillators are coupled with the Si photodiodes. The area around the scintillator is coated with a reflector to prevent the light emitted from the scintillator from escaping outside the photosensitive area [Figure 1-1]. Scintillator X-rays Protective resin PN junction Wire Base Disadvantage: Patterns and wires may be damaged when the scintillator is mounted. (b) Back-illuminated Si photodiode X-rays Scintillator Bump PN junction [Figure 1-1] Si photodiode with scintillator Advantages: There is no wiring so mounting the scintillator is easy. Multiple photodiodes can be tiled closely together. KPDC0037EA Scintillator Si direct photodiodes Reflector Photosensitive area Ceramic and the like Epoxy resin Si photodiode KSPDC0003ED Back-illuminated Si photodiodes have the PN junction on the side opposite to (on the backside of ) the light incident surface [Figure 1-2]. The photodiode surface bonded to the scintillator is flat and does not have wires. This prevents the photodiode from damage when the user attaches the scintillator. In addition, the detector can be made small because there is no area for wires as in a front-illuminated type. Furthermore, multiple photodiodes can be arranged with little dead space, so they can be used as a large-area X-ray detector. Because X-rays have no electric charge, they do not directly create electron-hole pairs in a silicon crystal. However, the interaction of silicon atoms with X-rays causes the release from ground state of electrons whose energy equals that lost by irradiated X-rays. The Coulomb interaction of these electrons causes electron-hole pairs to be generated, and these pairs are captured to detect X-rays. The probability that X-rays will interact with silicon atoms is therefore a critical factor when detecting X-rays directly. Si direct photodiodes can effectively detect X-rays at energy levels of 50 keV or less. Detection of X-rays less than 50 keV is dominated by the photoelectric effect that converts the X-ray energy into electron energy, so all energy of X-ray particles can then be detected by capturing the generated electrons with the Si photodiode. Detection of X-rays and gamma-rays from 50 keV up to 5 MeV is dominated by the Compton scattering, and part of the X-ray and gamma-ray energy is transformed into electron energy. In this case, the probability that the attenuated X-rays and gamma-rays will further interact with silicon (by photoelectric effect and Compton scattering) also affects the detection probability, making the phenomenon more complicated. Figure 1-3 shows the probabilities (dotted lines) of photoelectric effect and Compton scattering that may occur in a silicon substrate that is 200 µm thick, and the total interaction probabilities (solid lines) of silicon substrates that are 200 µm, 300 µm, and 500 µm thick. As can be seen from the figure, photodiodes created with a thicker Si substrate provide higher detection probability. With a 500 µm thick Si substrate, the detection probability is nearly 100% at 10 keV, but falls to just a few percent at 3 100 keV. The approximate range of electrons inside a Si direct photodiode is 1 µm at 10 keV and 60 µm at 100 keV. Si photodiode arrays 2. [Figure 1-3] Detection probabilities of Si direct photodiodes (Theoretical value) Interaction probability Photoelectric effect (200 μm) Compton scattering (200 μm) Total interaction probability Si su b thick strate n 500 ess μm 300 μm 200 μm X-ray energy (keV) KSPDB0018EA Baggage inspection equipment for examining the shapes and materials of items in baggage are used in airports and other facilities. Recently, high-accuracy CT baggage inspection equipment are being developed. Hamamatsu Si photodiode arrays with scintillators are widely used in these types of baggage inspection equipment. X-rays directed at baggage pass through objects and are converted into light by a scintillator. Then, the converted light is detected by the Si photodiode array. Hamamatsu Si photodiode arrays for baggage inspection feature low noise and consistent sensitivity and other characteristics between individual elements. The photodiode chips are mounted with high accuracy allowing highly accurate detection. Moreover, their sensitivity range matches the emission wavelength of scintillators making them suitable for baggage inspection. [Figure 2-1] Imaging example of baggage inspection equipment Structure 2-1 Many of the Hamamatsu Si photodiode arrays for baggage inspection equipment employ back-illuminated structure. Since back-illuminated Si photodiode arrays do not have patterns or wires on the surface that scintillators are bonded to, damage to patterns and wires when mounting scintillators can be avoided. Figure 1-2 shows cross sections for when a front-illuminated photodiode is combined with a scintillator and for when a back-illuminated photodiode [Table 2-1] Scintillator comparison table Parameter Condition Peak emission wavelength GOS ceramic Unit 560 512 nm 10 7 - 2.2 µs X-ray absorption coefficient 100 keV Refractive index At peak emission wavelength 1.74 1 3 100 ms after X-ray turn off 0.3 0.01 % 4.51 7.34 g/cm3 Transparent Light yellow-green - ±10 ±5 % Decay constant Afterglow Density Color tone Sensitivity variation 4 CsI(Tl) is combined with a scintillator. Examples of scintillators include CsI(Tl) and GOS ceramic. GOS ceramic features small variations in light emission and high reliability. We do not recommend CWO scintillators since they contain cadmium which falls under environmental management substances. [Figure 2-2] Spectral response of Si photodiode arrays and emission spectrum of scintillators (a) S11212-121 (Typ. Ta=25 °C) 100 100 S11212-121 80 CsI(TI) scintillator light emission characteristics 60 60 40 40 20 20 0 0 200 400 600 800 1000 Quantum efficiency (%) Relative light output (%) 80 0 1200 • Robustness Through the adoption of a back-illuminated structure, the photodiode array’s output terminals are connected to the circuit board electrodes using bumps without wires. Robustness is achieved by running the circuit wiring inside the board. • High reliability Since back-illuminated Si photodiode arrays do not have patterns or wires on the surface that scintillators are mounted on, scintillators can be mounted to the photodiode arrays without damaging the patterns and wires. High reliability is achieved since there are no wires, which could break due to temperature changes or be adversely affected in other ways. • Superior sensitivity uniformity In back-illuminated Si photodiode arrays (S11212/S11299 series), nonuniformity in sensitivity between elements are minimized, and the sensitivity variations at