Table of Contents
mse - calculates multiscale entropy (MSE) of one or multiple data sets
mse [ option ... ]
This program reads data (a column of
numbers in text form) from its standard input (or optionally from a list
of input files; see the -F option below), creates a set of coarse-grained
series with a user-selected range of scales, and calculates sample entropy
(SampEn) for each coarse-grained series. The user may specify the pattern
length (m) and the similarity criterion (r) for the SampEn calculations,
and may select any contiguous subset of the input data. The program can
also iterate over a set of user-specified m and r values. If multiple data
sets are provided, these calculations are performed separately for each
data set, and then the mean and sample deviation of these SampEn values
at each scale are calculated. The results are written to the standard output.
Options are:
- -a dn
- Set the scale increment to dn (1-40; default: 1).
- -b dm
- Set the m increment to dm (1-10; default: 1).
- -c dr
- Set the scale increment
to dr (>0; default: 0.05).
- -F listfile
- Read and process up to 100 data sets
from the files named in listfile (a text file containing a column of file
names). The output includes results for each file, followed by means and
sample deviations of the per-file results.
- -i n
- Begin the analysis with row
n of (each) data set. Rows are numbered beginning with 0; by default,
analysis begins with row 0.
- -I n
- Stop the analysis with row n of (each) data
set. By default, analysis ends at row 39999, or at the end of the data
set if there are fewer rows.
- -m n
- Set the minimum m (pattern length for SampEn)
to n (1-10; default: 2).
- -M n
- Set the maximum m to n (1-10; default: 2).
- -n n
- Set the maximum scale for coarse-graining to n (1-40; default: 20).
- -r n
- Set
the minimum r (similarity criterion for SampEn) to n (>0; default: 0.15).
- -R n
- Set the maximum m to n (>0; default: 0.15).
To obtain MSE results
with default settings for the first 40,000 RR intervals in record nsr047
of the Normal Sinus Rhythm RR Interval Database (nsr2db), use this command:
ann2rr -r nsr2db/nsr047 -a ecg -A -i s4 | mse
(ann2rr(1)
reads the ecg annotations for the record, calculates the RR
intervals, and writes them to its standard output, which is connected to
mse's standard input. mse reads the intervals and calculates SampEn for scales
1 through 20, for m = 2 and r = 0.15.)
For additional examples illustrating
the use of mse's options, see the tutorial at http://www.physionet.org/physiotools/mse/tutorial/.
Madalena Costa (mcosta@fas.harvard.edu)
http://www.physionet.org/physiotools/mse/mse.c
Table of Contents