AAPM/RSNA Tutorial on Equipment Selection: PACS Equipment Overview

Transcription

1 IMAGING & THERAPEUTIC TECHNOLOGY 879 AAPM/RSNA Tutorial on Equipment Selection: PACS Equipment Overview Display Systems 1 Aldo Badano, PhD Display systems are key components of the digital radiology department. Current display systems for medical imaging are based on cathode-ray tubes (CRTs) or active-matrix liquid crystal displays (AMLCDs). The CRT is a cathodoluminescent display: Light is generated by exciting a luminescent material with energetic electrons. AMLCDs are light-modulating devices that form the image in the screen by controlling the transparency of individual display pixels. Many image quality aspects of CRTs are determined by the way the pixel luminance is generated in the cathodoluminescent screen. The resolution properties of AMLCDs are much better than those of CRTs. In CRT devices, phosphor granularity and raster scanning patterns are the main components of spatial noise. In AMLCDs, the most notable feature of the noise characteristic is the subpixel structure of complex pixel designs used in medical displays. The small-spot contrast of CRTs is dominated mainly by veiling glare and reflections of ambient illumination. In addition to display reflectance, the contrast of medical AMLCDs is affected by crosstalk and by variations of the luminance at off-normal viewing angles. Abbreviations: AMLCD active-matrix liquid crystal display, CRT cathode-ray tube Index terms: Cathode ray tubes Computers Images, display Images, quality 2004; 24: Published online /rg From the Office of Science and Technology, Center for Devices and Radiological Health, Food and Drug Administration, Twinbrook Pkwy, HFZ-142, Rockville, MD From the AAPM/RSNA Tutorial on Equipment Selection at the 2002 RSNA scientific assembly. Received May 13, 2003; revision requested July 14 and received August 7; accepted August 21. Address correspondence to the author ( agb@cdrh.fda.gov). The mention of commercial products herein is not to be construed as either an actual or implied endorsement of such products by the Department of Health and Human Services. This is a contribution of the Food and Drug Administration and is not subject to copyright.

2 880 May-June 2004 RG f Volume 24 Number 3 Introduction The effectiveness of all diagnostic imaging modalities that use display devices is affected by the quality of the display system. Display image quality defines the relationship between the information contained in the image and the information conveyed to the observer through a luminance field in the screen. When all available information is transferred, the display system is considered to provide full imaging fidelity. Display systems always degrade the information content of the signal due to many limitations. Nevertheless, when the information conveyed matches the limitations of the observer s visual system, the display system can be described as a high-fidelity system, even when the system degrades image quality beyond the human visual capabilities (1). Current display offerings for diagnostic radiology systems are based on two competing technologies: the cathode-ray tube (CRT) and the active-matrix liquid crystal display (AMLCD). CRT technology is based on the generation of light by excitation of a cathodoluminescent phosphor by using focused, energetic electron beams. The AMLCD technology, based on an active array of transistors driving liquid crystal modulators, is a more recent technology. In this article, we briefly review the differences between these technologies in current offerings for monochrome medical display systems. Finally, we cover the key factors affecting display image quality. Technologies Many display concepts are constantly being reported in major display industry forums. However, only a subset of those are pursued to become efficient devices designed to satisfy the requirements of demanding applications. Today, only two technologies deliver the quality required to display radiographic images: the CRT and the AMLCD. Cathode-Ray Tubes The CRT is a cathodoluminescent display: Light is generated by exciting a luminescent material with energetic electrons (2,3). An electron gun located in the back of the device emits an energetic beam, which strikes a phosphor screen within a small spot. Electrons are extracted from the cathode by thermal emission from low-surface-potential materials (typically metallic oxides). The electron beam generated at the cathode is then accelerated, deflected, and focused by a series of electrostatic lenses and deflection coils (3). When high luminance is needed on a regular basis, the depletion of available electrons from the cathode material leads to cathode aging and eventually to image quality degradation and device failure. This can be remediated (up to a certain point) by adjusting the beam current over the lifetime of the CRT. Otherwise, the temporal stability of electron extraction from the cathode material in high-performance CRTs can be improved by using dispenser cathodes, which consist of a porous pellet impregnated with emissive oxide material (3). The emission from a dispenser cathode decreases to only about 95% during the first 3,000 hours under heavy cathode loading, whereas the emission from standard oxide cathodes drops to about 60% under the same conditions. The emissive structure is a key component of CRTs that greatly affects their image quality. It consists of all those elements responsible for the generation and delivery of light (see fig 7 of reference 4). Emissive structures vary according to the type of CRT. In general, they consist of a conductive coating (normally a thin aluminum overcoat), a cathodoluminescent phosphor (5), a black matrix layer, a glass faceplate, and sometimes, an antireflective coating. The choice of a particular phosphor for a medical CRT is an important element to consider when comparing monitors. The typical choices include P45 (a single-component phosphor) and P104 (a blended phosphor). These two phosphors differ in their luminous efficiency (percentage of luminance compared with that of a standard phosphor [P4] under specified conditions) and in their noise textures. P104 phosphors are about 54% more efficient than P45 screens. However, P104 phosphors are made from a mixture of grains of different size and color, which causes a granular appearance and affects the perceived noise (6,7). Phosphors degrade over time due to material changes in regions of high electron bombardment and high current density. The corresponding decrease in the brightness needs to be corrected for over the useful lifetime of the monitor by increasing the beam current. The maximum luminance of a CRT with a P45 phosphor is more stable and needs less adjustment of the electron beam current over the lifetime of the monitor when compared with that of a P104 phosphor screen.

