Wednesday, November 27, 2013




This document discusses methods for consistent color and grayscale presentation for digital color displays to be used in medical interpretation. There are advantages in using color displays over grayscale displays. Users can apply different modalities to the same display. Furthermore it is easier to use the application software on a color display as the navigation on a color display is more clear thus increasing users productivity. Above all, color displays are required in pathology, telemedicine and ophthalmology, because of the inherent color in these medical imaging methods.

Due to technical developments in display manufacture, more displays are and will be capable to display colors. LCD manufacturing costs were reduced drastically due to the high output volume in the TV market. Cold cathode fluorescent lamps (CCFLs) are being replaced by white LED, RGB LED backlight and OLED (Organic Light Emitting Diodes). One of the first real 10-bit format or panels comes from the consumer market in the formats of 16:9 and 16:10 with internal LUT's of up t0 14 LUT. These technologies provide display manufacturers with the capability to produce displays with bigger gamut, higher luminance, lower heat emission and lower price.

Why are there now more color applications

Large numbers of image rendering devices will allow generation of even larger numbers of clinical images.

Some of those images are displayed for making a diagnosis, reviewing live in conferences or treatment (endoscopy). Some images are true RGB color images and some are pseudo-color images (with COLOR PALETTE Photometric interpretation).

Why are color and luminance consistency important

Color and grayscale consistency makes sure that the same color is produced and reproduced across different output and input devices. These devices can be displays, printers, projectors, cameras or microscopes. This high accuracy in color and luminance reproduction is required due to the high sensitivity of the human observer with respect to luminance differences as well as color differences. Consequently images are required with equidistant color and gray levels, consistent and correct color-reproduction over the entire work-flow process in order to enable the observer to distinguish all color and gray shades that exist in the color image file. 

The need for Color Management in medical imaging

Managing color is a complex procedure and was addressed already in the pre-press industry. 

The International Color Consortium has developed a system to manage colors using a so called ICC profile and a CMM (Color Management Module). The ICC profile does characterise the imaging device and the CMM is a software algorithm that adjusts the numerical values, that get sent to or are received from different devices so that the perceived color they produce remains consistent. 

Figure 1: Color management system8

Figure 2: QUBYX PerfectLum Display calibration and verification software


QUBYX is a technology contributor to the Medical Imaging and Color Management Industries with various professional Calibration and Visualization solutions since 1996. QUBYX develops software solutions that are used in and for computer displays. Our solutions include OSD replacement software, Display Management systems, Color Calibrations and Verification software for the Medical imaging use. QUBYX has an extended knowledge in digital color technologies. Our knowledge extends to handling, calculating, mapping from different color spaces and creating ICC profiles. For More information, visit

QUBYX is working on a system to calibrate,verify and evaluate the accuracy of color and luminance calibration of medical displays and prints.

Figure 3: Calibration and verification

Verification includes deviations towards the target luminance response function and the achieved color reproduction.

Figure 4: Verification of JND (Just Noticeable Differences)

Figure 5: Verification of luminance response as described in AAPM TG182

Figure 6: Verification of Delta E towards the reference color-temperature over DDL (Digital Driving Level)

Figure 7: Color Image Reproduction Process for ICC Managed Color Medical Imaging


Part 14 of the DICOM standard3 describes a function on how an output device like a printer or display shall reproduce luminance output over driving level.

Figure 8: The Grayscale Standard Display Function presented as logarithm-of-Luminance versus JND-Index

Limitations of existing NEMA DICOM Part 14: 
  • applies only to grayscale images 
  • only grayscale consistency over devices 
  • no pseudo colors 

The DICOM Supplement 100 "Color Softcopy Presentation State"5 fills this void and guesses how color should be managed and gives a guideline on how to implement a color-reproduction pipeline and device-independent color spaces.

Basic recommandations of the supplement 100 are:

  • Use existing industry standard of ICC profiles 
  • Use the device-independent Profile Connection Space (PCS), CIEXYZ or CIELAB, as defined by the ICC 
  • The device depended RGB color space shall be sRGB 
  • The rendering intent is fixed to "perceptual" 
  • Space LUTs should be represented as 16 bit values, using LUT 16 type Tag, for greater precision. 
  • The chromaticAdaptationTag should be not set if the actual illumination source is D50. 
  • Color image files should have embedded ICC profiles 


The CIEL L* function is based on the CIE Lab colorspace11 and is reproducing the sensibility of the human observer9 on luminance and color shades.

