(Radiographics. 2000;20:1471-1477.)
© RSNA, 2000
IMAGING & THERAPEUTIC TECHNOLOGY |
The AAPM/RSNA Physics Tutorial for Residents 1
X-ray Image Intensifiers for Fluoroscopy
Jihong Wang, PhD and
Timothy J. Blackburn, PhD
1 From the Department of Radiology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9071. From the AAPM/RSNA Physics Tutorial at the 1999 RSNA scientific assembly. Received May 15, 2000; revision requested June 9 and received June 22; accepted June 27. Address correspondence to J.W. (jihong.wang@utsouthwestern.edu).
 |
Abstract
|
|---|
The x-ray image intensifier converts the transmitted x rays into a brightened, visible light image. Within an image intensifier, the input phosphor converts the x-ray photons to light photons, which are then converted to photoelectrons within the photocathode. The electrons are accelerated and focused by a series of electrodes striking the output phosphor, which converts the accelerated electrons into light photons that may be captured by various imaging devices. Through this process, several thousand light photons are produced for each x-ray photon reaching the input phosphor. Most modern image intensifiers use cesium iodide for the input phosphor because it has a high absorption efficiency and thus decreases patient dose. Image intensifiers come in various sizes, most having more than one input image size or magnification mode. Modern image intensifiers are specified by conversion factors, which is the measure of how efficiently an image intensifier converts x rays to light. Because of design restrictions, image intensifiers are subject to inherent and induced artifacts that contribute to image degradation. Both spatial and contrast resolution gradually decrease during the lifetime of the image intensifier because the brightness gain of an image intensifier decreases with time as the phosphor ages. A well-run quality control program for the image intensifier is needed to detect the inevitable changes in settings before they appear on clinical images.
Index Terms: Fluoroscopy • Physics
|
Introduction
|
|---|
Early fluoroscopic procedures produced visual images of low intensity, which required the radiologist's eyes to be dark adapted and restricted image recording. In the late 1940s, with the rapid developments in electronics and borrowing the ideas from vacuum tube technology, scientists invented the x-ray image intensifier, which considerably brightened fluoroscopic images (1). Commercial x-ray systems with image intensifiers were introduced in the mid 1950s. The x-ray image intensifier enabled the radiologist to visualize the output image without dark adaptation. The intensified visual image could be easily captured by film and television camera tubes. When the image intensifier was first introduced, it had a small input size and a glass vacuum case. Modern image intensifiers have input field sizes up to 57 cm in diameter with little image distortion, and the vacuum cases are usually made of metal.
An x-ray image intensifier has two major functions: (a) to intercept the x-ray photons and convert them into visible light photons and (b) to amplify or intensify this light signal. The image intensifier creates a large gain (or intensification) in luminance at the output screen compared with that at the input screen. The output screen image can be viewed with closed-circuit television or recorded with film.
In this article, we discuss the principles of operation of an x-ray image intensification system for fluoroscopy, its specific major components, their physical characteristics and relationships to overall image quality, patient dose, artifacts, and quality control.
|
Construction and Principles of Operation
|
|---|
In a modern fluoroscopy system, the image intensifier is located opposite the x-ray tube. The image intensifier is contained in a cylindrical protective case because it is a very delicate device under high vacuum and needs to be handled with care. At the entrance end of this protective case, there is usually a mechanical sensory device to prevent the image intensifier from pushing too hard on the patient or other objects, which may cause damage to the image intensifier. Image intensifiers come in various sizes depending on the specific application. Usually, the larger the image intensifier the higher the cost.
The operational principles of an image intensifier can be briefly described as follows. X-ray photons penetrate the input window of the vacuum case. The input phosphor absorbs the x-ray photons and converts them into optical photons (a phenomenon called luminescence). These optical photons are converted to photoelectrons at the photocathode. The photoelectrons are accelerated by the electric field produced by the strong electric potential difference of the image intensifier and are collected at the output phosphor. Each accelerated electron produces many optical photons at the output phosphor.
|
Image Intensifier Components
|
|---|
An image intensifier consists of the following major components: an input window, an input phosphor and photocathode, several electrostatic focusing lenses, an accelerating anode, an output phosphor screen, and a protective vacuum case (Fig 1).