the sensor’s end elements are suppressed. The sensitivity uniformity has been greatly improved as compared with the previous product (S5668 series) and enables high-quality X-ray images to be obtained. Wavelength (nm) KSPDB0330EB [Figure 2-3] Sensitivity uniformity (b) S11212-321 (Typ. Ta=25 °C) 110 (Typ. Ta=25 °C) 100 100 Back-illuminated Si photodiode array S11212/S11299 series 80 GOS ceramic scintillator light emission characteristics 60 60 40 40 20 20 Quantum efficiency (%) Relative light output (%) Relative sensitivity (%) S11212-321 80 100 Previous product S5668 series 90 80 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 0 0 200 400 600 800 1000 0 1200 Element KPDB0023EB Wavelength (nm) KSPDB0331EB 2-2 • Features Low cost • Allows tiling Back-illuminated Si photodiodes do not have space for wires as shown in Figure 1-2 (b), so multiple photodiodes can be tiled close together. [Figure 2-4] Tiling example The adoption of a back-illuminated structure simplifies scintillator mounting and other processes, and this leads to shorter manufacturing process. Moreover, back-illuminated Si photodiode arrays use bumps for their electrodes. Bumps are used in the manufacturing process of LCD monitors and the like and are suitable for high-volume production. The use of bumps has cut cost when compared with our previous products. 5 2-3 [Figure 2-6] Multiple arrangement example (S11212-121) Applications Dual energy imaging In normal X-ray non-destructive inspection, the X-ray transmitted through an object is detected by a single type of sensor, and the shape, density, and other characteristics of the object is made into an image using shading. In comparison, in dual energy imaging, high-energy image and low-energy image are captured simultaneously by two types of sensors, and the images are combined through arithmetic processing. This enables images that show detailed information about hard and soft objects to be obtained. Dual energy imaging is used in a wide range of fields such as security where specific chemicals, explosives, and other dangerous objects are detected and in the field of grain, fruit, meat, and other inspections. Hamamatsu back-illuminated Si photodiode arrays S11212/S11299 series support dual energy imaging. It is structured so that two types of Si photodiode arrays with scintillators can be combined to create top and bottom layers in order to simultaneously detect high-energy and low-energy X-rays. Moreover, its construction allows multiple arrays to arranged in close proximity to form a line sensor. This makes measurement of long and narrow objects possible. 2-4 New approaches Hamamatsu is currently developing a special ASIC that can be combined with the proven Si photodiode array for X-ray CT/baggage inspection. Hamamatsu ASICs are compact and operate on low power. They can be made into custom order products. We are developing an X-ray CT module that combines two ASICs and a 32 × 16 (512) element back-illuminated Si photodiode array. Four of these modules arranged side by side can be used in 128 slice X-ray CT scanners. Hamamatsu modules with ASICs feature high X-ray durability. We can also provide modules with heatsinks or GOS scintillators. [Figure 2-5] Example of combining back-illuminated Si photodiode arrays [S11212-421 (top) and S11299-121 (bottom)] [Figure 2-7] Dual energy imaging X-ray source Baggage under inspection Low-energy X-rays High-energy X-rays I Cross section Low-energy X-rays Phosphor screen Conveyor High-energy X-rays Si photodiode array for low-energy detection Photosensitive area Scintillator Si photodiode array for low-energy detection Photosensitive area Si photodiode array for high-energy detection The top and bottom photodiode arrays detect low-energy X-rays and highenergy X-rays, respectively, to capture two types of images simultaneously. Si photodiode array for high-energy detection KPDC0038EA 6 [Figure 2-8] Modules with ASICs (a) With heatsink (b) For 128 slice scanners (with GOS scintillators and heatsinks) 3. CCD area image sensors 3-1 Direct CCD area image sensors Windowless CCDs (front-illuminated type) are used for directly detecting X-rays from 0.5 keV to 10 keV. These CCDs cannot be used to detect X-rays whose energy is lower than 0.5 keV since an absorption layer exists on the CCD surface. A direct CCD (back-thinned type) must be used to detect X-rays whose energy is lower than 0.5 keV. To achieve high quantum efficiency in the energy region higher than 10 keV, a direct CCD with a thick depletion layer must be used. Direct CCDs are capable of both X-ray imaging and spectrophotometry. X-rays can also be detected in photon-counting mode (method for counting individual photons one by one). Direct CCDs are used in fields such as X-ray astronomy, plasma analysis, and crystal analysis. Principle of X-ray direct detection Photons at an energy higher than a specified level generate electron-hole pairs when they enter a CCD. If the photon energy is small as in the case of visible light, only one electron-hole pair is generated by one photon. In the vacuum-UV-ray and soft-X-ray regions where photon energy is greater than 5 eV, multiple electron-hole pairs are generated by one photon. The average energy required for silicon to produce one electron-hole pair is approx. 3.6 eV. So an incident photon at 5.9 keV (Kα of manganese), for example, generates 1620 electron-hole pairs in the CCD. The number of electrons generated by direct X-ray detection is proportional to the energy of the incident photons. Characteristics Figure 3-1 shows the result when X-rays (Mn-Kα/Kβ) emitted from a Fe-55 radiation source are detected by a CCD. Spectrum resolution is usually evaluated by using the FWHM (full width at half maximum). The Fano limit (theoretical limit of energy resolution) of Si detectors for Fe-55 is 109 eV. Major factors that degrade energy resolution are CCD charge transfer efficiency and CCD noise including dark current. When a CCD is sufficiently cooled down and is operated at a charge transfer inefficiency of 1 × 10-5 or less, the energy resolution is determined by the readout noise. To improve energy resolution, the CCD readout noise has to be less than 5 e- rms. The energy resolution of optimally adjusted Hamamatsu CCDs is below 140 eV for Fe-55. There are two modes for evaluating the CCD quantum efficiency in the X-ray region. One is the photon-counting mode, and the other is the flux mode that integrates all photons. The quantum efficiency in the visible region is usually evaluated in the flux mode [Figure 3-2]. 