Color Stability of Liquid Crystal Displays Used for Medical Imaging
 
Authors:
Alisa I. Walz-Flannigan, PhD, Mayo Clinic; Heather L. Weber, RT(R)(T)(QM); Scott F. Stekel; Diana M. Lanners; Renee S. Jonsgaard, RT(R); Teresa M. Peterson, RT(R)
 
Background:
Quality control (QC) and acceptance testing methodologies for displays used in medical imaging have largely been based upon grayscale displays, with the heaviest emphasis on grayscale luminance response. While color displays have become common-place in the reading room, QC standards specific to color liquid crystal displays (LCDs) have yet to be established.[1] The lack of published data on the color aging of displays hampers the development of evidence based QC guidelines.[2] Previous studies have quantified the maximum color differences allowed between indistinguishable displays, a quantity that can be used to color match pairs of displays at acceptance testing.[3] However, not enough is known about chromatic changes of aging color LCDs to guide if and how a QC program is needed to maintain color matching and uniformity. It has been shown that uniform yellowing of an LCD over time does not affect the diagnostic utility of a display.[4] However, anecdotal evidence from within our own practice has suggested that other types of non-uniform color aging occur, and the most common complaint received from our radiologists about the color displays has been of a color mismatch. In order to efficiently maintain our color displays through a QC program, we wish to better understand if and how LCDs change color with age.
 
Evaluation:
As part of our QC program for diagnostic displays in our MRI practice, quarterly measurements of color points were taken at five locations on each display of a pair. The average color point of each display was used to calculate the mean radial distance (MRD) between the displays in the 1976 CIE color coordinate system (u’v’). Based on previous study,[1] if the MRD exceeded 0.006, the pair of displays were deemed to be perceptibly different and would be recalibrated. There were a few cases with MRD < 0.006 that showed color mismatch, and those were also recalibrated.

We have accumulated data over the course of almost three years, for approximately 23 pairs of displays. From this data we have sought to understand whether the LCD panels age uniformly in color across a single panel or across the fleet of displays, and how quickly and by how much they change. We want to know if their color points move in a standard direction in CIE 1976 (u’v’) color space. Is it valid to assume that the color shifts for LCDs would be dominated by yellowing of the backlight, or is aging apparent in the LCD panel itself? We hope to use this data to guide the development of colorimetric aspects of our display QC practice. For instance, can one color measurement be used per display instead of 5? Is there sufficient, consistent change over time to warrant such regular testing, or are there just a few aberrant displays which could have been identified with a subjective visual assessment? Do displays demonstrate divergent color changes warranting continued measurement and tracking or do they change uniformly such that displays of the same age should appear similar in color?
 
Discussion:
Since the establishment of color measurements in our QC program, the number of color-related complaints about displays has greatly diminished, but the data extraction, recording, and analysis takes time. Color recalibration often results in reduced luminance output, so it may be questionable whether displays ought to be recalibrated, unless the color difference is readily, visually apparent. However, in the interest of establishing guidelines for standards of practice and uniform appearance of displays throughout the practice, having control limits, based on objective measurement, and a measurement frequency, based on realistic aging is worthwhile.
 
Conclusion:
In time, the color points of a LCD do appear to shift in a yellowing tendency. In general, it appears the amount of change from the original white point coordinates is perceivable after a couple thousand hours of usage. The color points at different locations within the same display tended to change in a similar fashion, such that the difference between their color points in time stays relatively constant. Different displays show the same tendency, such that different displays at the same age tend to show the same color shifts.

If displays are shown to be uniform and matching in color at Acceptance Testing, we do not feel it is worthwhile to continue measuring multiple intradisplay color points to identify non-uniform behavior or to measure color points across paired displays. However, a quick subjective visual analysis of color uniformity and matching during quality control inspection would still be worthwhile to identify aberrant display behavior.

 
References:
[1] Groth DS, Bernatz SN, Fetterly KA, Hangiandreou NJ. “Cathode ray tube quality control and acceptance testing program: initial results for clinical PACS displays.” Radiographics 21. 2001;719-732.

[2] ACR Technical Standard for Electronic Practice of Medical Imaging. November 2007. http://www.acr.org/SecondaryMainMenuCategories/quality_safety/guidelines/med_phys/electronic_practice.aspx

[3] 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. Medical Physics Publishing. Madison, WI. AAPM On-Line Report No. 03, April 2005.

[4] Krupinksi E, Roehrig H, Fan J. “Does the Age of Liquid Crystal Displays Influence Observer Performance?” Acad Radiol. 2007;14:463-467.