Input Window
The shape and choice of material for the input window results from a compromise among many factors, such as minimizing patient distance, x-ray absorption, x-ray scatter, manufacturing cost, and mechanical strength of materials. The input side of the image intensifier usually has a convex shape and is generally made of aluminum (Z = 13). The convex shape not only minimizes the patient distance thus maximizing the useful entrance field size (2), but it also gives the image intensifier better mechanical strength under atmospheric pressure. This aluminum input window is approximately 1 mm in thickness.
Input Phosphor and Photocathode
X rays transmitted through the input window are converted into fluorescent light photons by the input phosphor. The input screen is a substrate made of aluminum coated with a phosphor layer, an intermediate coupling layer, and finally the photocathode layer (Fig 1). The thickness of the input phosphor layer is a compromise between spatial resolution and x-ray absorption efficiency. A thicker phosphor layer has higher x-ray absorption efficiency, which means more x-ray photons can be absorbed and converted to light photons in the phosphor layer. A thicker phosphor layer requires fewer x-ray photons to generate the same amount of light photons at the image intensifier output window, thus reducing patient dose. However, with a thicker input phosphor layer, more light photons are scattered laterally within the phosphor layer, thus reducing the spatial resolution. Currently, the thickness of an input phosphor layer typically measures between 300 and 450 µm, depending on the image intensifier type and technology used.
To maximize the conversion efficiency from x-ray photons to photoelectrons, the mass attenuation coefficient of the input phosphor material should be matched with the spectrum of the x rays emerging from the patient. Ideally, the light spectrum of the input phosphor should also match the sensitivity profile of the photocathode. The initial phosphor used in early image intensifiers was zinc-cadmium sulfide (ZnCdS), whereas the current phosphor of choice is cesium iodide (CsI:Na). There are several reasons for replacing ZnCdS with CsI:Na as the input phosphor material. First, the mass attenuation coefficient of CsI:Na better matches the x-ray spectrum of the radiation transmitted from the patient. Figure 2a shows the mass attenuation coefficients of the two phosphors in relation to the relative spectral distribution of the transmitted radiation from the patient. The mass attenuation peaks in CsI:Na, compared with those of ZnCdS, are more closely matched to the transmitted x-ray spectrum, thus increasing the absorption of the transmitted x-ray photons. As mentioned, increasing the absorption efficiency decreases the patient's dose. A second advantage for using CsI:Na as the phosphor is that it has a high atomic number from Cs (Z = 55) and I (Z = 53), which also results in higher x-ray absorption. Consequently, most modern image intensifiers use CsI:Na for the input phosphor material.
View larger version (34K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Spectral matching comparisons of different input phosphor and photocathode materials. (a) The mass attenuation coefficients of CsI:Na, ZnS:CdS, and the relative spectral distribution of the radiation behind the patient are plotted as the function of x-ray quantum energy. The attenuation coefficient of CsI:Na is more closely matched with that of the relative spectral distribution of radiation, thus a better choice for the input phosphor. (b) Plot illustrates the spectral sensitivity matching of the photocathode material SbCs3 to that of CsI:Na and ZnCdS. Again, CsI:Na is a better match to the photocathode material.
|
|
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Spectral matching comparisons of different input phosphor and photocathode materials. (a) The mass attenuation coefficients of CsI:Na, ZnS:CdS, and the relative spectral distribution of the radiation behind the patient are plotted as the function of x-ray quantum energy. The attenuation coefficient of CsI:Na is more closely matched with that of the relative spectral distribution of radiation, thus a better choice for the input phosphor. (b) Plot illustrates the spectral sensitivity matching of the photocathode material SbCs3 to that of CsI:Na and ZnCdS. Again, CsI:Na is a better match to the photocathode material.
|
|
The photocathode layer is made of antimony-cesium (SbCs3). To maximize the conversion efficiency from light photon to photoelectron, light emitted from the input phosphor should match the sensitivity spectrum of the photocathode. As seen in Figure 2b, CsI:Na has a better spectral match to the antimony-cesium compound (SbCs3). This is another reason why CsI:Na is a better input phosphor material than ZnCdS. The photocathode has a thickness of about 20 nm and a photoelectron production efficiency of 10%–15%. Approximately 200 photoelectrons will be created for a single 60-keV x-ray photon absorbed in the input phosphor.
