C.10.9.1 Waveform Attribute Descriptions

C.10.9.1.1 Multiplex Group Time Offset

Multiplex Group Time Offset (0018,1068) specifies the offset time in milliseconds from a reference time to the first sample of the multiplex group. The reference time is the Acquisition DateTime (0008,002A), if present in the SOP Instance.

In all other cases, the offset is from an arbitrary reference time that is the same for all Multiplex Groups in the SOP Instance; i.e., the Multiplex Group Time Offset allows only relative time synchronization between Multiplex Groups in the SOP Instance. The arbitrary reference time may nominally be assumed to be the Content Time (0008,0033).

C.10.9.1.2 Trigger Sample Position

The Trigger Sample Position (0018,106E) specifies the sample which was digitized at the same time as a synchronization trigger. Sample positions are enumerated by channel, with the first sample enumerated 1. This provides a single trigger sample location for all channels of the multiplex group. Although channels may not have been sampled synchronously (as specified by Channel Time Skew or Channel Sample Skew), for the purpose of determining the location of the trigger with an integer value position, all channels are considered to be synchronous.

C.10.9.1.3 Waveform Originality

Waveform Originality (003A,0004) shall have the value ORIGINAL if the Waveform Data samples are the original or source data, and shall have the value DERIVED if the Waveform Data samples have been derived in some manner from the sample data of other waveforms.

Notes : 1. The Waveform Originality (003A,0004) attribute is comparable to the Image Type (0008,0008) attribute value 1 (see C.7.6.1.1.2). Within a single Multiplex Group, all channels shall have the same Originality value.

2. Waveform data which has been transcoded from a non-DICOM format may have Waveform Originality value ORIGINAL if the samples are unchanged from the originally acquired waveform samples.

C.10.9.1.4 Channel Definition Sequence

C.10.9.1.4.1 Channel Source and Modifiers

Channel Source Sequence (003A,0208) identifies the metric (quality being measured, e.g., voltage or pressure), the anatomical position of the sensor or probe, the function of the channel (e.g., measurement or stimulus), and any particulars of technique which affect those parameters (e.g., pull-back across multiple anatomic sites, or differential input from two distinct sites). If the full semantics of the source is not carried in a single coded entry (e.g., if it specifies the location but not the metric), additional qualifiers are identified in Channel Source Modifiers Sequence (003A,0209) coded entries.

When a single sensor channel is used to collect a waveform from two (or more) anatomic sites, e.g., in hemodynamic pull-back procedures, multiple Channel Source Modifier items will identify the sequence of sites, if not encoded in the semantics of the Channel Source Coded Entry. Transition times from one site to another may be indicated with an Annotation, or pull-back rate may be indicated with an Acquisition Context Sequence Item (see Section C.7.6.14).

The Baseline (default) Context IDs are defined by IOD in accordance with Section A.34. Restrictions in the IOD may also determine the pattern of specification of the waveform source, i.e., which item is to be encoded in the Channel Source sequence, and the order in which Channel Source Modifier items are to be encoded. Unless otherwise specified, pattern of specification of the waveform source shall be:

  1. If the function of the channel is not measurement, the function (and optionally additional parameters of the channel source) shall be encoded in the Channel Source item.

  2. If the function of the channel is measurement of a waveform originating in the patient (the implicit default function), the metric (and optionally additional parameters of the channel source) shall be encoded in the Channel Source item.

  3. If not encoded in the Channel Source item, and a particular technique needs to be encoded, that technique shall be encoded in the first Channel Source Modifier item.

Note: For example, an intracardiac measurement of a pressure waveform across the mitral valve by means of a catheter pullback may be encoded in one of the following three ways (using pseudo-coded terminology), depending on the availability of coded terms with sufficient expressive power:

Channel Source Channel Source Modifiers
X-2311 “pressure measurement” T-7663 “pullback” C-2001 “mitral valve”
X-2123 “pressure measurement, pullback” C-2001 “mitral valve”
X-1234 “pressure measurement, mitral valve, pullback” (none required)

C.10.9.1.4.2 Channel Sensitivity and Channel Sensitivity Units

Channel Sensitivity is the nominal value of one unit (i.e., the least significant bit) of each waveform sample in the Waveform Data attribute (5400,1010). It includes both the amplifier gain and the analog-digital converter resolution. It does not relate the vertical scaling of a waveform on a particular display.

