3. Epoching#
We are typically not interested in the full continuous EEG recording, but only in the EEG activity that happens around certain events of interest. Such events could be the onset of a stimulus or the onset of a response.
The events are usually stored in the EEG data as “triggers” (also called “markers” or “annotations”). These are numerical codes at certain timepoints that indicate the onset of an event. We can use them to “cut” the continuous EEG recording into smaller segments (called epochs) that begin a few hundred milliseconds before the event (to have a neutral “baseline” period) and end a few hundred milliseconds after the event (to capture brain activity related to the event).
Fig. 3.1 Continuous, epoched, and averaged EEG. Source: Luck (2022)#
Goals
Extracting event information from the data
Segmenting the data into epochs
Applying baseline correction
Rejecting high-amplitude epochs
Visualizing the epoched data
3.1. Load Python modules#
As before, we’ll make extensive use of the MNE-Python package(Gramfort et al, 2013). We’ll also use a new package called pandas (more on that below).
# %pip install mne hu-neuro-pipeline pandas
import pandas as pd
from mne import (Epochs, events_from_annotations, merge_events,
set_bipolar_reference)
from mne.io import read_raw
from mne.preprocessing import ICA
from pipeline.datasets import get_erpcore
3.2. Recreate preprocessing#
We repeat the preprocessing steps from the previous chapter in condensed form (without any intermediate plots).
Show code cell source
# Download data
files_dict = get_erpcore('N170', participants='sub-004', path='data')
raw_file = files_dict['raw_files'][0]
log_file = files_dict['log_files'][0]
# Preprocessing
raw = read_raw(raw_file, preload=True)
raw = set_bipolar_reference(raw, anode='FP1', cathode='VEOG_lower',
ch_name='VEOG', drop_refs=False)
raw = set_bipolar_reference(raw, anode='HEOG_right', cathode='HEOG_left',
ch_name='HEOG', drop_refs=False)
raw = raw.set_channel_types({'VEOG': 'eog', 'HEOG': 'eog'})
raw = raw.drop_channels(['VEOG_lower', 'HEOG_right', 'HEOG_left'])
raw = raw.set_montage('biosemi64', match_case=False)
raw = raw.filter(l_freq=0.1, h_freq=30.0)
raw_copy = raw.copy().filter(l_freq=1.0, h_freq=None, verbose=False)
ica = ICA(n_components=15)
ica = ica.fit(raw_copy)
eog_indices, eog_scores = ica.find_bads_eog(raw, ch_name=['VEOG', 'HEOG'],
verbose=False)
ica.exclude = eog_indices
raw = ica.apply(raw)
raw = raw.set_eeg_reference('average')
Show code cell output
Reading /home/runner/work/intro-to-eeg/intro-to-eeg/ipynb/data/erpcore/N170/sub-004/eeg/sub-004_task-N170_eeg.fdt
Reading 0 ... 649215 = 0.000 ... 633.999 secs...
EEG channel type selected for re-referencing
Creating RawArray with float64 data, n_channels=1, n_times=649216
Range : 0 ... 649215 = 0.000 ... 633.999 secs
Ready.
Added the following bipolar channels:
VEOG
EEG channel type selected for re-referencing
Creating RawArray with float64 data, n_channels=1, n_times=649216
Range : 0 ... 649215 = 0.000 ... 633.999 secs
Ready.
Added the following bipolar channels:
HEOG
Filtering raw data in 1 contiguous segment
Setting up band-pass filter from 0.1 - 30 Hz
FIR filter parameters
---------------------
Designing a one-pass, zero-phase, non-causal bandpass filter:
- Windowed time-domain design (firwin) method
- Hamming window with 0.0194 passband ripple and 53 dB stopband attenuation
- Lower passband edge: 0.10
- Lower transition bandwidth: 0.10 Hz (-6 dB cutoff frequency: 0.05 Hz)
- Upper passband edge: 30.00 Hz
- Upper transition bandwidth: 7.50 Hz (-6 dB cutoff frequency: 33.75 Hz)
- Filter length: 33793 samples (33.001 s)
Fitting ICA to data using 30 channels (please be patient, this may take a while)
Selecting by number: 15 components
Fitting ICA took 9.5s.
Applying ICA to Raw instance
Transforming to ICA space (15 components)
Zeroing out 2 ICA components
Projecting back using 30 PCA components
EEG channel type selected for re-referencing
Applying average reference.
Applying a custom ('EEG',) reference.
