Comparing Blink Detection by Eyelink vs. Algorithm by Hershman et al.

Reference: R. Hershman, A. Henik, and N. Cohen, “A novel blink detection method based on pupillometry noise,” Behav. Res. Methods, vol. 50, no. 1, pp. 107–114, 2018.

The Hershman algorithm is supposed to extend the blinks deteted by Eyelink to account for the opening and closing of eyelids.

import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
import pupeyes as pe

sns.set_theme('notebook', style='white')

# file name
path = 'data/sub001.asc'

# event marker format, specified as a dictionary {name: data type}   
msg_format = {'marker':str, 'event':str, 'block':str, 'trial':int} # e.g., start retention A 4
delimiter = ' ' # delimiter for messages

# start and stop notations for each trial
start_msg = 'start fixation'
stop_msg = 'end feedback'

# If you have any constant columns that you want to add
add_cols = {'subject':'sub001', 'condition': 'low'}

# read data
raw = pe.EyelinkReader(path=path, 
                       start_msg=start_msg, 
                       stop_msg=stop_msg, 
                       msg_format=msg_format, 
                       delimiter=delimiter, 
                       add_cols=add_cols
                       )
trial_idx = 2
samples = raw.get_samples()
samples = samples[samples.trial==trial_idx].reset_index(drop=True).iloc[0:10000] # just get a small part of the data

# get eyelink blinks
blinks_eyelink = raw.get_blinks()
blinks_eyelink = blinks_eyelink[blinks_eyelink.trial==trial_idx].reset_index(drop=True)


# get hershman blinks
from pupeyes.external.based_noise_blinks_detection import based_noise_blinks_detection
blinks_hershman = based_noise_blinks_detection(samples.pp, sampling_freq=1000)
blinks_hershman = pd.DataFrame(blinks_hershman)

# plot blinks
fig, ax = plt.subplots()
t = range(len(samples)) # normalize time to start at 0
ax.plot(t, samples.pp, color='blue')

for i, row in blinks_hershman.iterrows():
    hershman_onset = row['blink_onset']
    hershman_offset = row['blink_offset']
    if i == 0:
        ax.axvspan(samples.trackertime[hershman_onset]-samples.trackertime[0], samples.trackertime[hershman_offset]-samples.trackertime[0], color='red', alpha=0.3, label='hershman')
    else:
        ax.axvspan(samples.trackertime[hershman_onset]-samples.trackertime[0], samples.trackertime[hershman_offset]-samples.trackertime[0], color='red', alpha=0.3)

for i, row in blinks_eyelink.iterrows():
    blink_onset = row['starttime']
    blink_offset = row['endtime']
    if i == 0:
        ax.axvspan(blink_onset-samples.trackertime[0], blink_offset-samples.trackertime[0], color='green', alpha=0.3, label='eyelink')
    else:
        ax.axvspan(blink_onset-samples.trackertime[0], blink_offset-samples.trackertime[0], color='green', alpha=0.3)

ax.legend()

<matplotlib.legend.Legend at 0x790d5ab404a0>