Note: The Defined (default) Context Group for Channel Sensitivity Units Sequence is CID 3082 Waveform Units of Measurement, which includes all the commonly used measurement values. Units of measurement not included in the default list can be specified using the more general CID 82 Units of Measurement, or a local Coding Scheme. The Defined Context ID may be replaced in a specialization of the IOD.

Channel Sensitivity Correction Factor (003A,0212) is the ratio of the actual (calibrated) value to the nominal Channel Sensitivity specified in Data Element (003A,0210). Thus a waveform sample value multiplied by the Channel Sensitivity value provides the nominal measured value in Channel Sensitivity Units, and that nominal value multiplied by the Channel Sensitivity Correction Factor provides the calibrated measured value.

C.10.9.1.4.3 Channel Skew and Channel Offset

Skew is also known as a sub-sample time delay, typically caused by using a multiplexed analog to digital converter which switches from channel to channel. For analysis it may be important to know if the analog channels were all latched simultaneously or sequentially and then digitized. Skew may be represented as time offset in seconds, or a fractional number of samples.

Separate and additional to skew is an offset time adjustment (sometimes called latency) by which one waveform channel is displaced significantly relative to others before sampling.

Note: As an example, a hemodynamic pressure is measured at the external end of a catheter, and thus its measurement is delayed by the time for the pressure wave to propagate down the catheter. With a dual catheter measurement, two signals may be acquired at the same time, but one arrives by a longer distance (e.g., a pulmonary capillary wedge pressure, compared to a left ventricular pressure). To obtain an accurate comparison of the waveforms (e.g., the gradient across the mitral valve), one waveform has to be offset (perhaps as much as 30 ms) to synchronize them.

C.10.9.1.4.4 Waveform Bits Stored

Waveform Bits Stored (003A,021A) specifies the number of significant bits within the Waveform Bits Allocated of each sample, for signed or unsigned integers.

If Waveform Sample Value Representation is MB or AB, Waveform Bits Stored shall be 8.

C.10.9.1.4.5 Channel Minimum and Maximum Value

Channel Minimum and Maximum Value attributes (5400,0110) and (5400,0112) may be used to send the analog-to-digital converter limits (i.e., the clipping levels).

Note: These values do not represent the maximum and minimum values in the data set, but rather the valid range of values.

C.10.9.1.5 Waveform Bits Allocated and Waveform Sample Interpretation

Waveform Bits Allocated (5400,1004) specifies the number of bits allocated for each sample, and Waveform Sample Interpretation (5400,1006) specifies the data representation of each waveform sample. Waveform Bits Allocated shall be a multiple of 8. These data elements are related, and their defined terms are specified in Table C.10-10.

Table C.10-10

Waveform Bits Allocated and Waveform Sample Interpretation

Waveform Bits Allocated - Defined Terms Waveform Sample Interpretation - Defined Terms Waveform Sample Interpretation Meaning
8 SB signed 8 bit linear
UB unsigned 8 bit linear
MB 8 bit mu-law (in accordance with ITU-T Recommendation G.711)
AB 8 bit A-law (in accordance with ITU-T Recommendation G.711)
16 SS signed 16 bit linear
US unsigned 16 bit linear

Notes: 1. The set of valid values from within this table may be constrained by definition of the IOD (see Section A.34).2. mu-law and A-law encoding is without the alternate bit inversion used for PCM transmission through the telephone network.

This representation also applies to the Channel Minimum and Maximum Data Values, and Waveform Padding Value.

C.10.9.1.6 Waveform Padding Value

Equipment which produces digitized waveform curves may encode a specific value when the source is disconnected or otherwise invalid. This value is encoded like the Waveform Data attribute with one sample only.

The Waveform Padding Value need not be within the range specified by the Channel Minimum and Maximum Data Values.

C.10.9.1.7 Waveform Data

Each sample shall be encoded using the defined Waveform Sample Interpretation (5400,1006), using the defined number of Waveform Bits Stored (003A,021A) right justified in the sample. If the number of Waveform Bits Stored is less than the number of bits in Waveform Bits Allocated, the sign bit shall be extended to the highest order bit of the data sample.