[Parallel(n_jobs=1)]: Done 17 tasks | elapsed: 0.3s
This gives us the cleaned, continuous EEG data as a Raw object.
raw
General
| Measurement date | Unknown |
|---|---|
| Experimenter | Unknown |
| Participant | Unknown |
Channels
| Digitized points | 33 points |
|---|---|
| Good channels | 30 EEG, 2 EOG |
| Bad channels | None |
| EOG channels | VEOG, HEOG |
| ECG channels | Not available |
Data
| Sampling frequency | 1024.00 Hz |
|---|---|
| Highpass | 0.10 Hz |
| Lowpass | 30.00 Hz |
| Filenames | sub-004_task-N170_eeg.fdt |
| Duration | 00:10:34 (HH:MM:SS) |
3.3. Extract events#
We can extract the event codes that are stored within the raw data using the events_from_annotations() function.
events, event_id = events_from_annotations(raw)
Used Annotations descriptions: ['1', '10', '101', '102', '103', '104', '105', '106', '107', '108', '109', '11', '110', '111', '112', '113', '114', '115', '116', '117', '118', '119', '12', '120', '121', '122', '123', '124', '125', '126', '127', '128', '129', '13', '130', '131', '132', '133', '134', '135', '136', '137', '138', '139', '14', '140', '141', '142', '143', '144', '145', '146', '147', '148', '149', '15', '150', '151', '152', '153', '154', '155', '156', '157', '158', '159', '16', '160', '161', '162', '163', '164', '165', '166', '167', '168', '169', '17', '170', '171', '172', '173', '174', '175', '176', '177', '178', '179', '18', '180', '19', '2', '20', '201', '202', '21', '22', '23', '24', '25', '26', '27', '28', '29', '3', '30', '31', '32', '33', '34', '35', '36', '37', '38', '39', '4', '40', '41', '42', '43', '44', '45', '46', '47', '48', '49', '5', '50', '51', '52', '53', '54', '55', '56', '57', '58', '59', '6', '60', '61', '62', '63', '64', '65', '66', '67', '68', '69', '7', '70', '71', '72', '73', '74', '75', '76', '77', '78', '79', '8', '80', '9']
This functions returns two outputs:
eventsis a Numpy array with three columns: the sample number (timepoint) of each event, the duration of the event, and the event code (here indicating the specific face or car stimulus)event_id: a dictionary mapping the event codes to human-readable names
events
array([[ 4506, 0, 95],
[ 8822, 0, 115],
[ 9257, 0, 94],
...,
[641985, 0, 94],
[643049, 0, 49],
[643419, 0, 94]])
event_id
{'1': 1,
'10': 2,
'101': 3,
'102': 4,
'103': 5,
'104': 6,
'105': 7,
'106': 8,
'107': 9,
'108': 10,
'109': 11,
'11': 12,
'110': 13,
'111': 14,
'112': 15,
'113': 16,
'114': 17,
'115': 18,
'116': 19,
'117': 20,
'118': 21,
'119': 22,
'12': 23,
'120': 24,
'121': 25,
'122': 26,
'123': 27,
'124': 28,
'125': 29,
'126': 30,
'127': 31,
'128': 32,
'129': 33,
'13': 34,
'130': 35,
'131': 36,
'132': 37,
'133': 38,
'134': 39,
'135': 40,
'136': 41,
'137': 42,
'138': 43,
'139': 44,
'14': 45,
'140': 46,
'141': 47,
'142': 48,
'143': 49,
'144': 50,
'145': 51,
'146': 52,
'147': 53,
'148': 54,
'149': 55,
'15': 56,
'150': 57,
'151': 58,
'152': 59,
'153': 60,
'154': 61,
'155': 62,
'156': 63,
'157': 64,
'158': 65,
'159': 66,
'16': 67,
'160': 68,
'161': 69,
'162': 70,
'163': 71,
'164': 72,
'165': 73,
'166': 74,
'167': 75,
'168': 76,
'169': 77,
'17': 78,
'170': 79,
'171': 80,
'172': 81,
'173': 82,
'174': 83,
'175': 84,
'176': 85,
'177': 86,
'178': 87,
'179': 88,
'18': 89,
'180': 90,
'19': 91,
'2': 92,
'20': 93,
'201': 94,
'202': 95,
'21': 96,
'22': 97,
'23': 98,
'24': 99,
'25': 100,
'26': 101,
'27': 102,
'28': 103,
'29': 104,
'3': 105,
'30': 106,
'31': 107,
'32': 108,
'33': 109,
'34': 110,
'35': 111,
'36': 112,
'37': 113,
'38': 114,
'39': 115,
'4': 116,
'40': 117,
'41': 118,
'42': 119,
'43': 120,
'44': 121,
'45': 122,
'46': 123,
'47': 124,
'48': 125,
'49': 126,
'5': 127,
'50': 128,
'51': 129,
'52': 130,
'53': 131,
'54': 132,
'55': 133,
'56': 134,
'57': 135,
'58': 136,
'59': 137,
'6': 138,
'60': 139,
'61': 140,
'62': 141,
'63': 142,
'64': 143,
'65': 144,
'66': 145,
'67': 146,
'68': 147,
'69': 148,
'7': 149,
'70': 150,
'71': 151,
'72': 152,
'73': 153,
'74': 154,
'75': 155,
'76': 156,
'77': 157,
'78': 158,
'79': 159,
'8': 160,
'80': 161,
'9': 162}
An alternative way to get the events is to use an explicit events file that accompanies the raw data. This is useful if there are no events stored in the raw data or if they are erroneous (as is actually the case here for the N170 dataset!).