3 RG f Volume 24 Number 3 Badano 881 Active-Matrix Liquid Crystal Displays As opposed to CRTs, AMLCDs are light-modulating devices that form the image in the screen by controlling the transparency of individual display pixels (8). The base of this technology is the liquid crystal material, which exhibits properties typical of solids (ie, a highly ordered molecular arrangement) as well as properties associated with liquids (ie, viscosity) (9). Liquid crystal materials are typically long organic molecules with delocalized charge that tend to orient themselves along a main axis, forming a unique spatial configuration determined by elasticity, viscosity, and deformation constants. To modulate light transmission, the orientation of the molecules can be controlled with an external electric field. With the help of polarizer films, which allow transmission of light when the polarization vector and the axis of the film are aligned, liquid crystal cells can be designed to transmit (normally white) or block (normally black) light. In addition to the top and bottom substrates, liquid crystal display pixel structures require alignment layers, polarizer films, and electrodes. The main components of a typical medical AMLCD stack are shown in figure 8 of reference 4. The gap between the substrates (on the order of a few microns) is maintained by spherical glass beads, which act as spacers. Because of the multitude of elements that light needs to go through before generating an image in the front screen, liquid crystal displays are intrinsically inefficient devices. Typically, only 3% 5% of the total light generated by the backlight is seen at the front face of color liquid crystal displays. This fraction is higher for monochrome devices (8% 15%) due to the lack of absorption in the color filters. The modulation of the pixel luminance is achieved by controlling the voltage at each individual pixel. High-resolution displays used in diagnostic radiology with a large number of rows and columns (high pixel density) require active addressing methods with an array of nearly ideal switches to allow fast and accurate control of the pixel luminance. In AMLCDs, the active pixel element is typically a hydrogenated amorphous silicon (a-si:h) thin-film transistor (TFT), which is usually located in one of the corners of the pixel layout. Since the TFT circuitry is shielded from the high illumination coming from the backlight by an opaque coating, light is not transmitted over the TFT area. In addition, certain pixel areas can have very low light transmission (ie, metal electrode lines). The fraction of the total pixel area that allows transmission of light is called the aperture ratio. In consumer product displays, the aperture ratio can be as small as 50%, whereas in high-performance displays it can be as high as 80%. The aperture ratio affects the display power requirements, the control of the pixel luminance, and the noise characteristics of the panel. Factors That Affect Image Quality Display systems for radiology consist of a display device and a display driver. The specifications given for a system are valid only for that particular combination. As an example, one aspect of display image quality that strongly depends on the quality of the driver is the accuracy of the grayscale representation. The relationship between image values and screen luminance is determined by the gray-scale presentation function. The digital-to-analog converter (DAC) in the display controller determines the ability to finely modify the intrinsic response of the device to match a desired luminance response. Conventional controllers with 8-bit DACs have limited control over the display gray-scale function. This is particularly relevant for medical AMLCDs, where the gray-scale resolution is also affected by the intrinsic properties of the liquid crystal pixels, which are often limited to an 8-bit scale in the luminance output. In this case, a deeper gray-scale resolution can be achieved by subpixel modulation or by temporal modulation. Subpixel modulation uses the subpixel regions of AMLCDs originally designed for color applications to generate a look-up table, which provides additional depth to the gray scale (10,11). In temporally modulated AMLCDs, the actual pixel luminance is the combined luminance of two distinct luminance levels in two consecutive frames. Because the frame rate is high, human observers cannot discriminate between consecutive frames and therefore experience an average pixel luminance. Sharpness Many image quality aspects of CRTs are determined by the way the pixel luminance is generated in the cathodoluminescent screen. When an image is displayed, the scanning electron beam is required to modulate its intensity according to the gray-scale values representing the image. If large changes in image values (which will be translated into large changes in beam current and luminance output) are present, the electronics should be capable of modulating the beam with a time constant smaller than the time needed for the beam to excite the phosphor at that pixel location.