7 [Figure 3-1] CCD energy resolution in X-ray (Mn-Kα/Kβ) detection (typical example) 2500 • Mn-Kα (5.9 keV) 2000 Number of counts Features High sensitivity and low noise are achieved by use of FFT (full frame transfer) type CCD, which is widely used for analysis and measurement. 1500 • 1000 500 5.5 6.0 6.5 Structure and characteristics 7.0 Energy (keV) KMPDB0236EA [Figure 3-2] Quantum efficiency vs. photon energy (Typ. Td=-100 °C) 100 High-quality image type and low cost type available The high-quality image type CCD uses a CsI(Tl) scintillator to convert X-rays to visible light, and the low cost type CCD uses a GOS scintillator. Mn-Kβ (6.5 keV) 0 5.0 Highly detailed images • CCD area image sensors with FOS This CCD is coupled to an FOS which is an FOP with scintillator. This CCD with FOS utilizes CsI(Tl) as the scintillator to achieve high resolution. Flux mode Quantum efficiency (%) [Figure 3-3] Structure of CCD with FOS Scintillator CsI(Tl) FOP Photon-counting mode Base 10 1 KMPDC0295EA 1 10 Photon energy (keV) KMPDB0154EA 3-2 FOS CCD chip CCD area image sensors with scintillator Typically, CCD chips are damaged to some extent when exposed to X-rays. However, this CCD with FOS has an FOP on the CCD chip’s photosensitive area, and the FOP also serves as an X-ray shield to suppress damage by X-rays. Electric charges generated by X-rays incident near the surface of the CCD may cause noise, where white spots are seen at random positions. It degrades the image quality. This CCD with FOS, however, maintains high-quality images since the amount of X-rays incident on the CCD is small due to the X-ray shielding effect of the FOP [Figure 3-4]. [Figure 3-4] X-ray transmittance in FOP (X-ray tube voltage: 80 kV) 8 For panoramic/ cephalo imaging Transmittance (%) Besides visible, infrared, and ultraviolet light, a CCD can directly detect and image X-rays below 10 keV. However, in the X-ray region from several dozen to more than 100 keV used for medical diagnosis and industrial non-destructive inspection, scintillators are needed to convert the X-rays into visible light. In this case, CsI(Tl) and GOS scintillators are generally used, which convert X-rays into light at a peak of around 550 nm. The CCD then detects this light for X-ray detection. In X-ray imaging applications requiring large-area detectors, Hamamatsu provides front-illuminated CCD coupled to an FOS (fiber optic plate with scintillator). We also respond to requests for CCD coupled to an FOP (fiber optic plate) (scintillator to be implemented by the user). For intraoral imaging FOP thickness (mm) KMPDB0297EA [Figure 3-5] CsI(Tl) absorption coefficient [Figure 3-7] CsI(Tl) emission intensity Light output (relative value) Absorption coefficient (cm-1) (70 kVp, 4.8 mm Al filter) X-ray energy (keV) Scintillator thickness (mg/cm2) KMPDB0293EA KMPDB0298EA The resolution of a CCD with FOS is mainly determined by the following factors: · Pixel size · Scintillator specifications (material, thickness) · Gap between CCD chip and FOP (e.g., chip flatness) Due to the CCD structure, the resolution determined by the pixel size cannot be exceeded. The thicker scintillator results in higher emission intensity, yet the resolution deteriorates as the thickness increases (there is a trade-off here between emission intensity and resolution) [Figure 3-6, 3-7]. Since the resolution deteriorates as the gap between the chip and FOP becomes wider, technology for keeping this gap at a narrow width is essential. Note that the FOP flatness is superior to the chip flatness and so poses no problems. [Figure 3-6] CsI(Tl) resolution Contrast transfer function (%) (10 Lp/mm, 40 kVp, no filter) • Buttable configuration To obtain a long photosensitive area, panoramic imaging CCDs use two chips and cephalo imaging CCDs use three chips, with each chip being arranged in close proximity in a buttable configuration. There is a dead space between each chip. See Figure 3-8 for an example of an insensitive area caused by this dead space. 3-3 How to use There are two methods for capturing X-ray images: oneshot and TDI operation imaging. For one-shot imaging, in the CCD pixels, charges are constantly generated due to dark current, so those charges must be constantly drained when no X-rays are being input (standby state). When using TDI operation, the pixel transfer speed has to be made to match the motion speed of the object. (See chapter 5, “Image sensors.”) Image correction Scintillator thickness (mg/cm2) KMPDB0299EA CCDs may sometimes have pixel defects known as white spots where the dark current is large, and black spots where the output is low (low sensitivity). Scintillator and FOP performance also affect the image quality of CCDs with FOS. To achieve high image quality, we recommend using software to compensate for the dark current and sensitivity. See chapter 5, “Image sensors,” for information on compensating for pixel defects, dark current, and sensitivity. Multiple CCD chips are combined in CCDs for panoramic/ cephalo imaging and non-destructive inspection, and there is a dead space between each chip. Software compensation may help suppress effects from this dead space. 9 [Figure 3-8] Example of CCD for cephalo imaging and non-destructive inspection (S8658-01) X-ray detection side 220.0 (H) × 6.0 (V) 5.6 FOP 228.0 ± 0.3 FOP 3.0 15.24 60.96 B20 ← B14 *1 *2 Left chip 1.6 Center chip Left chip A1 → A13 B1 45.72 → 30.48 Right chip C1 → B13 Center chip Edge pixels C20 ← C14 25.4 28.0 ± 0.3 A20 ← A14 60.96 Dead area 1: 250 μm min. 350 μm max. C13 45.72 3.4 Scintillator 0.45 Enlarged view of portion *2 2.54 Center chip ±50 μm max. FOP 7.2 ± 0.2 223.0 ± 0.5 ±50 μm max. Enlarged view of portion *1 Right chip Edge pixels Dead area 2: 130 μm min. 200 μm max. KMPDA0251EA Precautions [Figure 3-9] Precautions when holding the sensor (a) Hold the board by the edges with fingers. Take the following precautions when using an X-ray CCD. (1) Anti-static and surge measures For measures to avoid electrostatic charge and surge voltage on an X-ray CCD, refer to “1-3 How to use” in section 1 “CCD area image sensors,” in chapter 5, “Image sensors.” (2) Operating and storage environment X-ray CCDs are not hermetically sealed, so avoid operating or storing them in high humidity locations. Also do not apply excessive vibrations or shock during transportation. (b) Do not apply force to the FOS. (3) Deterioration by X-ray irradiation Like other X-ray detectors, X-ray CCD characteristics deteriorate due to excessive X-ray irradiation. In some applications, CCDs need to be replaced as a consumable product. (4) Handling CCD with FOS · FOP is made from glass, so do not apply a strong force and shock to it. · Do not touch the scintillator section and photosensitive area. A scratched scintillator will cause changes in sensitivity. Bonding wires are coated with protective resin, but do not touch the resin as it can damage or break the wire. · When holding the sensor, hold the board by the edges with your fingers and make sure not to touch the exposed areas of the leads and wires as shown in the photos [Figure 3-9 (a)]. Touching the exposed areas of the leads and wires may damage the sensor due to static electricity. · Never apply force to the FOS. It may damage the scintillator [Figure 3-9 (b)]. 