For instance, the EEG data from each participant in the ERP CORE datasets (e.g., sub-004_task-N170_eeg.set) is accompanied by a tab-separated events file (e.g., sub-004_task-N170_events.tsv):
%cat data/erpcore/N170/sub-004/eeg/sub-004_task-N170_events.tsv
onset duration sample trial_type stim_file value
4.4004 0.0 4507 response n/a 202
8.6152 0.3 8823 stimulus n/a 39
9.0400 0.0 9258 response n/a 201
10.1318 0.3 10376 stimulus n/a 178
10.4883 0.0 10741 response n/a 201
11.5977 0.3 11877 stimulus n/a 28
11.9600 0.0 12248 response n/a 201
13.1641 0.3 13481 stimulus n/a 167
13.5283 0.0 13854 response n/a 201
14.7295 0.3 15084 stimulus n/a 79
15.1680 0.0 15533 response n/a 201
16.2793 0.3 16671 stimulus n/a 149
16.7119 0.0 17114 response n/a 201
17.7451 0.3 18172 stimulus n/a 21
17.9922 0.0 18425 response n/a 201
19.2783 0.3 19742 stimulus n/a 130
19.6240 0.0 20096 response n/a 201
20.7939 0.3 21294 stimulus n/a 76
21.0635 0.0 21570 response n/a 201
22.2109 0.3 22745 stimulus n/a 169
22.6641 0.0 23209 response n/a 201
23.8096 0.3 24382 stimulus n/a 8
24.1836 0.0 24765 response n/a 201
25.2930 0.3 25901 stimulus n/a 36
25.6641 0.0 26281 response n/a 201
26.8418 0.3 27487 stimulus n/a 32
27.2637 0.0 27919 response n/a 201
28.2920 0.3 28972 stimulus n/a 33
28.6318 0.0 29320 response n/a 201
29.8906 0.3 30609 stimulus n/a 159
30.3359 0.0 31065 response n/a 201
31.3408 0.3 32094 stimulus n/a 73
31.7197 0.0 32482 response n/a 201
32.8730 0.3 33663 stimulus n/a 12
33.2480 0.0 34047 response n/a 201
34.3066 0.3 35131 stimulus n/a 14
34.6797 0.0 35513 response n/a 201
35.8223 0.3 36683 stimulus n/a 153
36.2158 0.0 37086 response n/a 201
37.2549 0.3 38150 stimulus n/a 146
37.5996 0.0 38503 response n/a 201
38.6543 0.3 39583 stimulus n/a 13
38.9834 0.0 39920 response n/a 201
40.2373 0.3 41204 stimulus n/a 110
40.5430 0.0 41517 response n/a 202
41.7539 0.3 42757 stimulus n/a 114
42.1992 0.0 43213 response n/a 201
43.2363 0.3 44275 stimulus n/a 23
43.6074 0.0 44655 response n/a 201
44.7695 0.3 45845 stimulus n/a 120
45.2549 0.0 46342 response n/a 201
46.2852 0.3 47397 stimulus n/a 155
46.6230 0.0 47743 response n/a 201
47.8848 0.3 49035 stimulus n/a 47
48.2236 0.0 49382 response n/a 201
49.2842 0.3 50468 stimulus n/a 72
49.7354 0.0 50930 response n/a 202
50.8838 0.3 52106 stimulus n/a 177
51.2393 0.0 52470 response n/a 201
52.3662 0.3 53624 stimulus n/a 49
52.7754 0.0 54043 response n/a 201
53.8662 0.3 55160 stimulus n/a 118
54.2314 0.0 55534 response n/a 201
55.2822 0.3 56610 stimulus n/a 136
55.6475 0.0 56984 response n/a 201
56.7148 0.3 58077 stimulus n/a 55
57.0713 0.0 58442 response n/a 201
58.1807 0.3 59578 stimulus n/a 61
58.5029 0.0 59908 response n/a 201
59.7305 0.3 61165 stimulus n/a 67
60.0547 0.0 61497 response n/a 201
61.2461 0.3 62717 stimulus n/a 30
61.5830 0.0 63062 response n/a 201
62.7793 0.3 64287 stimulus n/a 176
63.1592 0.0 64676 response n/a 201
64.2451 0.3 65788 stimulus n/a 150
64.6230 0.0 66175 response n/a 201
65.8281 0.3 67409 stimulus n/a 62
66.1748 0.0 67764 response n/a 201
67.2441 0.3 68859 stimulus n/a 119
67.6553 0.0 69280 response n/a 201
75.0586 0.3 76861 stimulus n/a 4
75.5430 0.0 77357 response n/a 201
76.5078 0.3 78345 stimulus n/a 70
76.8711 0.0 78717 response n/a 202
77.9902 0.3 79863 stimulus n/a 64
78.2949 0.0 80175 response n/a 201
79.5234 0.3 81433 stimulus n/a 38
79.9424 0.0 81862 response n/a 201
80.9395 0.3 82883 stimulus n/a 134
81.5029 0.0 83460 response n/a 201
82.3721 0.3 84350 stimulus n/a 29
82.7266 0.0 84713 response n/a 201
83.7715 0.3 85783 stimulus n/a 80
84.1826 0.0 86204 response n/a 201
85.3379 0.3 87387 stimulus n/a 173
85.7100 0.0 87768 response n/a 201
86.8379 0.3 88923 stimulus n/a 45
