With the medical-imaging industry implementing filmless imaging at an ever-increasing rate—and given that one of the principal components of the digital-imaging chain is the final display device—review of quality issues in liquid crystal display (LCD) flat-panel devices should be of great interest. This month’s quality-comparison table is dedicated to reviewing some of the principal factors that contribute to image display quality on LCD flat-panel monitors.

John Weiser, PhD, and John Romlein, MS, are partners and cofounders of Qualiteering Labs LLP (QLABS LLP), Thurmont, Md. QLABS LLP is dedicated to the betterment of medical-imaging operations through quality-process reengineering and education. E-mail John Weiser at ; e-mail John Romlein at .

We have conducted a survey of representative LCD monitor vendors and created a table of the responses. The chart will verify what we all know to be true: There can be significant disparity in the method and detail of specifying monitor performance factors. This situation can make it difficult for the consumer to properly evaluate display characteristics in an apples-to-apples manner.

Without being too technical or specific, we developed a set of topic areas and related questions that we posed to a representative set of vendors. The topics and questions we posed follow, as well as a brief discussion of our findings, which are detailed in the table that provides the results of this vendor survey. The data is derived from both direct written responses to questions as well as a review of specification sheets for the proposed displays. The table is populated with a variety of monitor types, ranging from 1-megapixel (MP) color monitors to 5-MP grayscale monitors.

1) Brightness

Often, brightness is quoted as the maximum brightness capability of the display; however, this setting is not the setting at which the display would normally be operated. Usually, the manufacturer recommends a lower setting that should be used to optimize the useful life of the display.

Questions posed to vendors:

  • What is the maximum brightness of your display?
  • What is the recommended operating brightness of your display?

Discussion:

Maximum brightness is specified in all cases, but some vendors do not provide a recommended operating brightness, which typically ranges from 25% to 50% lower than the maximum brightness. Recommended operating brightness is used to preserve the monitor’s backlight over time and to provide the backlight stabilization circuitry with additional backlight output capacity to maintain the upper operating point upon which Digital Imaging and Communications in Medicine (DICOM) calibration is based. Section 4.3.4 of the American Association of Physicists in Medicine (AAPM) Task Group 18 (TG18) report calls for an operating level greater than 171 cd/m2. Table 7 of that same report provides minimum acceptance criteria for primary and secondary displays.1

2) Contrast Ratio

Contrast ratio can be measured in different ways, producing results that vary widely. It is important for consumers to know the conditions under which contrast ratio is determined.

Questions posed to vendors:

  • What is your method for determining the contrast ratio reported in your specifications? For example, do you use a method from the Video Electronics Standards Association (VESA) Flat-Panel Display Measurements Standard (FPDM)?2
  • What display menu settings, such as backlight, contrast, and color temperature, are used for determining the contrast ratio?
  • What is the white level luminance (cd/m2) at which the contrast ratio is determined?
  • What is the contrast ratio when the display is set to the recommended operating brightness?

Discussion:

Contrast ratio is most often measured using maximum brightness instead of recommended operating brightness. The resulting specification is not what will be achieved in the clinical setting under routine operating conditions, where luminance is set to the recommended operating level. The effect of minimum brightness and ambient lighting on the effective contrast is left out of this discussion, but they also play very significant roles. Only one vendor reported contrast ratio as a function of recommended operating brightness. Three vendors referenced the VESA FPDM method of contrast ratio measurement; only one provided the requested setup conditions used when determining contrast ratio.

3) Uniformity

It is important for medical displays to minimize the difference in image appearance between the center and the corners of the displays. Typically, uniformity is not quoted in monitor specifications.

Questions posed to vendors:

  • What is the minimum acceptable uniformity of your display?
  • What method do you use to determine uniformity?
  • What are the display menu settings at which uniformity is determined?
  • What is/are the luminance setting(s) at which uniformity is determined?

Discussion:

Some vendors provided a uniformity specification while others provided a nonuniformity specification. Displays were in the range of between 70% and 90% uniformity. The AAPM TG18 minimum acceptable uniformity for diagnostic (primary) or clinical review (secondary) displays is greater than or equal to 70%. A variety of descriptions were provided for the setup conditions and method for this test. Note that while, in most cases, maximum brightness was used to test for contrast ratio, uniformity was tested using the recommended operating brightness, which lessens the tendency for nonuniformities in brightness across the pixel matrix.

4) Pixel Defects

Depending on the location, clustering, and severity of the defect, pixel defects can cause observable image artifacts.

Question posed to vendors:

  • What criteria do you use during manufacture to determine if a display should be rejected due to pixel defects?

Discussion:

Only one vendor actually provided the defective pixel criteria used. The others that answered deferred to their suppliers’ specifications—and, in one case, deferred to the customer’s inspection specification—but they did not provide the actual criteria. The AAPM TG18 report discusses this topic in section 4.9.2.2, and the report refers to both the International Standards Organization (ISO) 13406-2 and VESA FPDM for quantitative methods of assessment.

5) DICOM Grayscale Conformance

There are many different ways to provide on-board DICOM Grayscale Standard Display Function (GSDF) conformance on a display. Examples include preset lookup table that applies to one display setting only, without internal brightness monitoring; preset lookup table that applies to one display setting only with internal brightness monitoring; preset lookup table that applies to multiple display settings with internal brightness monitoring (ie, display can be operated at various brightness settings while maintaining GSDF conformance); and no preset lookup table, but can be calibrated with external luminance measurement device and calibration software. In addition, the preset lookup tables can be individually determined for each display during manufacture, or determined for a general category of displays in a “one size fits all” approach.