4 882 May-June 2004 RG f Volume 24 Number 3 Table 1 Summary of Noise Components in Medical CRT and AMLCD Devices Type of Noise Component CRT AMLCD Spatial (fixed pattern) Phosphor granularity Nonuniformity of the liquid crystal Nonuniformity of the scan Thickness variations Raster Spacers Black mask (color) Subpixel structure Black mask (color) Temporal (random) Flicker and jitter Flicker Temporal (correlated)... Image lag or ghosting Therefore, the bandwidth requirements of the CRT signal amplifiers depend on the pixel array size. At low luminance, CRT spot sizes vary from 0.15 to 0.20 mm. The large beam current needed to generate higher luminance determines a larger spot size ( mm) due to the divergence of the beam caused by electrostatic repulsion. The spot size is not constant across the screen but increases at the edges relative to the center. To achieve uniform spot sizes, a dynamic focus adjustment performed by using deflection information can greatly improve the resolution uniformity of the monitor. On the other hand, the resolution properties of AMLCDs are much better than those of CRTs. Spatial modulation transfer functions measured with line patterns have been reported with close to ideal response up to the Nyquist frequency associated with the display pixel size (12). Table 2 Small-Spot Contrast Ratios for CRT and AMLCD Display Devices Display Device Noise Noise sources in a display device can be cataloged into random spatial and temporal variations, fixed-pattern spatial variations, and correlated temporal variations. Table 1 presents a summary of noise sources with examples for CRT and AMLCD devices. Spatial noise in a display device can reduce the detectability of small, low-contrast image features. The characteristics of spatial noise can be appreciated by using a magnifier lens to view the light emission pattern from a region with uniform, midgray brightness (Fig 1). In CRT devices, phosphor granularity and raster scanning patterns are the main components of spatial noise. Scanning patterns are regular periodic variations in the luminance, whereas granularity consists of random variations in the luminance. In AMLCDs, the most notable feature of the noise characteristic is the subpixel structure of complex pixel designs used in medical displays (Fig 2). This periodic structure introduces a highfrequency noise component, which interferes with traditional methods for measuring the noise characteristics of displays (14) and can affect the performance of visual tasks. Factors That Affect CRT Contrast From the list of many performance issues associated with image quality in display systems, the ability of the CRT device to achieve a large smallspot contrast ratio (C ss ) merits attention (15). The small-spot contrast ratio is defined as follows: C ss L w L b L ss L b, Small- Spot Contrast Ratio Medical AMLCD (Planar C3) (13,17) 750 Medical CRT (Clinton DS2000) (15) 152 Color AMLCD (Silicon Graphics SW1600) (15) 145 Medical CRT (Siemens Simomed) (15) 141 Medical CRT (Image Series M24L) (18) 89 Color CRT (Sony Trinitron Ultrascan) (15) 48 Color CRT (Hitachi Megascan) (18) 25 Note. Small-spot contrast ratios were measured for a 10-mm-diameter dark spot by using a collimated luminance probe and the methods described in references 15 and 16. A circular spot was used for CRTs, and a square spot was used for AMLCDs. where L w is the luminance from the small spot at the maximum luminance setting, L b is the background luminance, and L ss is the luminance from the small black spot (16). This metric relates to the ability to modulate signal in dark areas of the screen with bright areas elsewhere in the screen. Typical values for C ss of CRTs are shown in Table 2. The small-spot contrast of CRTs is

5 RG f Volume 24 Number 3 Badano 883 Figure 1. (a, b) Photographs of 30-mm-square regions of P104 (a) and P45 (b) CRT screens show the different appearances of noise due to phosphor granularity. (c) Photograph of the screen of a monochrome medical AMLCD, obtained at the same magnification, shows a fixed regular pattern due to the subpixel structure. Figure 2. Pixel structure for a dual-domain AMLCD. Individual display pixels consist of six subpixel regions in a chevron arrangement, which is determined by the dual domain and three color stripes. The simplified equivalent circuit shows one thin-film transistor (TFT) per color subpixel. C capacitor. (Adapted and reprinted, with permission, from reference 13.) Figure 3. Schematic of the three sources of veiling glare in CRTs: light diffusion, light leakage, and electron backscattering. AR antireflective coating. dominated mainly by veiling glare and reflections of ambient illumination. Veiling Glare. Veiling glare is commonly associated with multiple light scattering processes taking place in the emissive structures of CRTs, causing a contrast reduction, which is most significant in low luminance regions surrounded by bright areas. Figure 3 is a schematic depiction of the sources of veiling glare in medical CRTs:

