Monitor's megapixels do not affect image interpretation

June 3, 2005

Little or no difference exists between 1-megapixel and 5-megapixel medical-grade LCD monitors for reading cervical spine images for fractures, according to a study presented at the SCAR meeting Thursday. And consumer-grade LCDs may be good enough for reading CT images, a second study concluded.

Little or no difference exists between 1-megapixel and 5-megapixel medical-grade LCD monitors for reading cervical spine images for fractures, according to a study presented at the SCAR meeting Thursday. And consumer-grade LCDs may be good enough for reading CT images, a second study concluded.

A study led by Dr. David Hirschorn, a researcher at Massachusetts General Hospital, concluded that consumer-grade monitors can be just as effective as medical-grade monitors for interpretation of CTs.

The research, from the University of Maryland School of Medicine and Massachusetts General Hospital, suggests that many monitor purchasers may be overbuying when they insist on monitors with 3 MP or more for anything but the most demanding applications.

The studies also indicate that current American College of Radiology standards, which set a minimum monitor brightness standard of 50 foot lamberts and one-to-one scanner-to-monitor mapping of data, may need to be reconsidered, said Dr. Eliot Siegel, chief of radiology and nuclear medicine at the VA Maryland Health Care System and coauthor of the University of Maryland study.

That study followed earlier research that found no difference between 5-MP CRT monitors, the current gold standard, and 3-MP LCD monitors for the interpretation of cervical radiographs. Given that finding, the researchers sought to determine the minimum number of megapixels in LCD monitors that would be adequate for interpretation of a cervical radiograph, said Dr. Nabile M. Safdar, a fellow at the University of Maryland.

The study included 200 C-spine radiographs read by seven radiologists on 5-, 3-, 2-, and 1-MP medical-grade LCD monitors. Sensitivity and specificity were 81 and 77, respectively, on the 5-MP monitor; 82 and 73 on the 3-MP; 87 and 76 on the 2-MP; and 80 and 75 on the 1-MP.

The likely key to the good results from the monitors with smaller matrix sizes was high luminance, Siegel said. Each was set to approximately 160 foot lamberts of luminance, more than three times the minimum 50 foot lamberts specified by the ACR.

"For C-spine, I would be comfortable reading on a 1.3-MP medical-grade monitor," Siegel said.

Each monitor's identifying details were concealed during the study, and the radiologists doing the interpretations could not tell which was which, he said.

The mathematical changes to the images that were required to fit them into different megapixel matrices had no impact. CR images are typically acquired at 6 MP, but adjusting them to fit the different monitors seemed to make no difference, Siegel said.

Boosting the luminance requirement would probably eliminate the use of most or all CRT monitors and all but a handful of consumer-grade monitors for visually demanding imaging studies such as C-spine radiographs, he said.

But consumer-grade LCD monitors should be fine for CT images, according to Dr. David Hirschorn, lead researcher in the Massachusetts General Hospital study. His group compared a 21-inch 3-MP medical-grade gray-scale LCD monitor with a 19-inch 1.3-MP consumer-grade color LCD monitor to interpret 100 randomly selected CT scans.

Two attending radiologists searched for lesions and graded them for conspicuity. In all measures - number of lesions found, number of normals, and average conspicuity - there was no difference between the two monitors, Hirschorn said.

The next step will be to compare the two monitors in the interpretation of radiographs. Hirschorn said he expects that they will not be equivalent.