Biology Reference
In-Depth Information
20
0 decelerations
1 deceleration
2
3
4
5
6
7
8
15
10
5
5
0
10
15
20
Height of deceleration (S.D. of baseline)
FIGURE 6-10.
Increase in SA as a function of number (different curves) and magnitude of HR deceleration. (From
Kovatchev, B. P., Farhy, L. S., Hanging, C., Griffin, M. P., Lake, D. E., Moorman, J. R. [2003]. Sample
asymmetry analysis of heart rate characteristics with application to neonatal sepsis and systemic
inflammatory response syndrome. Pediatric Research, 54, 892-898. Used by permission.)
of approximate height 14 SD, SA is 5.7. Thus, SA is sensitive to
gradually increasing frequency and magnitude of transient
decelerations, which makes it a promising measure for abnormality in
HR data samples.
IX. TIME-DEPENDENT MEASURES: SAMPLE ENTROPY
It should be noted that the timing component in our discussion so far
appeared only in Figure 6-3, where consecutive samples of 4096 RRIs
were sequentially recorded. The measures subsequently introduced—
SD, skewness, and SA—are based solely on the statistical distribution of
RRIs. Thus, none of these measures takes into account the order of RRIs
within each sample and the potential contribution of the exact timing
sequence of the RRIs is ignored. Recall, however, Figure 6-3, which
suggests that a change in the time pattern of the RRIs may be indicative
of an upcoming sepsis episode. The statistical approach may be
insufficient for thoroughly addressing the problem. To emphasize this
point even further, if we take a sample of 4096 RRIs and calculate the
samples' SD, skewness, and SA, they will be exactly the same as the SD,
skewness, and SA of any reshuffling of the sample. Clearly, however, the
evolution of the RRI sequences with time is a factor that should not be
ignored, but to study this factor we shall need to employ methods