10 (c) Do not bend the cable excessively. (5) Handling the module with an assembled cable · Do not apply excessive force to the sensor section. Biting it, applying force to it, or dropping it may cause damage or failure. · Applying excessive force by bending or pulling on the cable may cause breakdowns such as the cable breaking internally, so please handle the cable with care [Figure 3-9 (c)]. 3-4 [Figure 3-12] Example of cephalo imaging Applications Non-destructive inspection (for industry) X-ray CCDs can be used to perform one-shot imaging or perform imaging in TDI operation to inspect objects moving on a conveyor and for other purposes. [Figure 3-10] Example of printed circuit board imaging Radiography (1) Panoramic imaging Panoramic X-ray imaging devices capture images by using the X-ray source and detector unit designed to rotate around the patient’s head. A TDI-CCD allows capturing panoramic diagnostic images that are greater than the photosensitive area. [Figure 3-11] Example of panoramic imaging (2) Cephalo imaging Cephalo X-ray imaging devices capture images of the head. These devices use TDI-CCDs for acquiring diagnostic images like panoramic imaging. 11 CMOS area image sensors 4. X-ray CMOS area image sensors are image sensors designed for intraoral imaging and non-destructive inspection. These image sensors make use of advantages offered by active pixel type CMOS devices, including high integration, sophisticated functions, and high S/N. They contain a timing generator, vertical and horizontal shift registers, readout amplifier, A/D converters, and LVDS [Figure 4-1]. The digital input and output make these image sensors very easy to use. These image sensors contain a global shutter function (integrates charges simultaneously in all pixels) that allows acquiring one shot of an X-ray image in synchronization with the X-ray irradiation timing. Since these image sensors have an internal A/D converter, analog video wiring can be kept short to reduce noise. The internal A/D converter also simplifies the external circuit and helps hold down the overall cost. [Figure 4-1] Block diagram (typical example) Timing generator Monitoring photodiode amplifier Low-speed A/D converter Column CDS Image amplifier High-speed A/D converter LVDS Photosensitive area CLK (P) LVDS Vertical shift register Monitoring photodiode Data (P) determines that the X-ray emission has started and instructs the CMOS area image sensor to start charge integration and readout. [Figure 4-2] Monitoring photodiode Monitoring photodiode (red section along circumference) KMPDC0308EA Low power consumption X-ray CMOS area image sensors have an internal high-speed A/D converter (14 bits) for image data, and a low-speed A/D converter (10 bits) for monitoring photodiode data. The highspeed A/D converter which uses up much current starts only when image data is transferred. Only the low-speed A/D converter which consumes low power is on during the long periods of standby for X-ray irradiation. This keeps the average power consumption lower [Figure 4-3]. [Figure 4-3] Block diagram CLK (N) Data (N) Monitoring photodiode charge Photosensitive area Horizontal shift register Pixel charge MCLK MST Vdd GND KMPDC0307EB Photosensitive area Monitoring photodiode amplifier Low-speed A/D converter Monitoring photodiode data Image amplifier High-speed A/D converter Image data Starts only when transferring image data ↓ Achieves low power consumption KMPDC0309EB 4-1 Features and structure Comparing CMOS and CCD In regions where signal levels are low, CCD area image sensors provide better image quality. CMOS area image sensors use CMOS technology, which makes them superior in terms of multifunctionality and easy to use. CMOS area image sensors also offer the advantage of a lower total cost because peripheral circuit functions can be built into the CMOS chip. Internal monitoring photodiode In the X-ray CMOS area image sensors, a monitoring photodiode for detecting the X-ray irradiation start timing (trigger) is mounted as a narrow strip along the entire circumference on the outer side of the photosensitive area [Figure 4-2]. Monitoring photodiode signals are transmitted repetitively at specific intervals. The output is sent to an external control circuit. When this output exceeds the specified threshold value, the external control circuit 12 Global shutter function Hamamatsu X-ray CMOS area image sensors employ a global shutter function. The global shutter enables integration of all pixels simultaneously and therefore produces high-resolution images even when X-rays are emitted during image data readout or in machine-vision and other applications where images of moving objects are captured, almost without any of their adverse effects. APS (active pixel sensor) type Unlike the charge transfer type image sensors exemplified by CCDs, X-Y address type CMOS area image sensors read out integrated pixel charges and thus have long data line wiring. This wiring capacitance becomes a large noise source when the transistors in each pixel switch. As such, Hamamatsu X-ray CMOS area image sensors employ an APS type that houses an amplifier in the pixel. Since the integrated charge is converted into voltage for each pixel, low-noise images can be achieved. [Figure 4-5] Wiener spectrum vs. spatial frequency (typical example) Detective quantum efficiency (DQE) is one of the parameters that define the performance of an X-ray detector. It shows the level of the output image signal S/N (SNROUT ) with respect to the S/N (SNRIN) of the X-ray irradiated on the X-ray detector. Since X-ray noise is closely related to the radiation dose, the DQE can be used as a measure of the photon detection efficiency of the X-ray detector. The DQE can be used as a measure for evaluating the incident X-ray photon detection efficiency and image quality. Higher DQE indicates higher efficiency in obtaining high-quality image from the incident X-rays. The DQE is given by equation (1) or (2). DQE = (Ta=25 °C) 10-4 Wiener spectrum (mm2) High detective quantum efficiency 10-5 10-6 10-7 