87.2627 0.0 89358 response n/a 201
88.3701 0.3 90492 stimulus n/a 175
88.7744 0.0 90906 response n/a 201
89.7695 0.3 91925 stimulus n/a 104
90.1982 0.0 92364 response n/a 201
91.2363 0.3 93427 stimulus n/a 56
91.6543 0.0 93855 response n/a 201
92.7188 0.3 94945 stimulus n/a 25
93.1660 0.0 95403 response n/a 201
94.2852 0.3 96549 stimulus n/a 113
94.6621 0.0 96935 response n/a 202
95.7842 0.3 98084 stimulus n/a 170
96.2227 0.0 98533 response n/a 201
97.3506 0.3 99688 stimulus n/a 57
97.7744 0.0 100122 response n/a 201
98.7666 0.3 101138 stimulus n/a 20
99.1426 0.0 101523 response n/a 201
100.2656 0.3 102673 stimulus n/a 6
100.5742 0.0 102989 response n/a 201
101.6992 0.3 104141 stimulus n/a 143
102.0459 0.0 104496 response n/a 201
103.1650 0.3 105642 stimulus n/a 5
103.4785 0.0 105963 response n/a 201
104.6309 0.3 107143 stimulus n/a 127
105.0381 0.0 107560 response n/a 201
106.1143 0.3 108662 stimulus n/a 125
106.4775 0.0 109034 response n/a 201
107.6465 0.3 110231 stimulus n/a 11
107.9736 0.0 110566 response n/a 201
109.2129 0.3 111835 stimulus n/a 135
109.5820 0.0 112213 response n/a 201
110.6953 0.3 113353 stimulus n/a 102
111.0781 0.0 113745 response n/a 201
112.1455 0.3 114838 stimulus n/a 58
112.4941 0.0 115195 response n/a 201
113.6611 0.3 116390 stimulus n/a 1
114.0137 0.0 116751 response n/a 201
115.1436 0.3 117908 stimulus n/a 144
115.5576 0.0 118332 response n/a 201
116.6436 0.3 119444 stimulus n/a 165
116.9902 0.0 119799 response n/a 201
118.2266 0.3 121065 stimulus n/a 158
118.5654 0.0 121412 response n/a 201
119.7090 0.3 122583 stimulus n/a 162
120.0381 0.0 122920 response n/a 201
121.1582 0.3 124067 stimulus n/a 10
121.5176 0.0 124435 response n/a 201
122.7080 0.3 125654 stimulus n/a 63
122.9492 0.0 125901 response n/a 201
124.1240 0.3 127104 stimulus n/a 78
124.4219 0.0 127409 response n/a 201
125.6738 0.3 128691 stimulus n/a 151
126.0293 0.0 129055 response n/a 201
127.1895 0.3 130243 stimulus n/a 172
127.5254 0.0 130587 response n/a 201
128.7559 0.3 131847 stimulus n/a 166
129.0693 0.0 132168 response n/a 201
130.2881 0.3 133416 stimulus n/a 145
130.5898 0.0 133725 response n/a 201
131.8379 0.3 135003 stimulus n/a 156
132.1094 0.0 135281 response n/a 201
133.3545 0.3 136556 stimulus n/a 109
133.6934 0.0 136903 response n/a 201
139.3857 0.3 142732 stimulus n/a 142
139.7969 0.0 143153 response n/a 201
140.8018 0.3 144182 stimulus n/a 53
141.1250 0.0 144513 response n/a 201
142.2510 0.3 145666 stimulus n/a 26
142.6133 0.0 146037 response n/a 201
143.7002 0.3 147150 stimulus n/a 3
143.9736 0.0 147430 response n/a 201
145.2334 0.3 148720 stimulus n/a 139
145.4375 0.0 148929 response n/a 202
146.7490 0.3 150272 stimulus n/a 65
147.0449 0.0 150575 response n/a 201
148.2656 0.3 151825 stimulus n/a 103
148.6211 0.0 152189 response n/a 202
149.7646 0.3 153360 stimulus n/a 17
150.0449 0.0 153647 response n/a 201
151.2979 0.3 154930 stimulus n/a 160
151.5410 0.0 155179 response n/a 201
152.8643 0.3 156534 stimulus n/a 179
153.1729 0.0 156850 response n/a 201
154.3799 0.3 158086 stimulus n/a 148
154.6934 0.0 158407 response n/a 201
155.9297 0.3 159673 stimulus n/a 154
156.2207 0.0 159971 response n/a 201
157.4453 0.3 161225 stimulus n/a 174
157.8125 0.0 161601 response n/a 201
159.0117 0.3 162829 stimulus n/a 24
159.3887 0.0 163215 response n/a 201
160.5947 0.3 164450 stimulus n/a 152
160.9570 0.0 164821 response n/a 201
162.0107 0.3 165900 stimulus n/a 161
162.3730 0.0 166271 response n/a 201
163.4766 0.3 167401 stimulus n/a 27
163.7969 0.0 167729 response n/a 201
164.9600 0.3 168920 stimulus n/a 141
165.3809 0.0 169351 response n/a 201
166.4092 0.3 170404 stimulus n/a 9
166.7002 0.0 170702 response n/a 201