In addition to its high absorption efficiency, CsI:Na can be evaporated onto the substrate in crystal needle form. These needles act like light pipes, in a manner similar to the light propagation in a fiber-optic faceplate, thus reducing cross scatter inside the phosphor screen and yielding better spatial resolution. A cross-sectional diagram of the input screen is shown in Figure 3. The CsI:Na needles are approximately 5 µm in diameter. Input phosphor screens in modern image intensifiers are approximately 300–500 µm in thickness and absorb 60%–70% at 60 keV. Because of the crystalline structure of the needles, the surfaces of the crystals, and the reflectivity of the substrate, approximately 2,600 luminescence photons are generated from each 60-keV x-ray quantum. Of these 2,600 luminescence photons, approximately 1,600 reach the photocathode.
View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Cross-sectional diagram of the input screen shows the CsI:Na crystal needles, which serve as the optical guide to the photons, preventing scattering of light photons across the needles and thus improving spatial resolution.
|
|
Electron Optics
Photoelectrons are accelerated from the photocathode to the output phosphor by the anode (Fig 1). The accelerated photoelectrons are focused down to the size of the output phosphor by a series of electrostatic focusing electrodes. The number of photoelectrons within the image intensifier will not increase: Only the speed of the photoelectrons will increase. The total current produced by these photoelectrons is approximately 600 nA (600 x 10-9 A).
The focusing electrodes are very sensitive to external electrical and magnetic fields. Extraneous electrical and magnetic fields (even the earth's magnetic field) may cause image distortions in the image intensifier. This effect must be monitored and controlled for fluoroscopes operated near magnetic resonance imaging units. Furthermore, the high voltages on the electrodes must be kept very stable to guarantee the image quality, since ripple in the voltage will be noticed as periodic variation in image diameter.
On the vacuum side of the output phosphor surface, the anode of the electron optics system has a thin aluminum film coating. This aluminum film allows electrons to pass through, but it is opaque to light photons generated on the fluorescent screen. It stops these photons from being scattered back into the image intensifier and exposing the photocathode. The film also serves as a reflector to increase the output luminance.
Output Phosphor and Window
The output phosphor of the x-ray image intensifier, which typically is called P20, is a fluorescent compound made of silver-activated zinc-cadmium sulfide (ZnCdS:Ag). The emission spectrum of P20 is at a maximum around 530 nm (green light). The P20 layer is very thin, having a thickness of 4–8 µm, and is deposited on the glass output window. Approximately 2,000 luminescence photons are generated for every accelerated 25-keV photoelectron. Because every electron was produced by one light photon, this represents a luminescence gain of 2,000. The output window is carefully designed so that the fluorescent photons reflected back to the input screen are minimized. The luminescence decay time of the output phosphor determines the temporal resolution of the image intensifier.
Image Intensifier Housing
The x-ray image intensifier is enclosed in a metal housing consisting of lead to absorb scattered radiation, mu-metal to shield the electron optics from extraneous magnetic fields, and an outer aluminum shell. On the input side of the housing, the aluminum shell protects the input window of the image intensifier. Although the housing will provide some shielding from external electromagnetic fields, the presence of strong magnetic or electrical fields too close to the image intensifier will degrade image quality.
|
Physical Characteristics
|
|---|
The input size, brightness gain, conversion factor, contrast ratio, magnification mode, and spatial resolution characterize an image intensifier. The size of an image intensifier is obviously the most visible property, and the larger the image intensifier, the larger the field of view. A large field of view allows one to visualize a larger area, which can be very helpful in some clinical procedures. However, a larger image intensifier is more difficult to make and thus more expensive.
Brightness Gain and Conversion Factor
The brightness gain comes from two sources that are completely unrelated: the minification gain and the flux gain. The minification gain is defined as the ratio of input area to the output area of the image intensifier. Because the number of photoelectrons leaving the photocathode is equal to the number striking the output phosphor, the number of photoelectrons per unit area at the output phosphor increases. The minification gain does not improve the statistical quality of the fluoroscopic image. It will not change the contrast of the image, but it will make the image appear brighter. A smaller output window size will just compress more photons into a smaller area, producing a smaller but brighter image.
Flux gain is defined as the number of photons generated at the output phosphor for every photon generated at the input phosphor. The flux gain results from the acceleration of photoelectrons to a higher energy so that they generate more fluorescent photons at the output phosphor. Each light photon generated at the input phosphor will generate approximately 100 photons at the output phosphor, resulting in a flux or luminance gain of 100. The total brightness gain of the image intensifier is the product of minification gain and flux gain (total brightness gain = flux gain x minification gain).
The size of the output window of an image intensifier is usually between 1.5 and 6.0 cm in diameter. The minification gain for a 23-cm image intensifier with an input entrance field size of 22 cm (380 cm2) and a 2-cm output window (3.14 cm2) is approximately 120. With a flux gain of approximately 100, the total brightness gain for this image intensifier would be approximately 12,000.