10-8 0 (SNROUT)2 ....... (1) (SNRIN)2 10 15 20 25 Spatial frequency (cycles/mm) KMPDB0387EA [Figure 4-6] DQE vs. spatial frequency (typical example) MTF2 DQE = 5 ....... (2) ϕWS (Ta=25 °C) 1.0 MTF : modulation transfer function ϕ : number of incident X-ray photons WS : Wiener spectrum 0.8 [Figure 4-4] MTF vs. spatial frequency (S11684, typical example) (Ta=25 °C) 1.0 0.8 MTF 0.6 0.4 0.6 DQE In an ideal imaging system without noise, the DQE is equal to 1. In a real imaging system, noise introduced in various processes such as noise generated by the pixels and electronic circuit increases the Wiener spectrum, and especially in the high-frequency region where the effect is great, the DQE decreases. Hamamatsu X-ray CMOS area image sensors use high-emission-efficiency and highresolution CsI(Tl) for the scintillator to achieve higher DQE, which produces high-quality image and lower X-ray radiation dose [Figure 4-4, 4-5, 4-6]. For information on MTFs, see “1-2 Characteristics and Resolution” in section 1 “CCD image sensors,” in chapter 5, “Image sensors.” 0.4 0.2 0 0 5 10 15 20 25 Spatial frequency (cycles/mm) KMPDB0388EA 4-2 How to use Since X-ray CMOS area image sensors have an internal timing generator, it is possible to monitor X-ray emission timing and also integrate and read out image data just by applying a master start pulse (MST) and master clock pulse (MCLK). Data from the monitoring photodiode and image data are switched by an internal switch so that they are transmitted from the same output wiring [Data(P), Data(N)] [Figure 4-7]. For information on image correction, see chapter 3, “CCD area image sensors.” 0.2 [Figure 4-7] Input/output wiring diagram 0 0 5 10 15 20 25 Timing generator High-speed A/D converter Spatial frequency (cycles/mm) KMPDB0386EA Cable Low-speed A/D converter Chip CLK(P) CLK(N) Data(P) Data(N) MCLK MST Vdd GND Base KMPDC0310EB 13 Precautions The precautions on using X-ray CMOS area image sensors are the same as those for X-ray CCDs (see section 3-3, “Precautions,” in chapter 3, “CCD area image sensors”). 4-3 Applications Intraoral imaging Detailed diagnostic images of two to three teeth can be obtained by inserting a CMOS area image sensor for intraoral imaging and non-destructive inspection into the patient’s mouth. For intraoral imaging, Hamamatsu provides CMOS modules that use a relatively large area CMOS area image sensor of 1000 (H) × 1500 (V) pixels or 1300 (H) × 1700 (V) pixels, both with a pixel size of 20 × 20 µm, assembled with a cable [Figure 4-8]. The scintillator uses CsI(Tl) that achieves a high resolution of 15 to 20 line pairs/mm. Coupling the FOS to the CMOS gives high durability against X-ray exposure. For example, these modules can operate up to 100000 times or more under X-ray irradiation of approximately 250 µGy at 60 kVp. Besides this feature, the sensor unit of these CMOS modules is thin and compact, allowing X-ray imaging even in a narrow section. 5. Flat panel sensors Flat panel sensors are X-ray imaging modules that use a large-area CMOS image sensor combined with a scintillator. The detector (two-dimensional photodiode array), highperformance charge amplifier, and scanning circuit are all integrated onto a large-area CMOS monolithic singlecrystal silicon chip. The A/D converters, memories, interface circuit, and the control signal generator that controls these components are assembled into a module. There is no need to use an external circuit to operate the device. The flat panel sensor can capture megapixel-class, high definition digital images which are distortion-free in both still and moving images. The thin profile and light weight make the flat panel sensor easy to install into other equipment. Flat panel sensors are now widely used in various types of X-ray imaging systems including CT. We also offer flat panel sensors that use advanced a-silicon, which features large photosensitive area and high-speed response, for the detector. [Figure 5-1] Flat panel sensors [Figure 4-8] CMOS module [Figure 5-2] Imaging examples (a) Hornet 14 (b) Fish • High sensitivity • High resolution This structure lowers the noise level, which is about one figure less than the current mode passive pixel type. Because of its low noise and high S/N features, the active pixel type flat panel sensor acquires high definition images from low energy X-rays. • High frame rate [Figure 5-3] Internal circuits of CMOS chip • Wide dynamic range • Distortion-free images • Directly deposited CsI(Tl) scintillator type is available. (a) Current mode passive pixel type Address switch ∙∙∙ U ∙∙∙ U Photodiode Structure ∙∙∙ U ∙∙∙ U ∙∙∙ Figure 5-3 shows the internal circuit of a CMOS chip for flat panel sensors. Two-dimensional X-ray image signals converted into fluorescence by a scintillator are accumulated as an electric charge in the junction capacitance of each photodiode with excellent linearity. The accumulated charges are then output one row at a time through the data line by the vertical shift register when the address switch turns on. Since the flat panel sensor operates in charge integration mode, the output video signal voltage V(t) is expressed by equation (1). ∙∙∙ 5-2 Vertical shift register Features Data line 5-1 Charge amp array ∙∙∙ ∙∙∙ Horizontal shift register Horizontal shift register Video signal Video signal Multiport readout Video signal U: unti-over flow circuit KACCC0224EB V(t) = G × Q(t) = G × I(t) × t1 = G × I(t) × 1/Sf .....(1) (b) Active pixel type G : amplifier gain Q(t): integrated charge I(t) : photodiode photocurrent t1 : integration time Sf : frame rate ∙∙∙ Photodiode Vertical shift register When a constant X-ray radiation dose is striking an object, the photocurrent generated in a photodiode is constant. The output voltage can be increased by slowing the frame rate (making the integration time longer). The frame rate can be controlled by the external trigger mode described later on. The saturation charge is determined by the photodiode junction capacitance. The maximum video output value after A/D conversion is set to the saturation charge value. Address switch Amplifier Data line ∙∙∙ Horizontal shift register Video signal KACCC0308EA Amplifier circuit There are two types of amplifier circuits for flat panel sensors: a current mode passive pixel type and an active pixel type. The current mode passive pixel type has an amplifier for each column of the photodiode array, where the amplifier is connected to each photodiode via address switch. The amplifiers are formed on one side of the two-dimensional photosensitive area as an amplifier array. The current mode passive pixel type allows a high fill