PC monitors allow diagnostic reading under optimal conditions and settings

Study underscores potential for PACS cost savings

The growing reliance on PCs in the medical imaging industry has elevated the status of these computers. But while their computing horsepower may be enough to run most postprocessing algorithms, their displays are not up to diagnostic standards, say PACS vendors, many of whom insist on marking images displayed on PC monitors "for reference only." Such branding may not be warranted, according to a study performed by researchers in New Zealand.

Although PC monitors do not perform at the same level as those found on diagnostic-quality workstations, they can generate diagnostic-quality images under the right conditions, according to Dr. Anthony Doyle, a clinical associate professor of radiology at the University of Auckland. His study (Acad Radiol 2002;9:646-653) found that PC-based systems provide results similar to those obtained with a workstation at considerably less cost, at least for evaluating computed radiographs of the hands in early rheumatoid arthritis.

The finding could have broad implications for the PACS industry, whose customers must balance the additional cost of diagnostic-

quality workstations against increasingly less expensive PCs. In many cases, customers mix expensive workstations and PCs and, eventually, begin using PCs for diagnostic purposes, according to Doyle.

"We had already started using images displayed on PCs for clinicians and were aware that in many cases the clinician would in fact effectively use these for diagnostic purposes before receiving the radiology report," he said. "We wanted to investigate the validity and safety of that approach."

Doyle and his colleagues found that, when appropriately used, PCs are an adequate diagnostic tool for viewing digital radiographic images, at least in skeletal applications. Some changes for optimizing the image involved the monitor itself, such as ensuring that the proper window and level settings were in place. Others involved altering the conditions under which PCs were used: shutting blinds to eliminate glare on the face of the monitor, for example.

Remedial efforts such as these were used to create optimal conditions for reading images on PC monitors in the New Zealand study. The results were then compared with those obtained with diagnostic-quality workstations.

"We found no significant differences in performance between the PC and workstation for interpretation of hand images in rheumatoid arthritis," Doyle said.

Once a PACS is introduced into an enterprise, many clinicians believe they need diagnostic-quality workstations, which is not always economically viable. But digital images viewed on a PC with the display optimized using available image tools should be adequate, according to Doyle.

"We encourage the validation of any equivocal findings using a (diagnostic-quality) workstation, but since doing the study, we have had very few requests for that and no instances of patient care being adversely affected by clinical use of PCs for image viewing," he said.