Radiologists are asked to keep track of an overwhelming amount of data and information, often working with as many as 100,000 images in one day. All that info can be too much to handle, according to a new analysis from the Journal of the American College of Radiology.

“Our human limitations are such that only a subset of those images can get real scrutiny, and only a fraction of the medical record and literature can be weighed,” wrote Murray A. Reicher, MD, Merge Healthcare, and Jeremy M. Wolfe, PhD, Brigham and Women’s Hospital. “The stress of an insurmountable job is contributing to radiology burnout. It is time for technology that can help radiologists thrive in an era of information overload.”

Reicher and Wolfe proposed several different ways the lives of radiologists could be enhanced if workstation technology was more focused on the basic science behind human perception and cognition. Just as researchers should never stop searching for new breakthroughs in medical technology, the authors wrote, they also should never stop seeking improved ways to interact with technology.

These are just some of their suggestions:

1. Allow easy ‘toggling’ between images

Going back and forth from one image to another in the same screen location is the ideal way to find differences between those two images, the authors explained. Perfectly aligning radiology images that are days, months, or even years apart won’t always be possible, but it still gives the radiologist the best possible chance to catch any potential problems.

“When nonexperts compare photographs or radiologists compare mammograms, observers are significantly faster at finding a change when images are toggled than when they are viewed side by side,” Reicher and Wolfe wrote. “This can probably be attributed to a reduction in the number of large saccadic eye movements that are needed in the toggle condition. In side-by-side viewing, any interesting feature in one image must be checked in the other image in a physically different location. In the toggle condition, that interesting feature can be examined without moving the eyes and without the possibility that the large eye movement might not land in exactly the right spot.”

2. Track and analyze eye movements

Eye movements in radiology have fascinated researchers for a long time, and radiologists who read CT scans have been separated into two categories: “drillers,” who keep their eyes in a limited area while looking through the depth of stacks of images, and “scanners,” who scan all over the image.

It is currently not clear if drillers or scanners have the “better” method for reading these images, but if a workstation could be aware of where the radiologist’s eyes are at all times, the authors said it may help improve the quality of care.

“Perhaps a workstation cognizant of eye movements could be programmed to automatically magnify or otherwise transform a region, if that region were persistently fixated (akin to tying computer actions to the position of the mouse, hovering over an advertisement, for example),” Reicher and Wolfe wrote. “A workstation that monitored the oculomotor behavior of a radiologist might be able to learn hanging protocols specifically suited to that radiologist.”

3. Limit interruptions (when possible)

Radiologists must be able to concentrate for an extended period of time, the authors wrote, but it can be difficult to focus when being bombarded with constant distractions. A “collaborative workstation” that can help the radiologists maintain his or her focus could go a long way toward improving performance.

“Interruptions will happen, but the workstation could serve as the personal assistant who protects the radiologist’s focus,” Reicher and Wolfe wrote. “For example, a messaging system built into a PACS might interrupt a radiologist only during the transition between examinations. Even more cleverly, a system might selectively route a technologist’s call to the radiologist who has been reading nonstop for more than an hour, thus intelligently meting out a break to the physician most in need of one. The system might monitor the radiologist’s computer activities in order to detect and warn about signs of fatigue.”

4. Compile information

Workstations could also help radiologists by summarizing key facts and statistics that help with determining a diagnosis.

Reicher and Wolfe compare patient information with that of a baseball player’s—if you read a detailed description of a player’s career, focusing on every important moment and statistic, it might take all night. But as long as you have a basic understanding of the sport, just glancing at the back of a player’s baseball card can provide a lot of the key information you need.