factor and high radiation durability. However, the input capacitance caused by the data line limits the reduction in amplifier thermal noise. The active pixel type structure eliminates the foregoing problem in the current mode passive pixel type. The active pixel type has an amplifier for each pixel, and the accumulated charges are converted into voltage there. Scintillator Flat panel sensors employ an indirect X-ray detection method that converts X-rays into light using a scintillator and then detects that light. By optimizing the wafer process technology, Hamamatsu has succeeded in developing a high-sensitivity photodiode that matches the spectral characteristics of the scintillator. The CsI(Tl) scintillator [Figure 5-4] used for most flat panel sensors has needle-like crystals through which scintillation light propagates, and flat panel sensors with CsI(Tl) scintillator therefore have superior resolution and emission intensity compared to flat panel sensors using other scintillators composed of grain (particle) crystals (such as GOS). 15 [Figure 5-4] Cross-sectional photo of CsI(Tl) scintillator showing needle-like crystals (b) Direct deposition type X-rays Directly deposited needle-like structure CsI(Tl) Photodiode CMOS chip KACCC0271EB Tiling [Figure 5-5] Cross-sectional photo of GOS scintillator showing grain crystals Some flat panel sensors have a large-format photosensitive area achieved by using the “tiling” technique. Highly precise tiling techniques now allow acquiring an X-ray image which contains no gaps at the tiling edges. Though the sensitivity of the pixels at the tiling edges decreases, it is still high enough to be corrected by software, so seamless images can be obtained. [Figure 5-7] Cross-sectional diagram of tiling edges of photosensitive area 1.9 to 2.3 mm CsI(Tl) scintillator Photodiode array B CsI(Tl) scintillator There are two scintillator-to-photodiode coupling methods. One method uses an FSP (flipped scintillator plate), which is a glass plate on which the scintillator is deposited, and the scintillator side of the FSP is attached in close contact with the photodiode. The other method is direct deposition of scintillator onto the photodiode. The method using a CsI(Tl) FSP has better fluorescent intensity and resolution than medical screens utilizing GOS. The direct deposition method further improves the resolution because it suppresses fluorescence scattering compared to the FSP method. Hamamatsu provides both FSP type and direct deposition type flat panel sensors that can be selected according to the application. [Figure 5-6] Cross-sectional structures of chip (a) FSP type Tiling position accuracy: 0 ± 60 μm Photodiode array A KACCC0386EB 5-3 Operating principle Signal readout method The following methods are generally used to read out digital signals. (1) Serial drive method This method reads out video data by serially driving all pixels, so the frame rate slows down when there are a large number of pixels. (2) Parallel drive method X-rays Glass substrate Needle-like structure CsI(Tl) • Air gap Photodiode CMOS chip KACCC0270EB 16 Single port readout method This method divides the monolithic photosensitive area into multiple blocks, and reads out video data through a single port by driving each block in parallel. Figure 5-8 shows a schematic of an photosensitive area divided into “n” blocks. Since flat panel sensors have many pixels numbering more than one million, the serial drive method causes the frame rate to drop. The single port readout method, however, offers high speed and easy processing of video data and so is used for most flat panel sensors. [Figure 5-9] Example of spectral response and CsI(Tl) scintillator emission spectrum This method reads out video data through multiple ports to achieve even higher speed drive than the single port readout method. Providing multiple ports for video data readout can increase the image data transfer amount per unit time, which is larger than that of the single port readout method. Some flat panel sensors use this method. [Figure 5-8] Schematic of parallel drive method a2 a3 b1 b2 b3 n1 n2 n3 Shift register Shift register Shift register A/D converter A/D converter A/D converter FIFO memory FIFO memory FIFO memory 100 Spectral response Vertical shift register a1 (Typ. Ta=25 °C) 50 40 80 30 60 20 CsI(Tl) scintillator emission spectrum 40 10 20 0 400 500 600 700 Relative emission output (%) Multiport readout method Quantum efficiency (%) • 0 800 Wavelength (nm) KACCB0062EB Linearity Digital signal KACCC0225EB Flat panel sensors exhibit excellent linearity versus the incident X-ray levels. Figure 5-10 shows the output linearity of a flat panel sensor (14-bit output). The upper limit of the 14-bit output is 16383 gray levels. [Figure 5-10] Output linearity (14-bit output, typical example) Flat panel sensors support the following video output interfaces: RS-422, LVDS, USB 2.0, and Gig-E. USB 2.0 and Gig-E support our digital camera interface DCAM. Binning mode Some flat panel sensor models have a binning mode function that simultaneously reads out multiple pixel data. Up to 4 × 4 pixels can be selected for binning though this depends on the sensor models. Increasing the number of binning pixels also increases the sensor frame rate. Note that the highest resolution is obtained by single operation (1 × 1 mode) without using binning mode. Video output (gray level) Video output interface Irradiation dose (relative value) KACCB0063EB Dark video output 5-4 Characteristics Spectral response The photosensitive area of flat panel sensors consists of a two-dimensional photodiode array. Figure 5-9 shows the spectral response of a typical flat panel sensor and the emission spectrum of a CsI(Tl) scintillator. To achieve high sensitivity, the photodiode array is designed to have high sensitivity in the vicinity of the peak emission wavelength of CsI(Tl). The X-ray energy range at which flat panel sensors are sensitive differs depending on the sensor model. Refer to their datasheets for details. When the integration time is set longer, dark video output slightly increases due to the photodiode dark current. Figure 5-11 shows the relationship between dark video output and integration time for a flat panel sensor (14-bit output). The photodiode dark current (ID) is expressed by equation (2). ID = K/G [C/s] ..........(2) K : increasing rate [gray levels/s] of dark video output versus integration time G : conversion gain 17 Dark video output (gray level) [Figure 5-11] Dark video output vs. integration time (14-bit output, typical example) ( ) V2tot(rms) = 1 Ct 8 kT gm Cf 3 Vtot(rms) = Ct Cf 2 β1 + ( ) Kf Ct Cox 2 W L Cf 2 β2 ....