The original definition of brightness gain is the output luminance level (or brightness) of an image intensifier divided by the output luminance level of a Patterson B-2 fluoroscopic screen when both are exposed to the same quantity of radiation. The Patterson B-2 fluoroscopic screen was typically used for fluoroscopy before image intensifiers were introduced. If the image intensifier gives 5,000 times brighter output than the Patterson B-2 fluoroscopic screen, the brightness gain is 5,000. The drawback of using this definition is the lack of reproducibility of the Patterson B-2 screen.
The International Commission on Radiological Units and Measurements (ICRU) has recommended another method of evaluation called the conversion factor. Today, most x-ray image intensifiers are specified by the conversion factor. The conversion factor is defined as the output luminance level of an image intensifier divided by its entrance exposure rate. It is a measure of how efficiently an image intensifier converts the x rays to light. Conversion factors have units of candela per square meter per milliroentgen per second ([cd/m2]/[mR/sec]). A typical 23-cm image intensifier has a conversion factor of approximately 200 cd/m2/mR/sec. The conversion factor usually equals to 1% of the brightness gain in value. Conversion factors tend to deteriorate (decrease) as image intensifiers age, resulting in higher patient dose for older image intensifiers. The higher the conversion factor, the more efficient the image intensifier.
Contrast Ratio
The contrast ratio of an image intensifier is defined as the brightness ratio of the periphery to the center of the output window when the center portion of an image intensifier entrance is totally blocked by a lead disk. It is a measurement of the veiling glare (discussed in the section Artifacts). The contrast ratio is typically specified in two ways: large area and small detail area. The large area or 10% area contrast ratio is measured by putting a lead disk, which has a surface area equal to 10% of the useful entrance area of the image intensifier, at the center of the input surface of the image intensifier. The small detail, or 10-mm area contrast, is measured by putting a 10-mm lead disk at the center of the input surface of the image intensifier (2). Measurements are made at 50 kVp without additional filtration. At higher peak kilovoltage values, the contrast decreases because of increased x-ray and light diffusion inside the image intensifier. The lead disk must be thick enough to block all x rays incident on it. The light is then measured at the center and adjacent to the disk at the output phosphor of the image intensifier. Currently, new image intensifiers have contrast ratios in the range of 10:1 to 30:1 and 15:1 to 35:1 for the 10% and 10-mm area contrast ratios, respectively.
Magnification Modes and Spatial Resolution
Changing the voltage applied to the electronic lenses inside an image intensifier will change the magnification mode of the image intensifier. In a magnification mode, a smaller area of the input phosphor is used, giving the effect of zooming in on the image. Because the input field size is reduced, the exposure to the input phosphor must be increased to maintain a constant brightness level at the output phosphor. In fact, to maintain the same noise level, the dose quadruples when the magnification is doubled. Modern image intensifiers can perform multiple steps of magnification. It is even possible to have stepless magnification. Each magnification mode yields a different dose rate to the patient. In general, the smaller the field size, the larger the magnification, and the higher the patient dose. The image intensifier exposure rate is typically set to 30 mR/sec for the 25-cm mode, 60 µR/sec for the 17-cm mode, and 120 µR/sec for the 12-cm mode (3).
Higher magnification modes produce increased spatial resolution. The spatial resolution of an image intensifier is in the range of 4–6 line pairs per millimeter (lp/mm). The final spatial resolution of the fluoroscopic system also depends on other imaging components in the whole imaging chain.
|
Artifacts and Quality Control
|
|---|
Image intensifiers come with a variety of imperfections or artifacts, including lag, vignetting, veiling glare, pincushion distortion, and S distortion. Some of these artifacts are caused by improper calibration and can usually be corrected.
Lag
Lag is the persistence of luminescence after x-ray stimulation has been terminated. Lag degrades the temporal resolution of the dynamic image. Older image intensifier tubes had phosphors with lag times on the order of 30–40 msec. Current image intensifier tubes have lag times of approximately 1 msec. Therefore, lag in modern fluoroscopic systems is more likely caused by the closed-circuit television system than the image intensifier.
Vignetting
A fall-off in brightness at the periphery of an image is called vignetting. Vignetting is caused by the unequal collection of light at the center of the image intensifier compared with the light at its periphery. As a result, the center of an image intensifier has better resolution, increased brightness, and less distortion.