167.8252 0.3 171854 stimulus n/a 69
168.1729 0.0 172210 response n/a 202
169.2412 0.3 173304 stimulus n/a 50
169.5723 0.0 173643 response n/a 201
170.6572 0.3 174754 stimulus n/a 2
171.1084 0.0 175216 response n/a 201
172.1572 0.3 176290 stimulus n/a 180
172.5322 0.0 176674 response n/a 201
173.7393 0.3 177910 stimulus n/a 133
174.0527 0.0 178231 response n/a 201
175.2725 0.3 179480 stimulus n/a 157
175.5967 0.0 179812 response n/a 201
176.8389 0.3 181084 stimulus n/a 138
177.1562 0.0 181409 response n/a 201
178.2549 0.3 182534 stimulus n/a 147
178.5889 0.0 182876 response n/a 201
179.8379 0.3 184155 stimulus n/a 42
180.1807 0.0 184506 response n/a 201
181.2373 0.3 185588 stimulus n/a 66
181.5400 0.0 185898 response n/a 201
182.6367 0.3 187021 stimulus n/a 37
182.9160 0.0 187307 response n/a 201
184.2197 0.3 188642 stimulus n/a 164
184.6045 0.0 189036 response n/a 201
185.8184 0.3 190279 stimulus n/a 163
186.1084 0.0 190576 response n/a 201
187.3018 0.3 191798 stimulus n/a 46
187.7158 0.0 192222 response n/a 201
188.7842 0.3 193316 stimulus n/a 18
189.2041 0.0 193746 response n/a 201
190.3174 0.3 194886 stimulus n/a 52
190.5801 0.0 195155 response n/a 201
191.8828 0.3 196489 stimulus n/a 60
192.1562 0.0 196769 response n/a 201
193.4160 0.3 198059 stimulus n/a 34
193.6924 0.0 198342 response n/a 201
194.8818 0.3 199560 stimulus n/a 16
195.1396 0.0 199824 response n/a 201
196.4482 0.3 201164 stimulus n/a 171
196.7637 0.0 201487 response n/a 201
197.9648 0.3 202717 stimulus n/a 71
198.3164 0.0 203077 response n/a 201
208.2607 0.3 213260 stimulus n/a 59
209.0518 0.0 214070 response n/a 202
209.6768 0.3 214710 stimulus n/a 15
210.1162 0.0 215160 response n/a 201
211.1768 0.3 216246 stimulus n/a 168
211.5400 0.0 216618 response n/a 201
212.5928 0.3 217696 stimulus n/a 77
213.0117 0.0 218125 response n/a 201
214.0586 0.3 219197 stimulus n/a 19
214.3721 0.0 219518 response n/a 201
215.6250 0.3 220801 stimulus n/a 122
215.8916 0.0 221074 response n/a 202
217.1738 0.3 222387 stimulus n/a 31
217.5557 0.0 222778 response n/a 201
218.6904 0.3 223940 stimulus n/a 35
219.0352 0.0 224293 response n/a 201
220.2568 0.3 225544 stimulus n/a 147
220.6836 0.0 225981 response n/a 201
221.7227 0.3 227045 stimulus n/a 108
222.0752 0.0 227406 response n/a 201
223.2383 0.3 228597 stimulus n/a 22
223.6191 0.0 228987 response n/a 201
224.6719 0.3 230065 stimulus n/a 68
225.0352 0.0 230437 response n/a 201
226.2539 0.3 231685 stimulus n/a 144
226.5557 0.0 231994 response n/a 201
227.6875 0.3 233153 stimulus n/a 40
227.9316 0.0 233403 response n/a 201
229.1035 0.3 234603 stimulus n/a 41
229.4277 0.0 234935 response n/a 202
230.6191 0.3 236155 stimulus n/a 124
230.8828 0.0 236425 response n/a 201
232.2188 0.3 237793 stimulus n/a 7
232.5068 0.0 238088 response n/a 201
233.7852 0.3 239397 stimulus n/a 112
234.1230 0.0 239743 response n/a 201
235.3174 0.3 240966 stimulus n/a 15
235.7236 0.0 241382 response n/a 201
236.8506 0.3 242536 stimulus n/a 22
237.0830 0.0 242774 response n/a 201
238.4170 0.3 244140 stimulus n/a 18
238.7549 0.0 244486 response n/a 201
239.8828 0.3 245641 stimulus n/a 128
240.2988 0.0 246067 response n/a 201
241.4824 0.3 247279 stimulus n/a 149
241.7148 0.0 247517 response n/a 202
242.9980 0.3 248831 stimulus n/a 137
243.4111 0.0 249254 response n/a 201
244.4814 0.3 250350 stimulus n/a 31
244.9395 0.0 250819 response n/a 201
246.0635 0.3 251970 stimulus n/a 38
246.4111 0.0 252326 response n/a 202
247.4805 0.3 253421 stimulus n/a 13
247.8105 0.0 253759 response n/a 201
248.9961 0.3 254973 stimulus n/a 74
249.2832 0.0 255267 response n/a 201
250.4961 0.3 256509 stimulus n/a 140
250.8105 0.0 256831 response n/a 201
251.9785 0.3 258027 stimulus n/a 132
