Original article
Technology and Perception in the 21st-Century Reading Room

https://doi.org/10.1016/j.jacr.2006.02.022Get rights and content

Radiology reading rooms have changed dramatically over the past 15 years, moving from analog-light-box-based environments to digital-display-based environments. Most of the focus in the early stages of this transition was on the technology, but it soon became obvious that it was not possible or even prudent to consider the technology without considering radiologists. The information being presented to radiologists in digital reading rooms is in many ways very different from that presented on traditional film. On one hand, the digital workstation display medium itself is very different from traditional film images hung on light boxes. On the other hand, without large-area light boxes, images such as those from computed tomography (CT) can no longer be displayed all at once in a series of film sheets. The digital world also introduces the possibility of manipulating image data in ways that were never possible with analog film. Not only can radiologists manipulate image data with various image-processing tools, but also, computers can analyze images and provide even more information to incorporate into the interpretation process. As a consequence of these differences, it has been necessary to focus attention on radiologists to discover ways to optimize the digital reading environment with respect to the human visual system and the way the eye-brain system processes information. This article reviews some of the important perceptual issues that have arisen in the digital reading rooms of the 21st century.

Section snippets

Perception and the transition from film to digital

Much of the early research in radiologic image perception centered on ways to better understand how and why radiologists made errors and why there is so much interobserver and intraobserver variation [1, 2, 3, 4]. Examinations of visual search patterns using eye-position recording techniques revealed much about the source of reader errors as well as the role of experience in image perception [5, 6, 7, 8, 9]. As the transition from film to the digital display of radiographic images began to take

Perceptual optimization of soft-copy image display

The question that lingered after the early equivalence studies was how to optimize soft-copy displays to yield the best possible display of images without introducing too much burden on radiologists in the form of excessive image processing to render images in an acceptable way. The majority of the early display evaluations were done with cathode ray tube (CRT) displays, and much has been written on how to physically characterize and calibrate them [14, 15]. Today, liquid crystal displays

Display luminance

For example, one study compared diagnostic accuracy as a function of display luminance and also recorded eye position to determine if luminance affected the visual search patterns of radiologists [21]. The study used a high-performance, monochrome CRT (140 ftL) with another high-performance display that had a 60% decrease in luminance (80 ftL). Performance in the higher luminance condition was better than in the lower luminance condition, although statistical significance was not achieved. The

Display calibration

Another crucial display parameter that has been found to influence both perception and performance is how a soft-copy display is calibrated. As soon as soft-copy displays began to be used, it was noticed that the same image displayed on one display often looked very different when displayed on another device. Even when an image was displayed on the same device, it could look different at different times. To address this issue, the Digital Imaging and Communications in Medicine (DICOM) Grayscale

CRTs vs LCDs

Every type of display has some advantages and some disadvantages. The CRT was the major electronic display for years before LCDs matured enough to be used for the primary interpretation of radiographic images. Cathode ray tubes are certainly well-suited for radiologic interpretation in terms of spatial resolution and luminance, but they are also bulky and have curved glass surfaces that produce reflections (even with antiglare coatings, there are some reflections) [23] and increase veiling

Compensating for display deficiencies

Luminance, calibration technique, and monitor technology are just a few of the physical characteristics of soft-copy displays that have been shown to influence diagnostic accuracy and the visual search performance of radiologists. As technology improves and as new display technologies are developed, further characterization, optimization, and evaluation of displays will have to take place to ensure that diagnostic accuracy is not affected negatively. In the meantime, ways to improve current

Image processing and interobserver variability

Many studies carried out in the laboratory setting generally show improved performance (of varying degrees) when a given image processing tool is implemented. Some of the more successful and often-used tools are those that use edge enhancement. For example, Goo et al [34] had radiologists review a series of neonatal chest images before and after edge enhancement was applied and found that the edge enhancement improved significantly the visualization of pneumothoraces, vascular catheters, the

21st-century perception opportunities

It is clear that digital reading rooms have already changed dramatically the image interpretation process and presented the opportunity for a wide variety of new tools for radiologists to use for interacting with and manipulating displayed image data. It has already been noted that the display medium itself is also likely to continue to change as new technologies emerge, and the tools that radiologists have will also continue to develop and change. One thing driving the need to develop more

Summary

Radiologists practicing 10 years from now will in all likelihood be working in a very different environment than today’s radiologists. Analog film will be a rare commodity compared with today, used for the most part for old comparison studies. There are likely to be new types of digital images, acquired using different technologies than exist today, and processed in ways that have yet to be conceived. Molecular imaging will surely have a huge impact on the types of studies being done and the

References (59)

  • A.R. Margulis

    Is academic radiology still attractive?

    Acad Radiol

    (2003)
  • M. Taljanovic et al.

    Academic radiologythe reasons to stay or leave

    Acad Radiol

    (2003)
  • W.B. Demartini et al.

    Computer-aided detection applied to breast MRIassessment of CAD-generated enhancement and tumor sizes in breast cancers before and after neoadjuvant chemotherapy

    Acad Radiol

    (2005)
  • H.O. Shin et al.

    Use of 2D histograms for volume rendering of multidetector CT datadevelopment of a graphical user interface

    Acad Radiol

    (2004)
  • C.C. Birkelo et al.

    Tuberculosis case findingcomparison of effectiveness of various roentgenographic and photofluorographic methods

    JAMA

    (1947)
  • L.H. Garland

    On the scientific evaluation of diagnostic procedures

    Radiology

    (1949)
  • L.H. Garland

    On the reliability of roentgen survey procedures

    Am J Roentgenol

    (1950)
  • W.J. Tuddenham et al.

    Visual search patterns in roentgen diagnosis

    Radiology

    (1961)
  • E. Llewellyn-Thomas et al.

    Visual search patterns of radiologists in training

    Radiology

    (1963)
  • H.L. Kundel et al.

    Visual scanning, pattern recognition and decision-making in pulmonary tumor detection

    Invest Radiol

    (1978)
  • C.F. Nodine et al.

    Time course of perception and decision making during mammographic interpretation

    Am J Roentgenol

    (2002)
  • B.S. Slasky et al.

    Receiver operating characteristic analysis of chest image interpretation with conventional, laser-printed, and high-resolution workstation images

    Radiology

    (1990)
  • L.F. Fajardo et al.

    Detection of pneumothoraxcomparison of digital and conventional chest imaging

    Am J Roentgenol

    (1989)
  • M.S. Frank et al.

    High-resolution computer display of portable, digital, chest radiographs of adultssuitability for primary interpretation

    Am J Roentgenol

    (1993)
  • E.A. Krupinski et al.

    Receiver operating characteristic evaluation of computer display of adult portable chest radiographs

    Invest Radiol

    (1994)
  • H. Roehrig

    The monochrome cathode ray tube display and its performance

  • E. Samei et al.

    Assessment of display performance for medical imaging systemsexecutive summary of AAPM TG18 report

    Med Phys

    (2005)
  • H. Jung et al.

    Assessment of flat panel LCD primary class display performance based on AAPM TG 18 acceptance protocol

    Med Phys

    (2004)
  • J. Forrester et al.

    The eyebasic sciences in practice

    (1996)
  • Cited by (0)

    View full text