Tag | (0040,08EA) |
---|---|
Type | Required (1) |
Keyword | MeasurementUnitsCodeSequence |
Value Multiplicity | 1 |
Value Representation | Sequence (SQ) |
Units of measurement.
Only a single Item shall be included in this Sequence.
See Section C.7.6.16.2.11.1 for further explanation.
Part of the Real World Value Mapping Functional Group Macro with usage: U
The Items in the Real World Value Mapping Sequence (0040,9096) may be used to translate stored values into Real World Values when there is such a relationship. The Real World Value Mapping Sequence (0040,9096) is independent of the Modality LUT (or Pixel Value Transformation Macro), as illustrated in Figure C.7.6.16-6.
Each Item specifies the range of stored values as well as the associated mapping function. Each Item can specify either a linear mapping, using Real World Value Slope (0040,9225) and Real World Value Intercept (0040,9224), or a non-linear mapping using Real World Value LUT Data (0040,9212). More than one Real World Value Mapping Item is allowed.
The range of stored pixel values specified by different Real Value World Mapping Sequence (0040,9096) Items can overlap (as illustrated in the example in Figure C.7.6.16-7).
Figure C.7.6.16-6. The Real World Value LUT and the Image Viewing Pipeline
For example, MR images may contain data that is not only the result of the physical/chemical properties of the scanned anatomy, but may also contain information that is representing Real World Values, such as, temperature [in degrees C], flow [in l/min], speed [in m/sec], relative activity [in %], relative contrast enhancement [in %], diffusion [in sec/mm^{2}], etc.
In some cases the conversion from Stored Values to Real World Values can be linear (through "slope" and "intercept") or non-linear (through look-up tables).
Both transformation methods can be applied to one range of stored values. Overlapped ranges might be used for different representations such as log versus linear scales or for different representations in units such as cm/sec versus mm/sec. Alternative methods can be identified by the labels assigned to the transformations.
The Real World Value First Value Mapped (0040,9216) and Real World Value Last Value Mapped (0040,9211) Attributes describe the range of stored pixel values that are mapped by the Sequence Item. Stored pixel values less than the first value mapped, or greater than the last value mapped have no real value attached.
When the Real World Value Intercept (0040,9224) and Real World Value Slope (0040,9225) Attributes are supplied, the stored value (SV) is converted to a Real World Value (RV) using the equation:
RV = (Real World Value Slope) * SV + Real World Value Intercept
When the Real World Value LUT Data (0040,9212) Attribute is supplied, Real World Values are obtained via a lookup operation. The stored pixel value of the first value mapped is mapped to the first entry in the LUT Data. Subsequent stored pixel values are mapped to the subsequent entries in the LUT Data up to a stored pixel value equal to the last value mapped.
The number of entries in the LUT data is given by:
Number of entries = Real World Value Last Value Mapped- Real World Value First Value Mapped + 1
If the stored pixel values that are encoded as floating point rather than integer values, only the equation and not the lookup operation is defined.
The Value Representation of Real World Value First Value Mapped (0040,9216) and Real World Value Last Value Mapped (0040,9211) (US or SS) is determined by the value of Pixel Representation (0028,0103) in the case of integer pixel data, or is SS in the case of floating point pixel data if these Attributes are present.
In practice, integer values may well be sufficient to define an input range that exceeds the actual stored floating point pixel value range. The reason to permit floating point alternatives to an integer range is that sometimes the stored floating point values may be very large and exceed what can be specified as an integer value.
The physical units for the Real World Values obtained from the Sequence Item are given by the Measurement Units Code Sequence (0040,08EA).
The quantity that the Real World Values represent may be described by the Quantity Definition Sequence (0040,9220), which consists of a list of name-value pairs, in which the coded concept name specifies what aspect of the physical quantity is being described.
For example, Relative Regional Blood Flow may be described by units and quantity as follows:
Measurement Units Code Sequence (0040,08EA) = ({ratio}, UCUM, "ratio")
Quantity Definition Sequence (0040,9220):
Additional information about how the relative blood flow was derived, e.g., the reference region used, can also be encoded as name-value pairs in the Quantity Definition Sequence (0040,9220). See the example in Annex EEEE “Encoding Diffusion Model Parameters for Parametric Maps and ROI Measurements (Informative)” in PS3.17.
For example, the Apparent Diffusion Coefficient (ADC) may be described by units and quantity as follows:
Measurement Units Code Sequence (0040,08EA) = (mm2/s, UCUM, "mm2/s")
Quantity Definition Sequence (0040,9220):
Additional information about how the ADC was derived, e.g., the model used, method of fitting and acquisition b-values used, can also be encoded as name-value pairs in the Quantity Definition Sequence (0040,9220). Other diffusion models and quantities are also defined. See the example in Annex EEEE “Encoding Diffusion Model Parameters for Parametric Maps and ROI Measurements (Informative)” in PS3.17.
The Quantity Definition Sequence (0040,9220) describes only the stored pixel values that are mapped using the Real World Values Mapping, and does not describe derived values from multiple pixels to which the Real World Values Mapping applies.
I.e., the mapping is a "point" operation, and as a consequence various modifiers that might be applied to a group of pixels, such as in an ROI, should not be used. E.g., an ROI encoded in a Structured Report using TID 1419 "ROI Measurements" might be the mean or maximum value (e.g., SUVbw mean or SUVbw max), and be encoded with (121401, DCM, "Derivation") = (373098007, SCT, "Mean") or (56851009, SCT, "Maximum"), respectively. These would not be appropriate to use within Quantity Definition Sequence (0040,9220), unless the individual pixel values were themselves derived in such a manner, e.g., when multiple images are averaged together. Thus the Content Items used in an SR to describe an ROI might be a superset of the name-value pairs used in Quantity Definition Sequence (0040,9220).
Figure C.7.6.16-7. Example of Mapping Stored Values to Real World Values