(3) 1 Kf 8 kT β1 + β2 .......(4) gm Cox2 W L 3 Ct = Cp + Cf + Cd ........................................................(5) Vtot : T : gm : β1, β2 : Kf : Cox : W : L : Integration time (s) KACCB0064EA A slight drift occurs in the dark video output after the power is turned on. Figure 5-12 shows examples from measuring this dark video output drift. In internal trigger mode (2 frames/s), the dark video output shows little change after the power is turned on. At a slow frame rate, however, the dark video output shifts. In applications where fluctuations in the dark video output cause problems, determine how often dark images should be acquired for correction to meet the allowable drift level range. Dark video output (gray level) [Figure 5-12] Drift characteristics of dark video output (14-bit output, typical example) total noise voltage absolute temperature [K] transconductance of charge amplifier first-stage transistor constants determined by charge amplifier 1/f noise constant of charge amplifier first-stage transistor gate oxide film capacitance of charge amplifier first-stage transistor W length of charge amplifier first-stage transistor L length of charge amplifier first-stage transistor The noise level of current mode passive pixel type CMOS image sensors depends on the pixel size and the number of pixels. The lower limit of flat panel sensor dynamic range is determined by noise and the upper limit by the saturation charge. This means that the dynamic range is derived from the ratio of saturation charge to noise. In the active pixel type, the video line parasitic capacitance is extremely low, so the noise is small. Resolution Resolution is a degree of detail to which image sensors can reproduce an input pattern in the output. The photosensitive area of a flat panel sensor consists of a number of regularly arrayed photodiodes, so the input pattern is output while being separated into pixels. Therefore as shown in Figure 5-13, when a square wave pattern of alternating black and white lines with different intervals is input, the difference between black and white level outputs becomes smaller as the pulse width of the input pattern becomes narrower. In such a case, the contrast transfer function (CTF) is given by equation (6). VWO - VBO × 100 [%].......... (6) VW - VB CTF = VWO: VBO : VW : VB : output output output output white level black level white level (when input pattern pulse width is wide) black level (when input pattern pulse width is wide) [Figure 5-13] Contrast transfer function characteristics Elapsed time after power-on (min) 1 line pair KACCB0065EA Noise and dynamic range Input pattern White Flat panel sensors were developed based on CMOS image sensors. CMOS image sensors transfer charges accumulated in the photodiodes to the readout circuit through the video line. In the current mode passive pixel type CMOS image sensors, noise is expressed by equation (4). The video line parasitic capacitance (Cd) is very large compared to the photodiode junction capacitance (Cp) and charge amplifier feedback capacitance (Cf ), so the video line parasitic capacitance becomes a dominant source of noise. 18 Output Black Pixel pitch KMPDC0070EA The fineness of the black and white lines on the input pattern is given by the spatial frequency of the input pattern. The spatial frequency is the number of black and white line pairs per unit length. In Figure 5-13, the spatial frequency corresponds to the reciprocal of the distance from one white edge to the next white edge in the pattern. It is usually represented in units of line pairs/mm. The finer the input pattern or the higher the spatial frequency, the lower the CTF will be. [Figure 5-14] Contrast transfer function vs. spatial frequency [pixel size: 50 × 50 µm, CsI(Tl) direct deposition, typical example] X-ray irradiation damage For example, on the C7942CA-22, if 80 kVp of X-rays are irradiated over 4 hours a day (1 × 1 mode, frame rate: 2 frames/s), the detector life is 152 days. Contrast transfer function (%) 5-5 How to use Connection method Spatial frequency (line pairs/mm) KACCB0193EA The resolution and sensitivity of flat panel sensors to X-rays depend on the scintillator thickness. Both are in a tradeoff relation. Our flat panel sensors are designed for optimal scintillator thickness by taking the application and pixel size into account to deliver high resolution and high sensitivity. Setup is simple. All that is needed is to connect the flat panel sensor to a PC and power supply using the data cable and power cable (some models require an external trigger input cable). Then supplying the voltage to the flat panel sensor will start real-time X-ray image acquisition from the PC control. Figure 5-15 shows a connection example of an X-ray imaging system using a flat panel sensor. Use a monotonically increasing series power supply with a transformer for the voltage source. [Figure 5-15] Connection example (C10500D-03) OS + Image acquisition software Video output (14-bit digital output) Frame grabber board Vsync, Hsync, Pclk PC/AT compatible computer (rear) Reliability In ordinary X-ray detectors, deterioration in performance such as a drop in sensitivity and an increase in dark video output occurs due to X-ray irradiation. Likewise, flat panel sensor characteristics deteriorate due to X-ray irradiation. For example, an FSP type flat panel sensor with an aluminum top cover intended for non-destructive inspection is designed for use at an X-ray energy from 20 kVp to 100 kVp, and can be used up to an accumulated irradiation dose of one million roentgens if used under 100 kVp X-ray energy. When the photosensitive area is uniformly irradiated with X-rays, the dark current also increases almost uniformly over the photosensitive area. The dark current might partially increase in the photosensitive area, but this can be eliminated by dark image correction. When the partial increase in dark video output caused by increased dark current has exceeded the dark image correction limit, the flat panel sensor should be replaced as a consumable part. The life of flat panel sensors can be extended by setting the X-ray dose to a lower level within the detectable range and by preventing X-rays from irradiating the flat panel sensor except during imaging. Another effective way to extend the detector life is to use pulsed X-rays. Monitor IntExtGrb ExtTrgGrb IntExtIO ExtTrgIO C10500D-03 X-ray source Voltage source (A.vdd, D.vdd) Object KACCC0642EA Trigger mode Flat panel sensors have two trigger modes (internal trigger mode and external trigger mode). In internal trigger mode, the sensor always operates at the maximum frame rate and constantly outputs the sync signals and video signal. To capture images in external trigger mode, apply external trigger pulses as shown in Figure 5-16. Vsync+, Hsync+, and video signal are output after time Tvd elapses from the rising edge of the external trigger pulse. To synchronize with the pulse X-ray source, apply X-rays during Txray. 