Report from SIIM: Informatics specialist eyes visual problems in radiologists

Radiologists' eyes can suffer physical injury from a reading environment that is not optimized for their benefit. Attention has focused on ergonomic posture and ambient light, but researchers are beginning to understand that many other factors can put a strain on visual acuity.

The reading environment is not well studied or understood, and fatigue and injury are becoming more prevalent. As a result, radiologists need better interventions before irreparable damage occurs, said Elizabeth A. Krupinski, Ph.D., an informatics specialist at the University of Arizona. She spoke at the 2007 Society for Imaging Informatics in Medicine meeting in Providence.

"I still find it amazing that it's not mandatory for radiologists to get their eyes checked annually," she said.

While carpal tunnel syndrome and repetitive motion disorder are still concerns, Krupinski is hearing more complaints about eye strain, dry eyes, glaucoma, headaches, corneal erosion and abrasion, and contact lens problems.

As part of a study, Krupinski polled participants at a meeting as to whether they needed some form of corrective eyewear. Her results revealed that about 97% of them did.

"That's far above the national average," she said.

She is presently collecting data that show a radiologist's resting focus moves back the longer he or she stares at a monitor. In other words, blurry items on the screen would become clear about a foot or so behind the screen.

Eye fatigue is one factor that affects resting focus. Reasons for eye fatigue include viewing distance, ambient lighting, resolution, viewing angle, length of viewing, and mental workload. It becomes especially difficult to focus on specific subtle changes that may be important for diagnostic accuracy, Krupinski said.

The longer one stares at a monitor, the slower the blinking rate. This can happen within 20 minutes. Without blinking, the eyes are not being moisturized and are subject to easier irritation. Dry eyes can be exacerbated by age, contact lenses, air conditioning or heating, geographic location, dust, and allergies.

A typical 65-year-old has one-third less light reaching the retina compared with a typical 20-year-old. Color perception also changes with age, resulting in a weaker distinction among violet, blue, and green.

"We have to optimize the reading environment to account for radiologists getting older," Krupinski said.

She highlighted a study from Duke University researchers presented at the 2005 RSNA meeting that revealed radiologists must contend with many items that reflect light onto their screen. These include bright wallpaper, white paper on the desk, radiographic film, wooden table, blackboard, carpet, and the white coat on the radiologist's back.

"Some of these may contribute more distractive reflection than others, but it's important for radiologists to know that they all have an effect on their reading environment and diagnostic accuracy," Krupinski said.

A study presented by Patrick Brennan, Ph.D., at the 2006 International Society for Optical Engineering (SPIE) meeting found that moderate lighting, 40 and 25 lux, led to the best performance of radiologists. High (480 lux) and low (under 25 lux) ambient lighting resulted in more false-positive and false-negative findings.

Studies have also found that display properties such as calibration, MTF and OTF compensation, window/level format, CRT versus LCD, flat versus curved surface, viewing angle, and color versus monochrome all affect the diagnostic accuracy of the radiologist.

When the luminance of the display is controlled for optimal distraction, the pupils are not contracting or expanding as much and, therefore, fatigue less easily, Krupinski said.

A study by Krupinski found that color images, when examined very closely, have more noise than black-and-white images. Does this have an effect on radiologist performance? Apparently not.

Krupinski used eye-tracking software to test the efficiency of radiologists as they interpreted chest images. She found no statistically significant difference in radiologist performance when using high-performance monochrome monitors, low-performance monochrome, or color. There were also no statistically significant differences in viewing time on each monitor

"I'm convinced we can use color monitors, at least for chest nodules, probably for fractures, and I'm willing to bet for mammography as well," she said.

For more online information, visit Diagnostic Imaging's SIIM 2007 Webcast.