spkit.mea.get_stim_loc

spkit.mea.get_stim_loc(X, fs=25000, ch=0, fhz=1, N=None, method='max_dvdt', gradient_method='fdiff', plot=1, verbose=False, figsize=(15, 3), sg_window=11, sg_polyorder=3, gauss_window=0, gauss_itr=1)

Get Stimulus Locations in given multi-channel

Get Stimulus Locations in given multi-channel X (nch,n)

loc: location of stim

    loc = max-dv/dt

                  loc
                   |        +ve peak of stim 
                   |-------| 
                   |       |
                   |       |  
                   |       |           
        ---|-------|-------|---------
           |       |
           |       |
           |       |
           |-------|

-ve peak of stim

This function uses spkit.get_activation_time function to identify the locations of stimulation. Check help(sp.get_activation_time) for more details.

Parameters:

Xnd-array

• with shape = (nch,n), where nch: number of channels, n: number of samples,

• multi-channel signal recording

fs: int, default=25000 (25KHz) for MEA

• sampling frequency of signal,

ch: int, default=0

• channel number of extract stimlus locations,

fhz: int, float,

• frequency of stimulus, e.g. fhz=1 means 1 stimulus per second

Nint, default=None,

• Number of stimulus to be extracted

• if None, N = number_samples/fs/fhz

method: str, default = ‘max_dvdt’

• One of {‘max_dvdt’, ‘min_dvdt’, ‘abs_dvdt’},

• Method to locate stimulus, could be one of {‘max_dvdt’, ‘min_dvdt’, ‘abs_dvdt’}

• For MEA, max_dvdt - maximum gradient works well, which is a middle transition of stimulus from negative to positive.

• Check above illustration.

gradient_method: str,default=’fdiff’

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

• method to compute gradient of signal

• for MEA, stim location, ‘fdiff’ works well.

• For more details, spkit.signal_diff

plot: int, default=1

• if 0, no plot are shown,

• if 1, final locations of stimulus are plotted

• if 2, plot of each stimulus with signal are shown, N-plots are shown

verbose:bool, default=False,

• if False, no information is printed, default if False

figsize: default=(15,3)

• figure size for final figure (if plot>0)

other parameters:

(sg_window, sg_polyorder, gauss_window, gauss_itr)parameters for gradient

• default (sg_window=11,sg_polyorder=3,gauss_window=0,gauss_itr=1)

• these parameters are used while gradient computation, only

• if gradient method is one of (“sgdiff”,”sgdrift_diff”,”sgsmooth_diff”, “gauss_diff)

• For more details: spkit.signal_diff

Returns:

stim_loc: array of loc

• array of locations (as index) of all the stimuli

stim_mag: array of mag

• magnitude of deflection, at stim locatio. It can be used to determine the False Positive

See also

spkit.get_activation_time, spkit.signal_diff

Notes

Sometimes stimuli are on the edge of the cycle duration, try testing or cropping first quarter of the cycle.

Examples

#sp.mea.get_stim_loc
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)

##############################
# Step 1: Read File
fs = 25000
X,fs,ch_labels = sp.io.read_hdf(file_name,fs=fs,verbose=1)

x = X[0]
t = np.arange(len(x))/fs
t1,t2 = int(fs*1.98), int(fs*2.03)

plt.figure(figsize=(10,4))
plt.subplot(211)
plt.plot(t,x)
plt.xlim([t[0],t[-1]])
plt.xticks(np.arange(len(x)/fs))
plt.grid()
plt.title('Channel 0')
plt.subplot(223)
plt.plot(t[t1:t2],x[t1:t2])
plt.xlim([t[t1],t[t2]])
plt.grid()
plt.title('zoomed in around 2s')
plt.tight_layout()
plt.show()

##############################
# Step 2: Stim Localisation
stim_fhz = 1

stim_loc,_  = sp.mea.get_stim_loc(X,fs=fs,fhz=stim_fhz, plot=1,verbose=1,N=None)

Examples using spkit.mea.get_stim_loc

MEA: Step-wise Analysis: Example

MEA: Step-wise Analysis: Example