252.3867 0.0 258445 response n/a 201
253.4443 0.3 259528 stimulus n/a 126
253.7627 0.0 259854 response n/a 201
254.9941 0.3 261115 stimulus n/a 17
255.2910 0.0 261419 response n/a 201
256.5771 0.3 262736 stimulus n/a 116
257.0352 0.0 263205 response n/a 201
257.9932 0.3 264186 stimulus n/a 172
258.2744 0.0 264474 response n/a 202
259.5088 0.3 265738 stimulus n/a 75
259.7705 0.0 266006 response n/a 201
261.0420 0.3 267308 stimulus n/a 107
261.4502 0.0 267726 response n/a 201
262.5576 0.3 268860 stimulus n/a 115
262.9150 0.0 269226 response n/a 201
264.1074 0.3 270447 stimulus n/a 16
264.4590 0.0 270807 response n/a 201
265.5068 0.3 271880 stimulus n/a 54
265.8350 0.0 272216 response n/a 201
266.9229 0.3 273330 stimulus n/a 40
267.3701 0.0 273788 response n/a 201
293.9639 0.3 301020 stimulus n/a 131
294.4580 0.0 301526 response n/a 201
295.3633 0.3 302453 stimulus n/a 101
295.6895 0.0 302787 response n/a 201
296.8291 0.3 303954 stimulus n/a 174
297.1299 0.0 304262 response n/a 201
298.2617 0.3 305421 stimulus n/a 5
298.5615 0.0 305728 response n/a 201
299.7617 0.3 306957 stimulus n/a 129
300.1699 0.0 307375 response n/a 201
301.3105 0.3 308543 stimulus n/a 51
301.6260 0.0 308866 response n/a 201
302.8936 0.3 310164 stimulus n/a 156
303.2334 0.0 310512 response n/a 201
304.3926 0.3 311699 stimulus n/a 121
304.7617 0.0 312077 response n/a 201
305.9258 0.3 313269 stimulus n/a 48
306.3057 0.0 313658 response n/a 201
307.4258 0.3 314805 stimulus n/a 111
307.8174 0.0 315206 response n/a 201
308.9746 0.3 316391 stimulus n/a 117
309.3301 0.0 316755 response n/a 201
310.4580 0.3 317910 stimulus n/a 106
310.7617 0.0 318221 response n/a 201
311.8740 0.3 319360 stimulus n/a 157
312.1855 0.0 319679 response n/a 201
313.3398 0.3 320861 stimulus n/a 44
313.6895 0.0 321219 response n/a 201
314.8223 0.3 322379 stimulus n/a 123
315.2979 0.0 322866 response n/a 201
316.4053 0.3 324000 stimulus n/a 105
316.7695 0.0 324373 response n/a 201
317.9053 0.3 325536 stimulus n/a 135
318.2012 0.0 325839 response n/a 201
319.4541 0.3 327122 stimulus n/a 43
319.8018 0.0 327478 response n/a 201
320.8867 0.3 328589 stimulus n/a 154
321.2969 0.0 329009 response n/a 201
322.4863 0.3 330227 stimulus n/a 176
322.8330 0.0 330582 response n/a 201
323.9697 0.3 331746 stimulus n/a 53
324.3213 0.0 332106 response n/a 201
325.4189 0.3 333230 stimulus n/a 39
325.7129 0.0 333531 response n/a 201
326.8848 0.3 334731 stimulus n/a 162
327.2412 0.0 335096 response n/a 201
328.3848 0.3 336267 stimulus n/a 3
328.6816 0.0 336571 response n/a 201
329.8340 0.3 337751 stimulus n/a 132
330.1855 0.0 338111 response n/a 202
331.2832 0.3 339235 stimulus n/a 71
331.5615 0.0 339520 response n/a 201
332.7666 0.3 340754 stimulus n/a 76
333.0410 0.0 341035 response n/a 201
334.1992 0.3 342221 stimulus n/a 118
334.6172 0.0 342649 response n/a 201
335.6982 0.3 343756 stimulus n/a 127
336.1211 0.0 344189 response n/a 201
337.1650 0.3 345258 stimulus n/a 28
337.4814 0.0 345582 response n/a 201
338.5977 0.3 346725 stimulus n/a 130
339.0488 0.0 347187 response n/a 201
340.0137 0.3 348175 stimulus n/a 112
340.3125 0.0 348481 response n/a 201
341.5303 0.3 349728 stimulus n/a 155
341.8652 0.0 350071 response n/a 201
342.9961 0.3 351229 stimulus n/a 165
343.2812 0.0 351521 response n/a 201
344.4121 0.3 352679 stimulus n/a 151
344.7207 0.0 352995 response n/a 201
345.8779 0.3 354180 stimulus n/a 27
346.3291 0.0 354642 response n/a 201
347.3613 0.3 355699 stimulus n/a 35
347.7529 0.0 356100 response n/a 201
348.8936 0.3 357268 stimulus n/a 113
349.2969 0.0 357681 response n/a 201
350.4600 0.3 358872 stimulus n/a 21
350.8408 0.0 359262 response n/a 202
351.8926 0.3 360339 stimulus n/a 14