19 [Figure 5-16] External trigger mode Frame time (Tvc to Tmax) 50% of frame time recommended ExtTrgLemo ExtTrgGrb (TTL) Tvd Tvc - Tvdpw Txray Tvl Vsync+ (RS-422) Hsync+, Pclk, and effective video output are the same as those for internal trigger mode. · Tmax is defined using the reciprocal of the minimum value of Sf(ext). · Txray = Frame time - Tvd - (Tvc - Tvdpw) · Tvl = Frame time - (Tvc - Tvdpw) KACCC0341ED Defect lines Charges accumulated in the photodiodes are transferred to the readout circuit through the data line by turning on the CMOS switch for each pixel using the gate line from the shift register. An open-circuit fault occurring in the gate line or data line will make it impossible to read out some pixels. These continuous pixels are called the defect line. Although defect lines are inevitable in image sensors with a large photosensitive area, correcting them by software based on values of the surrounding pixels makes it possible to eventually acquire images with no defects. Charges leaking out of a defect line might increase the output of the pixels adjacent to the defect line. This phenomenon can also be corrected by software. If using a flat panel sensor with a pulsed X-ray source, then setting the flat panel sensor to external trigger mode will be convenient. In external trigger mode, inputting an external trigger signal to the flat panel sensor allows reading out the charges that have been kept accumulated in the photodiodes up until then. The charges are in this case continually accumulated until an external trigger signal is input. To acquire an image in synchronization with the pulsed X-ray source, the X-ray source must emit X-rays at the appropriate trigger intervals. Figure 5-17 shows a timing chart for acquiring images with pulsed X-rays using an external trigger signal. Here, an external trigger signal is input prior to pulsed X-ray emission, and starts readout of charges integrated in the photodiodes up until that time (). Readout of the integrated charges ends after Tprdy from the rising edge of the external trigger signal (), and the photodiodes are reset. Refer to the datasheet for information on other parameters. Tprdy = Tvd + Tvc - Tvdpw ......... (7) Tvdpw: period during Vsync+ is at low level in internal trigger mode Pulsed X-rays are emitted in the period between and (rising edge of the next external trigger signal) on the timing chart. The next external trigger signal is input after the X-ray emission. The operation of to then repeats. [Figure 5-17] Timing chart Frame time = Integration time (Tvc to 10 s) Image correction ExtTrg(TTL) Flat panel sensors utilizing the latest CMOS process technology and CDS circuits can acquire images with very high uniformity, yet they also offer an even higher level of image quality by software correction. Vsync+ Tvd Tvc - Tvdpw Thd + Tdd Data1-12 (14 bits) Tprdy X-ray emission #1 X-ray emission X-ray image by X-ray emission #1 X-ray image Precautions Flat panel sensors deteriorate due to X-ray irradiation. After long term use or after use under large radiation doses, the sensor sensitivity decreases and the dark video output increases. Coping with this deterioration requires correcting the image by software to meet the desired detection accuracy, as well as periodically replacing the flat panel sensor as a consumable part. 5-6 Applications X-ray imaging using pulsed X-ray source In most X-ray imaging using a continuous X-ray source, there is no need to synchronize the detector with the X-ray source during use. However, in general, when using a pulsed X-ray source that emits a high radiation dose in a short time compared to continuous X-ray sources, the detector must be synchronized with the emission timing of the X-ray source to acquire an image. 20 KACCC0460EA Acquiring enlarged images of small objects Flat panel sensors can acquire an enlarged image since they capture images with no distortion and have high resolution. The image magnification is expressed by equation (8). Magnification = D1 ............ (8) D2 D1: distance between X-ray source focal point and flat panel sensor D2: distance between X-ray source focal point and object If the distance between the flat panel sensor and X-ray source is fixed, then the magnification will increase as the object is brought closer to the X-ray source. During enlargement, the image becomes fuzzier as the focal spot size of the X-ray source becomes larger. This means that using an X-ray source with a small focal spot size will yield sharp, clear images even when enlarged. [Figure 5-18] Image distortion by X-ray source with different focal spot size [Figure 5-19] Concept image of X-ray diffraction D1 D2 Diffracted X-rays X-ray source with small focal spot size Flat panel sensor Parallel X-rays Object Object Focal point KACCC0462EA [Figure 5-20] X-ray diffraction pattern example Ordinary X-ray source Flat panel sensor KACCC0245EB Cone beam CT As a method for making full use of the features of flat panel sensors with a large photosensitive area, there is a cone beam CT that uses a cone beam X-ray source capable of emitting X-rays over a wide area. The cone beam X-ray source and the flat panel sensor are installed opposite each other with the object positioned in the center. Images of the object are then acquired while the X-ray source and flat panel sensor are rotated at the same speed around the object. The two-dimensional image data acquired in this way is then reconstructed by a computer to create threedimensional X-ray transmission images. The cone beam CT can also acquire three-dimensional X-ray images of large objects in a short time by using high-frame-rate flat panel sensor with a large photosensitive area. X-ray diffraction Flat panel sensors are useful for analysis of X-ray Laue diffraction method because of a large photosensitive area and high resolution. As shown in Figure 5-19, parallel X-rays irradiate the object, and interference fringes formed by the X-rays diffracted by the object are detected with the flat panel sensor. In this way high definition images equivalent to those obtained with an imaging plate can be obtained. The flat panel sensor is used for applications including structural analysis of crystals and proteins. 21