Medical-grade PACS displays outshine consumer models

Consumer monitors carried by electronics stores may be functional and economical for radiology, but when put to the test by a year of use, they fall short compared with medical-grade monitors.

That's the conclusion of Dr. Elizabeth Krupinski, a research professor in the departments of radiology and psychology at the University of Arizona. She compared the two types of displays for a study commissioned by medical monitor vendor Barco.

The investigation, to be published in the Journal of Digital Imaging, was to determine whether commercial off-the-shelf (COTS) color displays are adequate for radiologists or whether radiologists should invest in more expensive high-performance medical-grade displays.

The answer, it seems, comes down to user preferences and finances.

Visible changes that reduce the utility of monochrome displays over time have been well documented, according to Krupinski. As monochrome liquid crystal displays age, slight degradations occur in diagnostic performance and visual search efficiency.

"Radiologists tend to take a little longer to render a positive decision and tend to fixate on missed lesions for less time with older displays," she said. "Whether the same types of performance degradations occur with medical-grade versus COTS color LCDs has yet to be explored."

Her goal was to compare decision and visual search performance in users of the two types of displays.

Six radiologists-three board-certified practitioners and three senior residents-were twice presented a set of 50 chest images (posteroanterior view), half with a solitary pulmonary nodule verified by CT and half nodule-free.

In one trial, the images were displayed on a 3-megapixel Barco color medical-grade display monitor. In another trial, the images were displayed on a color 2-MP monitor by Dell.

After viewing each image, the radiologists reported their decisions, using a six-point scale. They detailed whether a nodule was present and whether their confidence in that decision was possible, probable, or definite.

Each trial was timed from the point at which the image appeared on the display to the moment when the radiologist selected the next image button.

Twenty of the images were then used in an eye position study. Readers returned to the testing site after at least three weeks to view the images on both displays (Figure 1).

The differences were small, statistically. In instances in which the readers missed the nodule, however, the radiologists' eyes landed on the target in 1.62 seconds with the medical-grade display, compared with 1.87 seconds for the COTS display (Figure 2).

Ultimately, the medical-grade display performed better, Krupinski said. She attributes that advantage to built-in systems that calibrate for luminance over time, ensuring optimal continuing performance.

"Radiologists found more nodules and made fewer mistakes with medical-grade versus off-the-shelf," she said. "That in itself should say you've got to use the medical."

Her study drew early criticism.

Dr. David Hirschorn, a research fellow in radiology informatics at Massachusetts General Hospital and Harvard Medical School, has reported that consumer-grade monitors may work just as well as medical-grade displays, at a tenth of the cost.

His argument is not simply about the bottom line, however. Krupinski's report states that each display was set to a luminance that corresponds to backlight aging after one year's time."She is asserting that in her experience that value is 342 for the Dell display and 500 for the Barco," Hirschorn said. "In my experience, the Dell display does not typically fall to 342 but rather to 400 after one year, depending on how bright it started out. Furthermore, there is no clear evidence what brightness is optimal. Some prefer 400; some 500."

The Barco display is engineered not to fall in brightness over the lifetime of the display.

"I think it likely that if she were to have compared one pair of Barco monitors to another and set one pair at 342 and the other to 500, she would have gotten the same results," Hirschorn said.

Strictly speaking, the Barco monitor is better. But if the cost is higher than the value, it doesn't make sense to buy one, he said.

"The way I use the Dell display, it starts out at about 450 and falls to 375 over 18 months, at which point I replace it," Hirschorn said. "So I would never recommend performing diagnostic reads off of a 342 monitor, even though that was the highest brightness PACS displays used to be able to achieve in the late 1990s."

Krupinski acknowledged that consumer displays are evolving and manufacturers will likely soon catch up with the level of quality offered by superior medical-grade displays. Until then, radiologists who want to stick with Dell or other off-the-shelf brands can find tools on the market to manage luminescence levels.

"The key thing is to be aware of what the limitations are of a COTS display," Krupinski said. "If you are aware and take measures to monitor differences and take care of them when you can, then there should not be a problem."

-By Yomi Wrong