352.3525 0.0 360810 response n/a 201
397.4766 0.3 407017 stimulus n/a 141
398.3438 0.0 407905 response n/a 201
398.8760 0.3 408450 stimulus n/a 36
399.2236 0.0 408806 response n/a 201
400.3252 0.3 409934 stimulus n/a 70
400.6553 0.0 410272 response n/a 202
401.7578 0.3 411401 stimulus n/a 56
402.1191 0.0 411771 response n/a 201
403.3076 0.3 412988 stimulus n/a 134
403.7354 0.0 413426 response n/a 201
404.8076 0.3 414524 stimulus n/a 164
405.1758 0.0 414901 response n/a 201
406.2236 0.3 415974 stimulus n/a 73
406.5518 0.0 416310 response n/a 201
407.7227 0.3 417509 stimulus n/a 74
407.9990 0.0 417792 response n/a 201
409.2725 0.3 419096 stimulus n/a 166
409.5752 0.0 419406 response n/a 202
410.8555 0.3 420717 stimulus n/a 161
411.2793 0.0 421151 response n/a 201
412.2881 0.3 422184 stimulus n/a 47
412.6797 0.0 422585 response n/a 202
413.8369 0.3 423770 stimulus n/a 133
414.1992 0.0 424141 response n/a 201
415.2539 0.3 425221 stimulus n/a 67
415.5996 0.0 425575 response n/a 201
416.7695 0.3 426773 stimulus n/a 78
417.0791 0.0 427090 response n/a 201
418.3193 0.3 428360 stimulus n/a 34
418.6230 0.0 428671 response n/a 201
419.8848 0.3 429963 stimulus n/a 59
420.2549 0.0 430342 response n/a 201
421.4844 0.3 431601 stimulus n/a 19
421.8076 0.0 431932 response n/a 201
423.0508 0.3 433205 stimulus n/a 7
423.3516 0.0 433513 response n/a 201
424.5830 0.3 434774 stimulus n/a 2
424.8955 0.0 435094 response n/a 201
426.0664 0.3 436293 stimulus n/a 10
426.3994 0.0 436634 response n/a 201
427.5322 0.3 437794 stimulus n/a 126
427.9434 0.0 438215 response n/a 201
429.0654 0.3 439364 stimulus n/a 6
429.5352 0.0 439845 response n/a 201
430.5645 0.3 440899 stimulus n/a 60
431.0312 0.0 441377 response n/a 201
432.1143 0.3 442486 stimulus n/a 138
432.5029 0.0 442884 response n/a 202
433.5303 0.3 443936 stimulus n/a 45
433.8789 0.0 444293 response n/a 201
435.0127 0.3 445454 stimulus n/a 115
435.3906 0.0 445841 response n/a 201
436.5459 0.3 447024 stimulus n/a 171
436.8867 0.0 447373 response n/a 202
438.0283 0.3 448542 stimulus n/a 55
438.3184 0.0 448839 response n/a 201
439.4619 0.3 450010 stimulus n/a 107
439.8467 0.0 450404 response n/a 201
440.9941 0.3 451579 stimulus n/a 72
441.3428 0.0 451936 response n/a 201
442.3936 0.3 453012 stimulus n/a 122
442.8232 0.0 453452 response n/a 201
443.8936 0.3 454548 stimulus n/a 25
444.2783 0.0 454942 response n/a 201
445.3926 0.3 456083 stimulus n/a 150
445.7744 0.0 456474 response n/a 201
446.9258 0.3 457653 stimulus n/a 131
447.3027 0.0 458039 response n/a 201
448.4746 0.3 459239 stimulus n/a 159
448.7988 0.0 459571 response n/a 201
450.0576 0.3 460860 stimulus n/a 140
450.4229 0.0 461234 response n/a 201
451.5742 0.3 462413 stimulus n/a 9
451.9590 0.0 462807 response n/a 201
453.1230 0.3 463999 stimulus n/a 57
453.3984 0.0 464281 response n/a 201
454.6895 0.3 465603 stimulus n/a 109
454.9902 0.0 465911 response n/a 202
456.2559 0.3 467207 stimulus n/a 77
456.5742 0.0 467533 response n/a 201
476.6816 0.3 488123 stimulus n/a 42
477.0859 0.0 488537 response n/a 201
478.0977 0.3 489573 stimulus n/a 62
478.4141 0.0 489897 response n/a 201
479.4971 0.3 491006 stimulus n/a 178
479.9658 0.0 491486 response n/a 201
480.8965 0.3 492439 stimulus n/a 48
481.2539 0.0 492805 response n/a 201
482.4629 0.3 494043 stimulus n/a 110
482.7422 0.0 494329 response n/a 202
483.9287 0.3 495544 stimulus n/a 153
484.3418 0.0 495967 response n/a 201
485.5117 0.3 497165 stimulus n/a 44
485.8936 0.0 497556 response n/a 201
487.0449 0.3 498735 stimulus n/a 101
487.4297 0.0 499129 response n/a 201
488.5938 0.3 500321 stimulus n/a 173
488.9658 0.0 500702 response n/a 201
490.1602 0.3 501925 stimulus n/a 180
490.4541 0.0 502226 response n/a 201
