DICOM PS3.17 2019c - Explanatory Information

JJJJ.3 Classification of Multi-energy Images

The following Multi-energy image types and families are addressed in this supplement:

Classification of Multi-energy Images

Figure JJJJ.3-1. Classification of Multi-energy Images


Standard CT Image (CT Image IOD, Enhanced CT Image IOD)

Images created using ME techniques, for example, in case of the creation of conventional appearing CT images out of two energy spectra or images created with only one of the multiple energies acquired. No new image type definitions are needed but new optional Attributes are needed.

Objective Image Family
Virtual Monoenergetic Image

Each real-world value mapped pixel represents CT Hounsfield units and is analogous to a CT image created by a monoenergetic (of a specific keV value) X-Ray beam. In certain cases, the image impression (quality) will allow a better iodine representation and better metal artifact reduction. Monoenergetic images are sometimes colloquially referred to as monochromatic images.

Effective Atomic Number Image

Each real-world value mapped pixel represents Effective Atomic Number (aka. "Effective Z") of that pixel.

Electron Density Image

Each real-world value mapped pixel represents a number of electrons per unit volume (N) in units of 1023/ml or a relative electron density to water (N/NWater). Electron density is used commonly in radiotherapy.

Material Quantification Image Family

These image types characterize the elemental composition of materials in the image. They provide material quantification using a physical scale. Pixel values can be in HU or in equivalent material concentration (e.g., mg/ml). The following image types belong to this family:

Material-Specific Image

Each real-world value mapped pixel value represents a property of a material such as attenuation, concentration or density.

Material-Removed Image

An image where the attenuation contribution of one or more materials has been removed. Each real-world value mapped pixel may be adjusted to represent the attenuation as if the pixel was filled with the remaining materials. For pixels that did not contain any of the removed material(s), the pixel values are unchanged. For example, in virtual-unenhanced (VUE) or virtual-non-contrast (VNC) image the attenuation contribution of the contrast material is removed from each pixel.

Fractional Map Image

Each real-world value mapped pixel represents the fraction of a specific material present in the pixel. Since Fractional Map Images are generated as a set, the sum of the real-world values for all the Fractional Map Images is 1 for each pixel.

Value-Based Map Image

Each real-world value mapped pixel represents a certain value for a specified material (the exact interpretation of the value range has to be defined by the user).

Material Visualization Image Family

These image types allow visualizing material content, usually with colors (color maps, color overlays, blending, etc.).

Material-Modified Image

CT Image where pixel values have been modified to highlight a certain target material (either by partially suppressing the background or by enhancing the target material), or to partially suppress the target material. The image units are still HU, so they may be presented similarly to conventional CT Images. The values of some pixels in the Material-Modified Image are intentionally distorted for better visualization of certain materials (i.e. making tendon more visible). Thus, the image may not be used for quantification, unlike a Material-Removed Image, which can.

Color Image

Implementations of Material Visualization Images use existing DICOM objects (Blending Presentation State, Secondary Capture Image (used as fallback)).

DICOM PS3.17 2019c - Explanatory Information