spkit.mea.find_bad_channels_idx

spkit.mea.find_bad_channels_idx(X, fs=25000, thr=2, stim_fhz=1, mnmx=[None, None], plot=False, plot_dur=2, verbose=1)

Identify the Bad Channels of MEA based on stimuli

Identify the Bad Channels of MEA based on stimuli

This function compute the average time of stimuli over to the maximum voltage. In MEA, ideally, a stimulus is provided with 1ms at negative max voltage (~ -1000) mV/µV and 1ms positive (~ +1000), in one cylce So, if stimulus last more than 2ms (thr) per cycle on either side, it is flagged as a bad electode.

Parameters:
X: nd-array

  • array of shape = (nch,n) = (number of channels, number of samples), multi-channel signal recording

fs: int, default = 25KHz

  • sampling frequency of signal, for MEA

thrscalar, default=2 (ms)

  • threshold duration for stimuli, if over the threshold on either side, channel is flagged as BAD

    Note

    Lower the threshold, more channels will be flagged, higher the threshold less channels will be flagged

stim_fhz: int, float

  • stimuli frequency, number of stimuli per second

mnmx: list, [min, max], default=[None, None]

  • voltage of stimuli, if passed [None, None],

  • default, then for each chennels, it is computed by its min, max value

plot: bool, default=False

  • if true, all channels are plotted with ‘plot_dur’ duration

plot_dur: float

  • used for plooting channels

verbose: 0, Off mode

  • 1, a few information printed

  • 2, list of all channel printed with computed values and BAD flag label

Returns:
bad_channels_idx: list of channel index which are flagged as BAD

See also

egm_features

Examples

>>> #sp.mea.find_bad_channels_idx
>>> import numpy as np
>>> import matplotlib.pyplot as plt
>>> import os, requests
>>> import spkit as sp
>>> # Download Sample file if not done already
>>> file_name= 'MEA_Sample_North_1000mV_1Hz.h5'
>>> if not(os.path.exists(file_name)):
>>>     path = 'https://spkit.github.io/data_samples/files/MEA_Sample_North_1000mV_1Hz.h5'
>>>     req = requests.get(path)
>>>     with open(file_name, 'wb') as f:
>>>             f.write(req.content)
>>> fs = 25000
>>> stim_fhz=1
>>> X,fs,ch_labels = sp.io.read_hdf(file_name,fs=fs,verbose=1)

>>> bad_channels_idx_1 = sp.mea.find_bad_channels_idx(X,thr=2,stim_fhz=stim_fhz,fs=fs,
>>>                      plot=False,plot_dur=2,verbose=1)
>>> print(bad_channels_idx_1)
    [23, 25, 27, 28, 31, 34, 36]

Examples using spkit.mea.find_bad_channels_idx

MEA: Step-wise Analysis: Example

MEA: Step-wise Analysis: Example