spkit.get_activation_time

spkit.get_activation_time(x, fs=1, method='min_dvdt', gradient_method='fdiff', sg_window=11, sg_polyorder=3, gauss_window=0, gauss_itr=1, **kwargs)

Get Activation Time based on Gradient

Get Activation Time based on Gradient

Activation Time in cardiac electrograms refered to as time at which depolarisation of cells/tissues/heart occures.

For biological signals (e.g. cardiac electorgram), an activation time in signal is reflected by maximum negative deflection, which is equal to min-dvdt, if signal is a volatge signal and function of time x = v(t) However, simply computing derivative of signal is sometime misleads the activation time locatio, due to noise, so derivative of a given signal has be computed after pre-processing

Parameters:
x: 1d-array,

  • input signal

fs: int,

  • sampling frequency, default fs=1, in case only interested in loc

methoddefault = “min_dvdt”

  • Method to compute Activation Time, “min_dvdt” is the used in literature, however, it depends on the kind of signal. For Action Potential Signal, “max_dvdt” is considered as Activation Time for Unipolar Electrogram, “min_dvdt” is consodered. it can be chosen one of {“max_dvdt”, “min_dvdt”, “max_abs_dvdt”},

  • Some literation suggests to use max_dvdt instead of min_dvdt, but mostly agree on min_dvdt

gradient_method: default =”fdiff”,

  • {“fdiff”, “fgrad”,”sgdiff”,”sgdrift_diff”,”sgsmooth_diff”, “gauss_diff”}

  • Method to compute gradient of signal

  • one of {“fdiff”, “fgrad”, “npdiff”,”sgdiff”,”sgdrift_diff”,”sgsmooth_diff”, “gauss_diff”}

  • check signal_diff for more details on the method

  • if signal is noisy try “sgsmooth_diff” or “gauss_diff”

Parameters for gradient_method:

  • used if gradient_method in one of {“sgdiff”,”sgdrift_diff”,”sgsmooth_diff”, “gauss_diff”}

  • sg_window: sgolay-filter’s window length

  • sg_polyorder: sgolay-filter’s polynomial order

  • gauss_window: window size of gaussian kernel for smoothing,

  • check signal_diff for more details on the method

Returns:
atactivation time (ms)
loc: index
mag: magnitude of deflection at loc
dxderivative of signal x

See also

get_repolarisation_time, mea.activation_time_loc, mea.activation_repol_time_loc

Notes

#TODO

References

Examples

#sp.get_activation_time
import numpy as np
import matplotlib.pyplot as plt
import spkit as sp
x, fs = sp.data.ecg_sample(sample=1)
x = sp.filterDC_sGolay(x,window_length=fs//2+1)
#x = sp.filter_smooth_gauss(x,window_length=31) 
x = x[int(0.02*fs):int(0.4*fs)]
t = 1000*np.arange(len(x))/fs
# It is a good idea to smooth the signal or use gradient_method = gauss_diff
at,loc,mag,dx= sp.get_activation_time(x,fs=fs,method='min_dvdt',gradient_method='fdiff')

plt.figure(figsize=(10,4))
plt.plot(t,x)
plt.axvline(at,color='r',label=f'AT = {at:0.2f} ms')
plt.legend()
plt.xlabel('time (ms)')
plt.title('Activation Time (Depolarisation)')
plt.show()
../../_images/spkit-get_activation_time-1.png