Functional imaging can create Parametric Maps showing a functional relation between the anatomical region and the specific functional activity. For display purposes it is useful to show this functional activity with the use of a color LUT on the related anatomical image. To be able to do this it is necessary to include a Palette Color Lookup Table for the Parametric Map and define how to map the (floating point) values to a specific RGB value.
For a correct mapping it is important to know what range of continuous values needs to be mapped to the discrete range of RGB values of the LUT. For this the Minimum Stored Value Mapped (0028,1231) and the Maximum Stored Value Mapped (0028,1232) are defined. All values between the minimum and maximum will be distributed in a linear manner to the Palette Color Lookup Table that is supplied.
The usage of floating point values for the stored values removes the need for a Real World Value transformation other than the identity transformation.
This example illustrates BOLD fMRI activation data for a bipolar motor paradigm stored as a floating point parametric map encoding ‘t’ (statistical) Real World Values. Each voxel’s value represents how well the BOLD time series information at that location of the brain fits the general linear model (GLM) of the fMRI block paradigm pattern (right or left versus control, no movement). Right and left have been encoded as positive and negative t values, respectively.
The Double Float Minimum Stored Value Mapped and Maximum Stored Value Mapped in this case are -16.739 and 21.434, respectively. This range will be mapped to the low and high ends of the LUT applied to this activation map. In this case the Minimum Stored Value Mapped and Maximum Stored Value Mapped are equal to the RWV Minimum and Maximum, respectively, in the activation map data. Note several compelling reasons for the range to be different from the RWV Minimum and RWV Maximum:
-
Centering the RWV zero value on some desired index of the LUT; e.g., choosing -21.434 to +21.434 to properly center RWV zero on the middle of the LUT (presumably to match the LUT design).
-
Choosing a narrower range of Minimum Stored Value Mapped and/or Maximum Stored Value Mapped (negative and/or positive), i.e., windowing, to maximize the dynamic range of the LUT for critical RWV range(s).
-
Specifying a predetermined Minimum Stored Value Mapped and Maximum Stored Value Mapped regardless of the actual RWV data, in order to have key RWV transitions match LUT color effects, e.g., generally accepted hyperperfusion and hypoperfusion transition points in cerebral blood flow (CBF) maps.
For the purpose of this example, the full RWV range of the activation map is appropriate to display with the full range of the Spring Color Palette.
As the activation map without threshold suggests, areas outside the brain have been masked off. These would be coded with Padding values in the parametric map.
Thresholding (not part of the parametric map) will be applied for positive and/or negative ranges. Note that this operation does not change the color mapping (i.e., RWV x corresponds to LUT entry j) but only the opacity of voxels outside the range (forcing A=0 or transparent).
Other visualization methods such as smoothing and overall opacity may be applied to the colored, thresholded activation map.