6 884 May-June 2004 RG f Volume 24 Number 3 Figure 4. Specular and diffuse reflections for a CRT. The thick lines indicate the position of the electron beam and the luminance that it generates when it impinges on the phosphor layer. The specular reflections occur mostly at the front surface of the faceplate. The reflective coating, which is designed primarily to increase the light output of the phosphor, also increases the diffuse component of the display reflections. optical scattering, light leakage, and electron backscattering. Color CRTs typically have a lower display image quality when compared to monochrome CRTs with similar electron optics design. In addition to increasing the degradation in contrast by veiling glare, the light and electron scattering processes that take place within the emissive structure contribute to degrade color saturation. Color purity is obtained by increasing optical absorption in the emissive structure and by reducing electronic glare using low backscattering materials as mask coatings (19 21). Reflections of Ambient Light. The reflections of ambient light from CRT devices can be represented by the addition of a specular and a diffuse component (Fig 4) with different effects on the quality of the image displayed. More generally, reflections have to include a third component called haze, which becomes important in flatpanel displays. Transmission through the faceplate of medical monitors is typically 20% 50% to reduce reflections from ambient light (Fig 5). The glass absorption reduces veiling glare through dampening optical scattering within the faceplate (16). Medical monitors of good quality typically have a thin-film surface coating that provides conduction (to eliminate static charge and reduce dust collection), abrasion resistance, and antireflective properties. It has also been shown that antireflective coatings reduce veiling glare in CRTs (22). However, by decreasing the reflection of incident light, antireflective coatings may increase diffuse reflections, since more light enters the faceplate. The effectiveness of antireflective coatings is associated with a compromise between the specular and diffuse components of ambient light reflection. Factors That Affect AMLCD Contrast Owing to the thin faceplate that AMLCD and flat-panel displays in general have, these devices do not suffer from veiling glare (23). However, in addition to display reflectance, the small-spot contrast (C ss ) of medical AMLCDs is affected by crosstalk and, most importantly, by the variations of luminance at off-normal viewing angles. Crosstalk. Crosstalk is a general term used to describe two phenomena that degrade display contrast. On the one hand, optical crosstalk is a short-range effect with a characteristic distance of less than 10 pixels. On the other hand, electronic crosstalk has complex spatial characteristics (13,24) that depend on orientation (vertical vs horizontal wiring scheme). Electronic crosstalk is associated with unwanted modification of the pixel voltage effectively applied to the liquid crystal cell caused by incomplete pixel charging, leakage currents in the thin-film transistor, and parasitic capacitive coupling. Accordingly, crosstalk is more important in large panels with high spatial and gray-scale resolution (25,26). Methods used for reducing electronic crosstalk employ modified driving techniques to bracket the desired voltage at each individual pixel in the active-matrix array. The effect of crosstalk is seen as a shift in the display pixel luminance in a region where there are significant variations in the desired luminance across the vertical or horizontal direction. Figure 6 shows small-spot contrast ratio measurements for a variety of medical CRTs and AMLCDs. For a spot size of 10 mm, the measured contrast ratio for CRTs is lower than 100, whereas medical AMLCDs can achieve ratios of 800 due to the lack of veiling glare and controlled crosstalk (Table 2).

