Optimal monochromatic color combinations for fusion imaging of FDG-PET and diffusion-weighted MR images

Abstract

Objective

To investigate the optimal monochromatic color combination for fusion imaging of FDG-PET and diffusion-weighted MR images (DW) regarding lesion conspicuity of each image.

Methods

Six linear monochromatic color-maps of red, blue, green, cyan, magenta, and yellow were assigned to each of the FDG-PET and DW images. Total perceptual color differences of the lesions were calculated based on the lightness and chromaticity measured with the photometer. Visual lesion conspicuity was also compared among the PET-only, DW-only and PET-DW-double positive portions with mean conspicuity scores. Statistical analysis was performed with a one-way analysis of variance and Spearman’s rank correlation coefficient.

Results

Among all the 12 possible monochromatic color-map combinations, the 3 combinations of red/cyan, magenta/green, and red/green produced the highest conspicuity scores. Total color differences between PET-positive and double-positive portions correlated with conspicuity scores (ρ = 0.2933, p < 0.005). Lightness differences showed a significant negative correlation with conspicuity scores between the PET-only and DWI-only positive portions. Chromaticity differences showed a marginally significant correlation with conspicuity scores between DWI-positive and double-positive portions.

Conclusions

Monochromatic color combinations can facilitate the visual evaluation of FDG-uptake and diffusivity as well as registration accuracy on the FDG-PET/DW fusion images, when red- and green-colored elements are assigned to FDG-PET and DW images, respectively.

Appendix

The Yxy system and CIE L*a*b* (CIELAB) system proposed by the International Commission on Illumination (CIE, Vienna, Austria) [11] were adopted in this analysis. The Yxy system is a widely used method to specify colors, where the parameter Y indicates the lightness of a color, and the chromaticity of a color is specified by the remaining two parameters x and y. CIELAB is a complete color space that describes all the colors that the human eye perceives. The three coordinates of CIELAB stand for the lightness of a color (L*), its position between red/magenta and green (a*), and its position between yellow/blue (b*).

The calculation of color differences requires the conversion of Yxy values to L*a*b* values. This can be done with a white reference point, which is the maximum white level on the monitor of interest. The color difference indicated by the Euclidean distance between the two colors in CIELAB space is known as the CIE 1976 formula. The mathematical details are as follows.

First, the measured Yxy values are converted into XYZ values.

$$\begin{gathered} X=Y \cdot x/y \hfill \\ Z=Y \cdot (1 - x - y)/y. \hfill \\ \end{gathered}$$

L*a*b* values are calculated according to the following equations:

$$L*=116f\left( {\frac{Y}{{{Y_n}}}} \right) - 16$$

$$a*=500\left\{ {f\left( {\frac{X}{{{X_n}}}} \right) - f\left( {\frac{Y}{{{Y_n}}}} \right)} \right\}$$

$$b*=200\left\{ {f\left( {\frac{Y}{{{Y_n}}}} \right) - f\left( {\frac{Z}{{{Z_n}}}} \right)} \right\}$$

$$f\left( {\frac{t}{{{t_n}}}} \right)={\left( {\frac{t}{{{t_n}}}} \right)^{1/3}}\quad {\text{if}}\quad \frac{t}{{{t_n}}}>0.008856$$

$$f\left( {\frac{t}{{{t_n}}}} \right)=7.787\left( {\frac{t}{{{t_n}}}} \right)+\frac{{16}}{{116}}\quad {\text{if}}\quad \frac{t}{{{t_n}}}>0.008856,$$

where X_n, Y_n, and Z_n are the values of X, Y, and Z for the appropriate white reference value.

Finally, color differences are computed with the following equations.

Total color difference:

$${{\Delta}}\,E=\sqrt {{{\left( {L_{1}^{*} - L_{2}^{*}} \right)}^2}+{{\left( {a_{1}^{*} - a_{2}^{*}} \right)}^2}+{{\left( {b_{1}^{*} - b_{2}^{*}} \right)}^2}} .$$

The chromaticity difference is the color difference after accounting for lightness differences.

$${\text{Lightness difference:}}\quad {{\Delta}}\,L=L_{1}^{*} - L_{2}^{*}$$

$${\text{Chromaticity difference:}}\quad {{\Delta}}\,ab=\sqrt {{{\left( {a_{1}^{*} - a_{2}^{*}} \right)}^2}+{{\left( {b_{1}^{*} - b_{2}^{*}} \right)}^2}} .$$