CLEVELAND - With more healthcare institutions adopting PACS for image distribution, just how reliable are the cathode ray tube (CRT) soft-copy displays used for image interpretation? The answer depends a great deal on the program an institution chooses for quality assurance (QA) of its workstation displays, according to Catherine Kim Ly of Texas Children’s Hospital (TCH) in Houston.
No fewer than three groups in the U.S. are now working on image quality guidelines for interpretation from soft-copy displays: the American College of Radiology (ACR), the American Association of Physicists in Medicine Task Group 18, and Working Group 11 of the joint ACR-National Electronic Manufactures Association. In addition, there are several international organizations working on the same issues.
"Unfortunately, soft-copy diagnosis is in its infancy, and there is no general agreement of the specifics of how to ensure optimal electronic display," Ly said in a presentation Thursday at the Symposium for Computer Applications in Radiology.
Ly and her colleagues have developed a QA program for TCH that collects data on the frequency of monitor calibration, the useful life of the displays, model-dependent limits on maximum and minimum luminance, and symptoms of degradation.
During the first year of the monitor QA program, Ly and her team collected data only quarterly. "In January 2001, on the basis of poor results from the QA tests, 41 monitors had to undergo extensive recalibration. From that point, we started collecting data monthly," said Ly.
A biomedical engineer at the facility measures its 51 monitors each month for luminance data from the display of a Society of Motion Picture and Television Engineers (SMPTE) test pattern and records the data in a spreadsheet. In addition, the engineer makes an evaluation of sharpness, geometric distortion, and artifacts. "It takes a good 15 minutes to perform the tests on each monitor and record the data for later evaluation," noted Ly.
If the luminance of a monitor falls outside a 10% limit of maximum or minimum luminance, the institution puts in a service call. A vendor-performed calibration on the monitors is also conducted each quarter. This check-up includes the width and height of the display; image focus and position; luminance from the SMPTE pattern; and a graph of the DICOM Part 14 grayscale display function index that is generated by the video driver card software.
Since the recordkeeping began, Ly has assembled a compelling body of data on monitor lifespan. Monitors that couldn't be calibrated to perform within manufacturer specifications, or had other gross defects, were replaced at TCH. In 2001, 13 of the 51 monitors in service at the facility were replaced, a 25% replacement rate per year.
Maintenance records show that the replaced monitors had been in service from 9 to 60 months. Two of 10 high-resolution monitors (2K x 2K) replaced had life spans of 11 and 14 months, while the 11 low-resolution monitors (1K x 1K) lasted an average of 35 months.
Ly cautioned her audience that a year’s worth of collection has produced a voluminous data set that is very time-consuming to analyze. Her group is currently working on ways to automate both the collection and analysis of calibration and performance data for the displays.
And even though the work is demanding, Ly will be putting the results to good use. "Our ongoing analysis of the lifetime of the displays will be important when it is time to negotiate purchase contracts, because grayscale CRTS are expensive," she said.
By Jonathan S. BatchelorAuntMinnie.com staff writer
May 3, 2002
Related Reading
LCD just as good as conventional monitors for chest CR, April 30, 2002
Bigger not necessarily better for digital chest x-ray matrix, January 15, 2002
Flat-panel combo provides less noise, lower dose, November 29, 2001
ER study calls soft-copy as reliable as film, October 1, 2001
Copyright © 2002 AuntMinnie.com