7 RG f Volume 24 Number 3 Badano 885 Figure 5. To reduce reflections from ambient light, transmission through the faceplate of a medical monitor is typically If a transmission of 0.3 is assumed, diffuse reflections are reduced to at least 0.09, resulting in improved black levels. The display brightness diminishes only to 0.3. Absorption also reduces veiling glare by dampening the scattering within the faceplate. Figure 6. Contrast ratio (CR) measurements for CRTs and AMLCDs as a function of the dark spot size. Dashed lines monochrome CRTs, dotted line color CRT. (Adapted and reprinted, with permission, from reference 27.) Figure 7. The light transmission and intensity modulation that occur in an AMLCD result in a non-lambertian luminous emission from the screen. The electro-optic effect in the liquid crystal cell that determines the pixel luminance is highly dependent on the relative orientation of the input light (from the backlight) and the liquid crystal molecules and polarizer films in the liquid crystal display stack, as well as the path length associated with each direction of emission. I intensity. Non-Lambertian Emission. CRTs, like most emissive displays (28), emit light in such a way that the angular luminous intensity approximately follows Lambert s cosine law. Consequently, the display luminance remains approximately constant across all viewing directions, which is a fundamental property of Lambertian surfaces. However, this is not the case in AMLCDs (Fig 7). The luminance and contrast of AMLCDs vary with the viewing direction. In some AMLCDs, at large off-normal angles, the variations can be severe enough to cause an inversion of the gray scale, a

8 886 May-June 2004 RG f Volume 24 Number 3 Figure 8. Contrast ratio measurements for a medical AMLCD. In this case, the contrast ratio is the ratio of maximum to minimum luminance for a 20% region in the midgray background. Figure 9. Changes in luminance (a) and contrast (b) for a medical AMLCD at different angles along a diagonal direction of viewing (29). The data points labeled 0 correspond to perpendicular viewing. JND just noticeable difference, L luminance. condition that is unacceptable in diagnostic display devices. Figure 8 shows measured contrast variations for a medical AMLCD (29). Figure 9 shows the same luminance and contrast data from Figure 8, as viewed from different angles along the bottom-right to top-left diagonal direction. The measured luminance increases with respect to normal viewing by a factor of about 10 when the viewing direction moves along the oblique axes. The slope of the curve in Figure 9a, which is associated with luminance contrast, is significantly reduced in the low luminance region. This is confirmed by analyzing the contrast response plots in Figure 9b. The available contrast at each JND (just-noticeable-difference) index decreases in the diagonal directions. (The JND index refers to the corresponding change in image data associated with a just noticeable difference in the screen luminance, as given by a specific gray-scale presentation model [ie, the DICOM Grayscale Display Standard Function] [30]). Even at 40, the contrast response in the low luminance region falls outside of the 25% tolerance limits (indicated by dotted lines). The effect of these changes on lesion detectability is not well understood. However, by using measured luminance data such as those presented in Figure 9, it is possible to simulate the effect of angular variations in luminance in AMLCDs on images (Fig 10). These simulated images can then be used in model and human observer studies (31). The implications of the viewing angle problem in medical imaging monitors are twofold. First, a single user of the device will experience its effect when looking at different areas of the display screen, depending on the dimension of the screen surface (which can reach more than 30 cm on one of the sides). In this scenario, the more severe changes in the luminance presentation curve and available contrast associated with different viewing directions are likely to occur between the center and the corners of the screen (Fig 11). The second aspect of this problem arises when more than one individual is looking at the same image displayed on the same screen or in the complementary case of many displays in a tile arrangement. In this case, the variations can be much larger due to the larger angles involved.

