4  Signal processing

4.1 ECG

The ECG records the electrical signal from the heart. In a clinical record it is performed as a 12 lead recording. The ECG contains an initial P wave which originates from the atrium. The P wave is absent in patients with atrial fibrillation. The RR interval is used to measure the heart rate. A false assumption is that the heart rate is constant but measurement shows beat to beat variation. Heart rate variability is a feature of health.

4.1.0.1 ECG data

The ecg dataset of one patient contains 2048 observations collected at a rate of 180 samples per second.

library(ade4)
data("ecg") #ts object
#Time Series:
#Start = 0.31 
#End = 11.6822222222222 
#Frequency = 180
head(ecg)
[1] -0.104773 -0.093136 -0.081500 -0.116409 -0.081500 -0.116409
plot(ecg)

Let’s look now at the Hart dataset.

ECG<-read.csv("../../DataMining/python_journey/Heart/ECG/data.csv")

#ECG is a dataframe object

plot(as.ts(ECG)) # from base R

create a ts object by providing starting and end time and frequency. the plot of the data now has a new x limit.

ECG1<-ts(ECG$hart, start=c(0.31,13.7), frequency=180)

plot(ECG1)

EEGECG<-read.csv("../../DataMining/python_journey/Heart/ECG/eeg_stroke_ecg.csv")

#this data has 2 columns time and ECG

plot(as.ts(EEGECG$ECG))

ECG2<-ts(EEGECG$ECG, start=c(0.31,6.75), frequency=60)
plot(ECG2)

Data from Samsung smart watch series 5, acquired for 30 seconds at 500 Hz

library(tidyverse)
── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.2     ✔ readr     2.1.4
✔ forcats   1.0.0     ✔ stringr   1.5.0
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.2     ✔ tidyr     1.3.0
✔ purrr     1.0.1     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
ecg<-read.csv("./Ext-Data/20240226094349.csv") 
ecg1<-ts(ecg$Lead, frequency = 500)
plot(ecg1)

Here we illustrate the use of RHRV package to analyse ECG signal (Rodriguez-Linares et al. 2017). The code is provided on the RHRV website https://rhrv.r-forge.r-project.org/documentation.html. The example.beats data was download from this website and stored in the Data-Use folder. Here we make changes regarding the path of the data.

library(RHRV)
Loading required package: waveslim

waveslim: Wavelet Method for 1/2/3D Signals (version = 1.8.4)

Attaching package: 'waveslim'
The following object is masked from 'package:lubridate':

    pm
Loading required package: nonlinearTseries
Registered S3 method overwritten by 'quantmod':
  method            from
  as.zoo.data.frame zoo 

Attaching package: 'nonlinearTseries'
The following object is masked from 'package:grDevices':

    contourLines
Loading required package: lomb
hrv.data = CreateHRVData()
hrv.data = SetVerbose(hrv.data, TRUE)

#setwd("C:/RHRV")
hrv.data = LoadBeatAscii(hrv.data, "example.beats.txt", 
                         RecordPath = "./Data-Use/" )
 Loading beats positions for record: example.beats.txt 
 Path: ./Data-Use/ 
 Scale: 1 
 Date:  01 / 01 / 1900 
  Time:  00 : 00 : 00 
 Number of beats: 17360 
plot(hrv.data$Beat$Time)

hrv.data = BuildNIHR(hrv.data)
 Calculating non-interpolated heart rate 
 Number of beats: 17360 
PlotNIHR(hrv.data)
 Plotting non-interpolated instantaneous heart rate 
 Number of points: 17360 

hrv.data = FilterNIHR(hrv.data)
 Filtering non-interpolated Heart Rate 
 Number of original beats: 17360 
 Number of accepted beats: 17248 
PlotNIHR(hrv.data)
 Plotting non-interpolated instantaneous heart rate 
 Number of points: 17248

4.2 EEG signal processing

The eegUtils package has useful methods for plotting EEG. https://craddm.github.io/eegUtils/index.html. According to the site, it has functions for importing data from Biosemi, Brain Vision Analyzer, and EEGLAB.

#devtools::install_github("mne-tools/mne-r")
#remotes::install_github("craddm/eegUtils@develop")
library(eegUtils)

Attaching package: 'eegUtils'
The following object is masked from 'package:stats':

    filter
plot_butterfly(demo_epochs)
Creating epochs based on combinations of variables: epoch_label participant_id 

topoplot(demo_epochs, 
         time_lim = c(.22, .25 ))
Creating epochs based on combinations of variables: epoch_label participant_id 
Using electrode locations from data.
Plotting head r 95 mm

The following data is from https://datashare.ed.ac.uk/handle/10283/2189

library(eegUtils)
library(R.matlab)
limo_test <- import_set("limo_dataset_S1.set")
limo_cont <- R.matlab::readMat("continuous_variable.mat")
limo_cat <- readr::read_csv("categorical_variable.txt",
                            col_names = c("cond_lab"))

The MNE-R package is an interface to MNE package in Python. It is an excellent package for signal processing.