Veiling Glare
Scattering of light and the defocusing of photoelectrons within the image intensifier are called veiling glare. Veiling glare degrades object contrast at the output phosphor of the image intensifier. As mentioned, the contrast ratio is a good measure of determining the veiling glare of an image intensifier. X-ray, electron, and light scatter all contribute to veiling glare.
Pincushion Distortion
Pincushion distortion is a geometric, nonlinear magnification across the image. The magnification difference at the periphery of the image results from the projection of the x-ray beam onto a curved input surface. The distortion is easily visualized by imaging a rectangular grid with the fluoroscope. Figure 4a shows a representation of pincushion distortion.
S Distortion
Electrons within the image intensifier move in paths along designated lines of flux. External electromagnetic sources affect electron paths at the perimeter of the image intensifier more so than those nearer the center. This characteristic causes the image in a fluoroscopic system to distort with an S shape (Fig 4b). Larger image intensifiers are more sensitive to the electromagnetic fields that cause this distortion. Manufacturers include a highly conductive mu-metal shield that lines the canister in which the vacuum bottle is positioned to reduce the effect of S distortion.
Quality Control
As with any other radiologic equipment, it is crucial (and required by law in many states) to have an ongoing quality assurance and quality control program for fluoroscopic systems. Because the settings on image intensifiers drift over time, the quality of patient care may be compromised. A well-run quality control program should detect the changes in the settings before they appear on clinical images. The quality control checks must be performed on a regular basis. The checks should include some basic measurements on high-contrast spatial resolution, low-contrast resolution, and entrance exposure rates.
Both spatial and contrast resolution gradually decrease during the lifetime of the image intensifier because the brightness gain of an image intensifier decreases with time as the phosphor ages. Eventually, the image intensifier will need to be replaced.
|
Summary
|
|---|
The function of the x-ray image intensifier in the fluoroscopic imaging system is to convert the x-ray spectrum transmitted through the patient into a highly visible image. The image is produced by converting the x-ray photons into light photons at the image intensifier input phosphor, converting the visible light photons into electrons at the photocathode, accelerating and focusing the electrons through use of electrodes, and finally, converting the electrons back into visible light at the output phosphor. The intensity of the final image is several thousand times brighter than the initial image created at the input phosphor. Image intensifier tubes come in many sizes, and most have multiple magnification modes associated with them. Conventional vacuum-based image intensifiers have inherent artifacts associated with them and are sensitive to extraneous electromagnetic fields. Additional information on image intensifiers may be found in a variety of sources (4–9).
|
References
|
|---|
-
Coltman JW. Fluoroscopic image brightening by electronic means. Radiology 1948; 51:359-367.
-
de Groot PM. Image intensifier design and specifications In: Proceedings of the AAPM Summer School: specification, acceptance testing, and quality assurance of diagnostic x-ray imaging equipment—1991. New York, NY: American Institute of Physics, 1994; 429-460.
-
Chakraborty DP. Routine fluoroscopic quality control In: Proceedings of the AAPM Summer School: specification, acceptance testing, and quality assurance of diagnostic x-ray imaging equipment—1991. New York, NY: American Institute of Physics, 1994; 569-595.
-
Blume H, Colditz J, Eckenbach W, et al. Image intensifier and x-ray exposure control systems. In: Balter S, Shope TB, eds. Syllabus: a categorical course in physics—physical and technical aspects of angiography and interventional radiology. Oak Brook, Ill: Radiological Society of North America, 1995; 87-103.
-
Bushberg JT, Seibert JA, Leidholdt EM, Jr, Boone JM. The essential physics of medical imaging Baltimore, Md: Williams & Wilkins, 1994.
-
Curry TS, III, Dowdey JE, Murry RC, Jr. Christensen's physics of diagnostic radiology 4th ed. Philadelphia, Pa: Lea & Febiger, 1990.
-
Huda W, Sloan RM. Review of radiologic physics Baltimore, Md: Williams & Wilkins, 1995.
-
Krestel E, ed. Imaging systems for medical diagnostics: fundamentals and technical solutions; x-ray diagnostics, computed tomography, nuclear medical diagnostics, magnetic resonance imaging, sonography, biomagnetic diagnostics Berlin, Germany: Siemens-Aktiengesellshaft, 1990.
-
Sprawls P, Jr. Physical principles of medical imaging 2nd ed. Madison, Wis: Medical Physics, 1993.
Top
Abstract
Introduction