Question posed to vendors:

  • Considering the types of conformance methodologies listed above, or possible others, what method does your display use to provide DICOM GSDF conformance?

Discussion:

All vendors provided some method of creating a DICOM GSDF conformance curve. Other features included internal luminance sensors, front luminance sensors, quality control (QC) management software, and front-panel light-emitting diode (LED) status indicators. DICOM GSDF lookup tables either were preset or required an external photometer and calibration software. References for this function are available in DICOM 3.143 as well as the AAPM TG18 report’s sections 1.2.2, 3.4.6, and 4.3.

6) Chromaticity

Consistent color temperature between displays is especially important for multiple display workstations.

Questions posed to vendors:

  • What criteria do you use to ensure color matching of your displays?
  • Can your displays be ordered in matched sets if they are intended to be used on a multiple-display workstation?
  • If a replacement display is needed for a multiple-display workstation, can the replacement be ordered to match the existing displays?

Discussion:

Color matching is an important factor for both color and grayscale monitors. Most vendors referenced the Commission Internationale d’Eclairage (CIE) 1931 Standard,4 and they provided color coordinates and, in some cases, tolerances; however, the vendors provided various levels of detail as to the measurement method against the standard as it applies to their product. Some vendors offer color-matched display sets and replacements; others do not. See section 4.8.4 of the AAPM TG18 report for more information on color-matched display sets and replacements.

7) Response Time

The ability to display motion and respond quickly to rapid image change has become increasingly important in medical imaging. The pixel response time—how fast the pixel can turn on and turn off—is a measurement of the display’s ability to provide high-quality images of motion studies.

Questions posed to vendors:

  • What is the response time of your display?
  • What method do you use to measure response time?

Discussion:

All but one vendor supplied response time performance data. The values ranged from 100 ms to 8 ms. The method of measurement was not consistently described and, in some cases, referred to the panel manufacturer without further detail. The use of displays for quickly driving through stacks of CT or MRI images or for viewing color video images, such as visible light video, could be negatively affected by slow response times.

8) Viewing Angle

The acceptable viewing angle of a display is important in medical applications where more than one person could be viewing a displayed image or where there are multiple displays on a single workstation, which introduces some off-angle viewing. Vertical viewing angle is critical in situations where people of varying heights are using the same workstation, such as a radiologic technologist QC station.

Question posed to vendors:

  • What criteria do you use to determine the published vertical and horizontal viewing angles for your display?

Discussion:

All but two vendors reported that the viewing angle was consistently the point that the contrast ratio reaches 10:1; one vendor provided additional detail. There was no significant difference between specified viewing angles. The range was from 85° to 89° in both vertical and horizontal directions. Reference this topic in section 4.4.4.1.2 of the AAPM TG18 report.

9) Continuing Service and Warranty

Question posed to vendors:

  • Does your warranty and maintenance plan address the previously discussed eight issues as well as take into account changes in these characteristics for the life of the monitor?

Discussion:

The degree of information as to the coverage of these performance factors during warranty maintenance periods varied widely. One vendor provided a warranty of most of the performance factors; however, the majority of vendors specified only backlight longevity. One vendor specified color-matched replacements and three levels of extended service plans for up to 7 years. None of the vendors explicitly expressed warranty support of all performance factors for an extended period.

Conclusion

Presently, the state of the industry for testing and specifying monitor performance is unsettled and somewhat in conflict with itself. Good standards and testing protocols are available from a variety of US-based and international bodies—including the American College of Radiology (ACR), AAPM, Deutsches Institut fur Normung (DIN), VESA, ISO, DICOM, and the National Electrical Manufacturers Association (NEMA)—but their application is not consistent across vendors or called for by customers. Furthermore, users and vendors seem to be on different wavelengths regarding the actual performance requirements. Users are interested in the performance of displays in the clinical setting; however, vendors and manufacturers, in most cases, are specifying their devices under laboratory conditions and in a manner that optimizes one performance factor at a time, while, in some cases, degrading other performance factors. The specification of contrast ratio at maximum brightness versus contrast ratio at recommended operating level is a good example of this.

In most cases, there was little consistency and detail in the warranty and maintenance programs support of prolonged operation of the display devices at the specifications listed by the vendors. If reliability and total cost of ownership are to be estimated, there needs to be a basis for estimating the useful life expectancy, maintenance cost, and replacement cost. Without vendors’ guarantee that their product will support the desired operating performance levels for a specified period of time, such a cost can not be estimated.

It falls to the customer to understand the implications and tradeoffs of the current lack of uniform specification and testing methodology. Future development in the standardization of testing and specification methods—coupled with increased education to manufacturers, vendors, and customers alike—will drive the industry to a point where an apples-to-apples comparison can be made.

References

  1. Samei E, Badano A, Chakraborty D, et al. Assessment of display performance for medical imaging systems, report of the American Association of Physicists in Medicine (AAPM) Task Group 18. April 2005. Available at: http://www.aapm.org/pubs/reports/OR_03.pdf. Accessed May 31, 2006.
  2. Video Electronics Standards Association FFDM Task Group. Flat panel display measurements standard, version 2.0. May 19, 2005. Available at: http://www.vesa.org/public/Fpdm2/FPDMUPDT.pdf. Accessed May 19, 2006.
  3. National Electrical Manufacturers Association. PS 3.14-2006 Digital imaging and communications in medicine (DICOM) part 14: grayscale standard display function. Available at: http://medical.nema.org/dicom/2006/06_14pu.pdf. Accessed May 19, 2006.
  4. Commission Internationale d’Eclairage. 1931 CIE chromaticity diagram. Available at: http://www.biyee.net/v/cie_diagrams/index.htm. Accessed May 19, 2006.