Skip to content

cbrnr/HeartBeats.jl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

License

HeartBeats.jl

HeartBeats.jl provides a heartbeat detector based on the approach described by Pan & Tompkins (1985). It is based on the implementation available in the Python package SleepECG.

Installation

Use the package manager to add HeartBeats.jl by typing ] add HeartBeats in the Julia REPL.

Example

HeartBeats.jl contains a short example ECG dataset taken from scipy.misc.electrocardiogram(). The function example_ecg() returns this data, which was sampled with a sampling frequency of 360 Hz, as a Vector{Float64}. We can use this dataset to showcase the detect_heartbeats() function:

using HeartBeats

ecg = example_ecg()
fs = 360  # sampling frequency

beats = detect_heartbeats(ecg, fs)

The beats array will contain all detected R peak locations.

Benchmark

The detector is based on the Python implementation available in SleepECG. It is about 18× faster than the Python implementation and only 2× slower than the C implementation. Follow these steps to reproduce the benchmark:

  1. Export all data records used in the 'runtime' benchmark by including export_records = True in config.yml (refer to the SleepECG documentation for details on how to set up and run the benchmarks). This will generate 15 text files.
  2. Move those text files to a folder that you can access from Julia (i.e. set the Julia working directory accordingly).
  3. Run the following code snippet to benchmark the runtime for 60 minute data segments (note that you need to add the CSV package):
using CSV, HeartBeats

function run_benchmark()
    total = 0
    fs = 128
    files = filter(x -> endswith(x, ".txt"), readdir(".", join=true))
    for file in files
        println(file)
        data = CSV.File(file, comment="#")
        ecg = view(data[:ecg], 1:60*60*fs)  # 60 minutes
        stats = @timed detect_heartbeats(ecg, fs)
        total += stats.time
    end
    println("Total: $total, Average: $(total / length(files))")
    total
end

run_benchmark()