SoftCopy Display Quality Assurance Program at Texas Children's Hospital

With growing dependence on picture archiving and communication systems for viewing images, a quality assurance program to monitor the condition of workstation displays has become increasingly important. At present there is no universally accepted program for PACS, but there are groups such as DICOM Working Group 11 of the ACR-NEMA and AAPM Task Group 18 that are working on image quality guidelines for interpretation from soft-copy displays. Texas Children's Hospital (TCH) is developing our own quality assurance program. Data is being collected to determine the appropriate frequency of calibration, the useful life of the displays, appropriate manufacturers, and model-dependent limits on maximum and minimum luminance (black level), symptoms of degradation, and monitor cleanliness. Our system includes a variety of monitors manufactured by Sun, AFP, Siemens, Image Systems, Barco, and Orwin. We are presently collecting data on individual monitor luminance functions but have not yet initiated service calls based on deviation from the DICOM Part 14 Grayscale Display Function (GSDF). The GSDF was intended to produce a grayscale in which driving levels produce changes in luminance that are perceptually equivalent throughout the entire luminance range for a specific test target. Our data is based on measurements of luminance from a digital Society of Motion Picture and Television Engineers (SMPTE) test pattern, which is a standard used by many other institutions. TCH's biomedical engineer measures luminance data each month from the display of the SMPTE pattern and record the results in a spreadsheet. The engineer also makes subjective evaluations of sharpness, geometric distortion, and artifacts. When a monitor's luminance falls outside of arbitrary 10% limits of maximum or minimum luminance, then a service call is placed to the vendor. The luminance check by the biomedical engineer is used to verify both routine and unscheduled calibrations. In addition to the monthly monitor checks, the vendor calibrates the monitor every three months. The vendor checks the following: width and height of display, focus and position of image, SMPTE test pattern, and a graph of the GSDF Index produced by an automatic calibration software by the video driver card. Monitors are also cleaned occasionally for dust, fingerprints, ink, and pencil marks, which accumulate with routine use. This paper reports data collected to date on luminance drift, as well as conclusions on appropriate frequency of test and calibration, and our experience with monitor useful life.