Figure 9: The CIE L* function

ICC color profile standard

The International Color Consortium promotes a standard for color Specification, an ICC profile, that can be associated with images or devices3 . The profile permits interpretation of numeric values in a multi-channel image as true colors. An ICC profile is a set of data that characterizes a color input or output device, or a color space, according to standards promulgated by the International Color Consortium (ICC). Profiles describe the color attributes of a particular device or viewing requirement by defining a mapping between the device source or target color space and a profile connection space (PCS). This PCS is either CIELAB (L*a*b*) or CIEXYZ. Mappings may be specified using tables, to which interpolation is applied, or through a series of parameters for transformations.

To profile the monitor, again a color sensor is attached flat to the display’s surface. Profiling software sends color and grayscale shades to the display and compares the values actually sent against readings from the calibration device, creating a matrix for the display.

Depending on the calibration software, different versions of ICC profiles are created like version2 or version4 and with different tags either CIELAB (L*a*b*) or CIEXYZ. Mappings may be specified using tables, to which interpolation is applied, or through a series of parameters for transformations.

Every device that captures or reproduce color can have its own profile.

Depending on the profiling application and the device, different TAG´s are embedded into an ICC profile.
AToB0Tag, AToB1Tag, AToB2Tag, BToA0Tag, BToA1Tag, BToA2Tag

Many named color profiles exist: sRGB and Adobe RGB are in common use as the RGB workspace.

Figure 10: CIE 1931 Chromaticity diagram6. 
The range of colors perceptible to the human visual system is represented by the color gamut shown here in CIE 1931 x,y coordinates.

Methods of Display and Printer calibration and profiling


The aim of color calibration is to measure and adjust the color response of a device (input or output) to establish a known relationship to a standard Luminance response like DICOM GSDF or CIE L* and to a standard white point like D65.

Color calibration is a requirement for all devices taking an active part of a color/luminance managed workflow. A measurement device is needed to measured colors and luminance of the Displays and Printers. The measured colors are described in standard color spaces like CIE Lab or CIE XYZ.

Figure 11: CIE xy color space with sRGB profile Gamut view and with colortemperature curve of the blackbody

Display calibration

For calibrating the monitor a color sensor is attached flat to the display's surface, shielded from all ambient light. The calibration software sends a series of color and grayscale shades to the display and compares the values that were actually sent against the readings from the calibration device. This establishes the current offsets in color display. Depending on the calibration software and type of monitor used, the software will load this correction table into the Graphics Board or into the Display.

Figure 12: Luminance response before calibration (Red) CIE L* Target (Green) and result after calibration (Blue)

Figure 13: Gamma correction that was calculated and applied to reach the target color and luminance response

Display profiling

For profiling of the monitor again a color sensor is attached flat to the display's surface. The profiling software sends a series of color and grayscale shades to the display and compares the values that were actually sent against the readings from the calibration device. This creates a matrix for the display. Depending on the calibration software different versions of ICC profiles are created like version2 or version4 and with different Tags.

Figure 14: ICC Profile map

Figure 15: The Gamut of an Display ICC profile in 3D view

Figure 16: Once the Display ICC profile is created it has to be selected as the default ICC profile in the Operating system.

Printer calibration and profiling

For calibrating the printer a spectrophotometer or densitometer is used to measure the ink on the paper. The calibration software create a document with a series of color and grayscale shades and prints it. The print is read with the spectrophotometer or densitometer. The software compares the values that were actually sent against the readings from the measurement device. This establishes the current offsets for the printer. Depending on the calibration software and type of printer used, the software will load a correction table into the print server (RIP) or creates an ICC Profile only. The calibration needs to be redone once paper or ink is changed inside the printer.

Figure 17: Color or grayscale printers can be calibrate to DICOM GSDF

Figure 18: A spectrophotometer or densitometer is used to calibrate and profile a printer

Color management workflow for different modalities

Color Images

Grayscale Images

Pseudocolor Images

Figure 19 (all 4 images together)

The correction LUT (Look Up Table) of the calibration is applied in the display, graphicboard, camera, printer, microscope. The LUT is adding an offset to the color input values and create a new output. The ICC profiles needs to be selected in the system and/or inside the images application or printing and camera driver.