Data values are encoded interleaved, incrementing by channel and then by sample (i.e., C1S1, C2S1,C3S1, ... CnS1, C1S2, C2S2, C3S2, ... CnSm), with no padding or explicit delimitation between successive samples. Cx denotes the channel defined in the Channel Definition Sequence Item in item number x.

Notes: 1. With 8-bit Waveform Data, there may be an odd number of channels and an odd number of samples; see PS3.5 for rules on encoding.

2. The sign bit extension rule differs from the rules for pixel data, which do not require sign extension.

C.10.9.1.8 Waveform Data Display Scale

Waveform Data Display Scale (003A,0230) specifies the recommended display scale in the time dimension for the waveform data in units of mm/s. The display application needs to know the horizontal pixel scaling of the display device to effectively apply this attribute. The Waveform Data Display Scale (003A,0230), divided by the Sampling Frequency (003A,001A), times the display scaling in pixels/mm gives the number of horizontal pixels between subsequent samples.

Notes: 1. This presumes a conventional waveform display with time in the horizontal dimension, and waveform measured quantity in the vertical dimension.

2. This attribute is useful for size consistency between softcopy display and classical hardcopy ECG printouts. E.g., for the channels of a multiplex group whose Sampling Frequency is 400 Hz, and whose Waveform Data Display Scale (003A,0230) is 25 mm/s, and which is being displayed on a monitor at 4.1 pixels/mm, the horizontal interval between samples is

25 / 400 * 4.1 = 0.25625 pixels

How the display application renders samples at fractional pixel spacing is implementation dependent.

C.10.9.1.9 Channel Position

Channel Position (003A,0245) specifies the position of the channel display baseline in the Presentation Group vertical display dimension. The channel position is relative to the top of the display area. Each Presentation Group display area is defined with vertical dimension positions of 0.0 (top of display area) to 1.0 (bottom of display area).

Notes: 1. This presumes a conventional waveform display with time in the horizontal dimension, and waveform measured quantity in the vertical dimension.

2. Compare Bounding Box Annotation Units (0070,0003) value DISPLAY in Section C.10.5 Graphic Annotation Module.

Channel sample value 0 maps to the Channel Position (003A,0245) display baseline. The real world value associated with the baseline is specified in Channel Baseline (003A,0213).

Multiple channels may overlap, and may share the same Channel Position (003A,0245).

C.10.9.1.10 Channel Display Scale

Fractional Channel Display Scale (003A,0247) or Absolute Channel Display Scale (003A,0248) specifies the vertical scaling of the channel display. If both attributes are provided, the selection of which to use for display control is implementation dependent.

Fractional Channel Display Scale (003A,0247) specifies the fraction of the Presentation Group vertical display dimension used to offset each unit of the waveform sample value from the baseline Channel Position (003A,0245). Positive offset values are towards the top of the display area.

Note: E.g., for a channel whose Fractional Channel Display Scale (003A,0247) value is 0.004, and whose Channel Position (003A,0245) value is 0.5, a sample whose value is -37 would be displayed at a position

0.5 - (-37 * 0.004) = 0.648

of the total vertical display dimension below the top of the Presentation Group display area.

Absolute Channel Display Scale (003A,0248) specifies the nominal physical distance on a display device (monitor) used to offset each unit of the waveform sample value from the baseline Channel Position (003A,0245). The display application needs to know the vertical pixel scaling of the display device to effectively apply this attribute. The Channel Sensitivity (003A,0210) value in Channel Sensitivity Units (003A,0211), divided by Absolute Channel Display Scale (003A,0248) value, gives the real-world scaling of the intended display.

Notes: 1. This attribute is useful for size consistency between softcopy display and classical hardcopy waveform printed charts. E.g., a channel whose Channel Sensitivity (003A,0210) value is 44 uV (per least significant bit), and whose Absolute Channel Display Scale (003A,0248) value is 0.44 mm, has a real-world display scale of 0.1 mV/mm.

2. E.g., for a channel whose Absolute Channel Display Scale (003A,0248) value is 0.44 mm, and which is being displayed on a monitor at 4.1 pixels/mm, a sample whose value is 107 would be displayed at a position

107 * 0.44 * 4 = 193.028

pixels above the Channel Position (003A,0245) baseline.

Display for waveform samples that would lie outside the Presentation Group display area is unspecified.

Negative values of Fractional Channel Display Scale (003A,0247) and Absolute Channel Display Scale (003A,0248) are allowed.