9 RG f Volume 24 Number 3 Badano 887 Figure 10. Effect of viewing angle on white noise images. (a) Image shows the pattern as seen in the perpendicular direction. (b, c) Corresponding images obtained with the technique described in the text show the changes in the gray scale luminance relationship due to a viewing angle of 45 in the horizontal plane (b) and in one of the diagonal planes (c). Figure 11. Effect of viewing angle on the luminance calibration functions of AMLCDs for a fixed centered observer. The luminance output for perpendicular viewing (right graph) is distorted at the corners and the edge of the display screen (left graphs). If a lesion were present in these screen locations, its detectability would be different than if it were in the center. Several solutions have been developed to compensate for the angular variations of the display luminance. The approaches come from recognizing that the anisotropy of the light modulation is the dominating factor in defining the viewing angle characteristics of the device. The most commonly used solutions include compensation films, multiple pixel domains, and modified liquid crystal alignment modes. In current AMLCD designs, many if not all of these classes are combined and employed in the same device. The benefits of using one of these solutions are compounded by the addition of other solutions. However, each of these solutions has drawbacks. Compensation films are designed to compensate for the anisotropy introduced by the liquid crystal alignment with respect to the different directions of light transmitted by the cell (32,33). Since the films are static in the sense of not being dynamically adjustable with the pixel gray level, the compensation is optimal only for a single luminance level. The viewing angle of AMLCDs can be significantly improved by dividing the pixel area into multiple subpixel domains having different liquid crystal director orientation. Since each domain has an asymmetric response as a function of the viewing direction, the net effect is an average emission, which tends to reduce the luminance variations with angle (34,35). The main challenge associated with this technology is the stability of the liquid crystal alignment and increased fabrication costs.

10 888 May-June 2004 RG f Volume 24 Number 3 If the liquid crystal molecules remain in the display plane for all gray-scale states, the asymmetry for the different angles is minimized. This is the basis of the in-plane switching (IPS) structure used in many medical imaging AMLCDs. In IPS, the pixel electrodes are located on the same bottom glass plate (as shown in fig 8 of reference 4) (36,37). Although this improves the angular constancy of the luminance, it also reduces the transmission through the liquid crystal stack due to the presence of interdigitated electrodes, resulting in a lower available luminance. Another design that improves the viewing angle performance is the vertically aligned (VA) liquid crystal mode. This arrangement can be achieved by oblique electrical fields with displaced electrodes or with pyramid-shaped protrusions on both substrate plates (38). Conclusions The display system is a key component of a digital radiology implementation with respect to ensuring appropriate image quality. The choice of a display technology for a particular imaging application is affected by the fidelity of the image presentation. The selection of a system is affected also by the display size, ergonomic considerations, cost, service availability, and lifetime. As technology continues to advance and new display solutions become available, systems capable of delivering the image quality needed for multimodality applications will become widespread. Techniques such as holography, stereo, retinal, and large projection displays, along with new fundamental display technologies such as the lightemitting organic display, will allow full color (both for imaging techniques and for assistance tools) and fast dynamic applications with excellent gray-scale and spatial resolution. Some technical challenges have already come to light when these high-performance systems are required to be portable, to deliver images more conveniently to radiologists and other health care providers. Acknowledgments: The author thanks the many collaborators who have contributed to the material reviewed in this article, including M. J. Flynn, S. Martin, J. Kanicki, R. J. Jennings, R. M. Gagne, K. J. Myers, E. Muka, H. Blume, and K. Compton. References 1. Flynn MJ, Kanicki J, Badano A, Eyler WR. Highfidelity electronic display of digital radiographs. 1999; 19: Keller PA. The cathode-ray tube: technology, history and applications. New York, NY: Palisades Press, Compton K. Image performance in CRT displays. Bellingham, Wash: SPIE Press, Samei E, Siebert JA, Andriole K, Badano A, Crawford J, Reiner B. General guidelines for purchasing and acceptance testing of PACS. Radio- Graphics 2004; 24: Ozawa L. Cathodoluminescence: theory and applications. Tokyo, Japan: Kodansha, Muka E, Mertelmeier T, Slone RM. Impact of phosphor luminance noise on the specification of high-resolution CRT displays for medical imaging. Proc SPIE 1997; 3031: Krupinski EA, Roehrig H. Pulmonary nodule detection and visual search: P45 and P104 monochrome versus color monitor displays. Acad Radiol 2002; 9: Depp SW, Howard WE. Flat-panel displays. Sci Am 1993; 3(40): Collings PJ. Liquid crystals: nature s delicate phase of matter. Princeton, NJ: Princeton University Press, Flynn MJ, Compton K, Badano A. Luminance response calibration using multiple display channels. Proc SPIE 2001; 4319: Wright SL, Millman S, Wu C, et al. Color and luminance management for high-resolution liquidcrystal displays. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 2003; Blume HR, Steven PM, Cobb ME, et al. Characterization of high-resolution liquid crystal displays for medical images. Proc SPIE 2002; : Martin S, Badano A, Kanicki J. Characterization of a high quality monochrome AM-LCD monitor for digital radiology. Proc SPIE 2002; 4681: Badano A, Drilling S, Imhoff B, Jennings RJ, Gagne RM, Muka E. Noise in flat-panel displays with sub-pixel structure. Med Phys (in press).

