Determining the sharpness of electronic image displays: an evaluation of three methods.

The authors evaluated 3 methods developed to assess the level of monitor cathode ray tube (CRT) sharpness. Results include a comparison of 2 observer-based methods to a charged coupled device (CCD) digital camera-based method for the purposes of CRT equipment comparison, acceptance testing, and routine CRT quality control. Three methods designed to measure a monitor's sharpness were evaluated on a single 20-inch CRT monitor. We defined signal-to-noise ratio (SNR) to be the overall signal difference measured by each method from the highest to lowest values divided by the average standard deviation of the measurements. Comparing the results of the 3 methods, the authors found that the digital CCD camera-based method provided a much higher SNR than the 2 observer-based methods and, therefore, is the preferred of the 3 methods for measuring the sharpness of CRT displays. The SNR values for the CCD, Cx and line target methods were 151.5, 4.9, and 4.3, respectively. The Cx target observer-based method (a novel target and scoring routine dubbed the "Cx" target because of its appearance) had a higher SNR than the line target observer-based method. The average time and standard deviation required to score the Cx and the line targets were 5.45 +/- 2.15 minutes and 8.34 +/- 2.95 minutes, respectively. The observer-based method results (and variability) versus the camera-based method results (and variability) indicate strong linear relationships. Exploring this finding and the optimization of the camera-based method are the subjects of future research.

Cathode ray tube quality control and acceptance testing program: initial results for clinical PACS displays.

The use of picture archiving and communication systems (PACS) and primary soft-copy interpretation in radiology is growing rapidly. The authors present a cathode ray tube (CRT) acceptance test and quality control (QC) program developed over a 5-year period on the basis of experience with multiple PACS and CRT vendors. The CRT QC procedures address monitor cleanliness and setup, qualitative image quality, and quantitative luminance and color measurements. Required materials include a photometer with luminance and color probes and 100%-video, flat-field window and test images from the Society of Motion Picture Test Engineers (SMPTE). Luminance was found to change over time for all gray-scale CRTs examined, which necessitated quarterly recalibration. The phosphor color of these monitors was also found to change, but changes were consistent and slow enough to warrant only annual measurements. Color measurements were found to be especially useful at initial setup and for CRT replacement. Use of this program allowed standardization of absolute luminance of individual CRTs, matching of phosphor color for multimonitor workstations, and systematic tracking of image artifacts. Implementation of a QC program is strongly recommended owing to the dynamic nature of CRT displays.

Measurement of the dose deposition characteristics of x-ray fluoroscopy beams in water.

The purpose of this study was to characterize the x-ray dose distribution of fluoroscopy beams by measuring their percent depth dose curves and lateral dose profiles in a water phantom. Percent depth dose curves were measured near the surface with an Attix parallel plate chamber and deep within the water phantom with a Farmer-type cylindrical chamber. Percent depth dose curves were compared to published data where applicable. Lateral dose profiles were measured at depths of 2, 5, 10, and 15 cm in phantom with a Farmer chamber. Pulsed, 50 mA x-ray beams with peak tube potentials of 60, 80, 100, and 120 kV and half value layers of 1.89, 2.52, 3.20, and 4.09 mm Al, respectively, were investigated.

Effects of x-ray spectra on the DQE of a computed radiography system.

The effect of incident x-ray beam quality on the measured detective quantum efficiency (DQE) of a computed radiography system was investigated. The incident x-ray beams used had peak tube potentials of 70, 95, and 120 kVp, were filtered with various thicknesses of a "patient equivalent phantom" (PEP), aluminum, and copper, and provided a consistent exposure to the storage phosphor. For each peak tube potential and filter combination, the one-dimensional modulation transfer function and noise power spectrum were measured and the square of the incident signal-to-noise ratio was estimated. The spatial frequency dependent DQE was calculated from these data. The DQE was integrated to provide an overall estimate of the efficiency and frequency response of the computed radiography system for the various x-ray beams. There was found to be a wide range of integral DQE (IDQE) values for the peak tube potential and filter combinations used. For example, the IDQE ranged from 3.0 to 0.9 mm(-2) using the peak tube potential and filter combinations 70 kVp with 5.1 cm PEP and 120 kVp with 30.3 cm PEP, respectively. Finally, peak tube potential and filter combinations 70 kVp with 10.2 cm PEP and 120 kVp with 20.2 cm PEP were chosen as standard x-ray beams that will be used at our facility to measure the DQE of digital radiographic imaging systems for evaluation and acceptance testing.

Image quality evaluation of a desktop computed radiography system.

The modulation transfer function (MTF), noise power spectrum (NPS), and detective quantum efficiency (DQE) of the Lumisys ACR-2000 desktop computed radiography (CR) reader were measured and compared to equivalent measurements acquired from a Fuji AC-3 CR system. The one-dimensional (1D) MTF was measured from an image of a sharp edge and the 1D NPS was derived from a 2D NPS measured from a uniform field exposure. The energy dependent ideal input signal to noise ratio of the incident x-ray beams was estimated using published x-ray spectra and attenuation coefficients. Measurements were acquired using Agfa, Fuji, and Kodak storage phosphor plates and it was concluded that use of the Fuji plates resulted in the highest system DQE for the ACR-2000. The DQE was measured using exposures of 0.10, 1.0, and 10.0 mR from 70 and 120 kVp x-ray beams filtered with aluminum. The DQE of the Lumisys ACR-2000 was lower than that of the Fuji AC-3.

Optimization of a contrast-detail-based method for electronic image display quality evaluation.

The authors previously reported a general technique based on contrast-detail methods to provide an overall quantitative evaluation of electronic image display quality. The figure-of-merit reflecting overall display quality is called maximum threshold contrast or MTC. In this work we have optimized the MTC technique through improvements in both the test images and the figure-of-merit computation. The test images were altered to match the average luminance with that observed for clinical computed radiographic images. The figure-of-merit calculation was altered to allow for contrast-detail data with slopes not equal to -1. Preliminary experiments also were conducted to demonstrate the response of the MTC measurements to increased noise in the displayed image. MTC measurements were obtained from five observers using the improved test images displayed with maximum monitor luminance settings of 30-, 50-, and 70-ft-Lamberts. Similar measurements were obtained from two observers using test images altered by the addition of a low level of image noise. The noise-free MTC and MTC difference measurements exhibited standard deviations of 0.77 and 1.55, respectively. This indicates good measurement precision, comparable or superior to that observed using the earlier MTC technique. No statistically significant image quality differences versus maximum monitor luminance were seen. The noise-added MTC measurements were greater than the noise-free values by an average of 4.08 pixel values, and this difference was statistically significant. This response is qualitatively correct, and is judged to indicate good sensitivity of the MTC measurement to increased noise levels.