Rendering intent

When the gamut of source color space exceeds that of the destination, saturated colors are liable to become clipped (inaccurately represented), or more formally burned. The color management module can deal with this problem in several ways. The ICC specification includes four different rendering intents: absolute colorimetric, relative colorimetric, perceptual, and saturation.

Absolute colorimetric

If the output device has a much larger gamut than the source profile, i.e., all the colors in the source can be represented in the output, using the absolute colorimetry rendering intent would "ideally" (ignoring noise, precision, etc.) give an exact output of the specified CIELAB values. Perceptually, the colors may appear incorrect, but instrument measurements of the resulting output would match the source. Colors outside of the proof print system's possible color are mapped to the boundary of the color gamut. Absolute colorimetry is useful to get an exact specified color for example like in pathology or pseudo colors.

Figure 20

Relative colorimetric

The goal in relative colorimetry is to be truthful to the specified color, with only a correction for the visual adaptation due to the media (paper color in the case of a printer) or white point in the case of a display. Relative colorimetry is useful in proofing applications, since you are using it to get an idea of how an image on one device will appear on a different device.

Perceptual and Saturation

Perceptual rendering is recommended for color separation. In practice most used is relative or perceptual intent, as for natural images, absolute causes color cast, while saturation produces unnatural colors. Relative intent handles out-of-gamut colors by clipping (burning) these colors to the edge of the gamut, leaving in-gamut colors unchanged, while perceptual intent smoothly moves out-of-gamut colors into gamut, preserving gradations, but distorts in-gamut colors in the process. If an entire image is in-gamut, relative is perfect, but when there are out of gamut colors, which is more preferable depends on a case-by-case basis. Saturation intent is most useful in charts and diagrams, where there is a discrete palette of colors which one wishes to have saturated, but where the specific hue is less important.

Figure 21

Figure 22

Display White and Black level and ambient light

Display white and black level must be constant over time as well as the ambient light. The ambient light should be low and no light reflection should be visible on the display panel.

Still unexplored and missing items.

  • PACS applications and DICOM viewers miss ICC profile support. Without ICC profile from the imaging application the color correction will not be applied. Software developers need to integrate the ICC support. 
  • The DICOM Supplement 100 "Color Softcopy Presentation State” does not specify the luminance response a color display shall be calibrated. Should the luminance response be GSDF or CIE L* or gamma? Should color be corrected on a color display where grayscale images are displayed? Should an ICC profile be created for color displays where only grayscale images are displayed? 
  • The DICOM Supplement 100 "Color Softcopy Presentation State” fixed the rendering intend to “Perceptual”. Perceptual will make sure that the perception of the entire images stays intact. But for applications where the exact color is more important than the global perception like in applications like Endoscopy, Microscopy or in case of pseudo-colors the “absolute” rendering intent would be here more appropriate. 
  • The DICOM Supplement 100 "Color Softcopy Presentation State” does not specify a white point or any standard illuminant10 to calibrate too. Without a specific white point there can not be consistency from display to display. 
  • The DICOM Supplement 100 "Color Softcopy Presentation State” does not specify how to verify and evaluate the colorcalibration like it is done for DICOM GSDF. Adjustment and verification of the white point over the entire dynamic range is not described in DICOM 


1) H. Roehrig et al:”Color calibration and color-managed displays: Does the calibration method matter?” Proc SPIE Vol 7627; pp 76270K1-13. 
2) AAPM TG18 report ~ samei/tg18 
3) Digital Imaging and Communications in Medicine (DICOM) 
4) Part 14: Grayscale Standard Display Function
5) Supplement 100: Color Softcopy Presentation State 
6) CIE Space ~ hoffmann/ciexyz29082000.pdf 
7) ICC profiles 
8) Colormanangement
9) Colorimetry -- Part 1: CIE standard colorimetric observers ISO 11664-1:2007 (CIE S 014-1/E:2006) 
10) Colorimetry -- Part 2: CIE standard illuminants ISO 11664-2:2007 (CIE S 014-2/E:2006)
11) Colorimetry -- Part 4: CIE 1976 L*a*b* Colour space ISO 11664-4:2008 (CIE S 014-4/E:2007)

No comments:

Post a Comment