11 RG f Volume 24 Number 3 Badano Badano A, Flynn MJ, Kanicki J. Accurate smallspot luminance measurements. Displays 2002; 23: Badano A, Flynn MJ. A method for measuring veiling glare in high performance display devices. Appl Opt 2000; 39: Martin S, Badano A, Kanicki J. High-resolution medical imaging AM-LCD: contrast performance evaluation. In: Proceedings of the International Display Research Conference. San Jose, Calif: Society for Information Display, 2002; Flynn MJ, Badano A. Image quality degradation by light scattering in display devices. J Digit Imaging 1999; 12: de Vries GC. Contrast-enhancement under low ambient illumination. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1995; van Oekel JJ. Improving the contrast of CRTs under low ambient illumination with a graphite coating. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1995; van Oekel JJ, Severens MJ, Timmermans GMH, et al. Improving contrast and color saturation of CRTs by Al 2 O 3 shadow mask coating. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1997; Badano A, Flynn MJ, Muka E, Compton K, Monsees T. Veiling glare point-spread function of medical imaging monitors. Proc SPIE 1999; 3658: Badano A, Flynn MJ. Monte Carlo modeling of the luminance spread function in flat panel displays. In: Proceedings of the International Display Research Conference. San Jose, Calif: Society for Information Display, 1997; Badano A, Kanicki J. Characterization of crosstalk in high-resolution active-matrix liquid crystal displays for medical imaging. Proc SPIE 2001; 4295: Libsch FR, Lien A. A compensation driving method for reducing crosstalk in XGA and higherresolution TFT-LCDs. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1995; Libsch FR, Lien A. Understanding crosstalk in high-resolution color thin-film-transistor liquid crystal displays. IBM J Res Dev 1998; 42: Badano A. Principles of cathode-ray tube and liquid crystal display devices. In: Samei E, Flynn MJ, eds. Syllabus: a categorical course in diagnostic radiology physics. Oak Brook, Ill: Radiological Society of North America, 2003; Lee SJ, Badano A, Kanicki J. Monte Carlo modeling of organic polymer light-emitting devices on flexible plastic substrates. Proc SPIE 2003; 4800: Badano A, Flynn MJ, Martin S, Kanicki J. Angular dependence of the luminance and contrast in medical monochrome liquid crystal displays. Med Phys 2003; 30: Digital Imaging and Communications in Medicine (DICOM), part 3.14: grayscale standard display function. Rosslyn, Va: National Electrical Manufacturers Association, Badano A, Gallas BD, Myers KJ, Burgess AE. Effect of viewing angle on visual detection in liquid crystal displays. Proc SPIE 2003; 5029: Hoke CD, Mori H, Bos PJ. An ultra-wide-viewing angle STN-LCD with a negative-birefringence compensation film. In: Proceedings of the International Display Research Conference. San Jose, Calif: Society for Information Display, 1997; Mori H, Bos PJ. Application of a negative birefringence film to various LCD modes. In: Proceedings of the International Display Research Conference. San Jose, Calif: Society for Information Display, 1997; M88 M Nam MS, Wu JW, Choi YJ, et al. Wide-viewingangle TFT-LCD with photo-aligned four-domain TN mode. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1997; Chen J, Bos PJ, Bryant DR, et al. Four-domain TN-LCD fabricated by reverse rubbing or double evaporation. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1995; Masutani Y, Tahata S, Hayashi M, et al. Novel TFT-array structure for LCD monitors with inplane switching mode. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1997; Wakemoto H, Asada S, Kato N, et al. An advanced in-plane switching mode TFT-LCD. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1997; Ohmuro K, Kataoka S, Sasaki T, et al. Development of super-high-image-quality vertical-alignment-mode LCD. In: Proceedings of the Society for Information Display. San Jose, Calif: Society for Information Display, 1997;