491.6426 0.3 503443 stimulus n/a 32
491.9658 0.0 503774 response n/a 201
493.1094 0.3 504945 stimulus n/a 152
493.4219 0.0 505265 response n/a 201
494.6914 0.3 506565 stimulus n/a 105
495.0381 0.0 506920 response n/a 201
496.1914 0.3 508101 stimulus n/a 20
496.5176 0.0 508435 response n/a 201
497.7402 0.3 509687 stimulus n/a 66
498.0371 0.0 509991 response n/a 201
499.3066 0.3 511291 stimulus n/a 1
499.6299 0.0 511622 response n/a 201
500.8066 0.3 512827 stimulus n/a 137
501.2051 0.0 513235 response n/a 201
502.3887 0.3 514447 stimulus n/a 168
502.7012 0.0 514767 response n/a 201
503.8223 0.3 515915 stimulus n/a 124
504.1973 0.0 516299 response n/a 201
505.3047 0.3 517433 stimulus n/a 30
505.7334 0.0 517872 response n/a 201
506.7539 0.3 518917 stimulus n/a 114
507.1572 0.0 519330 response n/a 201
508.2373 0.3 520436 stimulus n/a 24
508.6289 0.0 520837 response n/a 201
509.7529 0.3 521988 stimulus n/a 79
510.1172 0.0 522361 response n/a 202
511.2188 0.3 523489 stimulus n/a 136
511.6055 0.0 523885 response n/a 201
512.7520 0.3 525059 stimulus n/a 169
513.1016 0.0 525417 response n/a 201
514.2178 0.3 526560 stimulus n/a 41
514.5811 0.0 526932 response n/a 201
515.6514 0.3 528028 stimulus n/a 4
515.9336 0.0 528317 response n/a 201
517.1338 0.3 529546 stimulus n/a 158
517.5654 0.0 529988 response n/a 201
518.6494 0.3 531098 stimulus n/a 49
518.9326 0.0 531388 response n/a 201
520.1162 0.3 532600 stimulus n/a 163
520.5088 0.0 533002 response n/a 201
521.6152 0.3 534135 stimulus n/a 111
521.9971 0.0 534526 response n/a 201
523.0820 0.3 535637 stimulus n/a 142
523.3809 0.0 535943 response n/a 201
524.6309 0.3 537223 stimulus n/a 106
524.9170 0.0 537516 response n/a 201
526.1143 0.3 538742 stimulus n/a 68
526.3887 0.0 539023 response n/a 202
527.5967 0.3 540260 stimulus n/a 128
528.0371 0.0 540711 response n/a 201
529.1963 0.3 541898 stimulus n/a 103
529.5488 0.0 542259 response n/a 201
530.6455 0.3 543382 stimulus n/a 51
531.0605 0.0 543807 response n/a 201
532.0615 0.3 544832 stimulus n/a 29
532.4209 0.0 545200 response n/a 201
533.5449 0.3 546351 stimulus n/a 61
533.9170 0.0 546732 response n/a 201
535.0938 0.3 547937 stimulus n/a 175
535.4443 0.0 548296 response n/a 202
569.3320 0.3 582997 stimulus n/a 63
569.7959 0.0 583472 response n/a 201
570.7480 0.3 584447 stimulus n/a 52
571.0918 0.0 584799 response n/a 201
572.2979 0.3 586034 stimulus n/a 33
572.6523 0.0 586397 response n/a 201
573.7139 0.3 587484 stimulus n/a 75
574.0596 0.0 587838 response n/a 201
575.2295 0.3 589036 stimulus n/a 119
575.5801 0.0 589395 response n/a 201
576.7959 0.3 590640 stimulus n/a 23
577.1074 0.0 590959 response n/a 201
578.3115 0.3 592192 stimulus n/a 80
578.6436 0.0 592532 response n/a 201
579.8779 0.3 593796 stimulus n/a 65
580.1875 0.0 594113 response n/a 201
581.4443 0.3 595400 stimulus n/a 58
581.7480 0.0 595711 response n/a 201
582.8770 0.3 596867 stimulus n/a 179
583.2754 0.0 597275 response n/a 201
584.4434 0.3 598471 stimulus n/a 37
584.7881 0.0 598824 response n/a 201
585.9258 0.3 599989 stimulus n/a 146
586.2920 0.0 600364 response n/a 201
587.3750 0.3 601473 stimulus n/a 54
587.7080 0.0 601814 response n/a 201
588.8418 0.3 602975 stimulus n/a 12
589.1396 0.0 603280 response n/a 201
590.4238 0.3 604595 stimulus n/a 11
590.7080 0.0 604886 response n/a 201
591.8906 0.3 606097 stimulus n/a 117
592.2275 0.0 606442 response n/a 201
593.3730 0.3 607615 stimulus n/a 145
593.6680 0.0 607917 response n/a 202
594.8057 0.3 609082 stimulus n/a 160
595.1953 0.0 609481 response n/a 201
596.2891 0.3 610601 stimulus n/a 8
596.6436 0.0 610964 response n/a 201
597.7051 0.3 612051 stimulus n/a 139
598.0039 0.0 612357 response n/a 202
599.3047 0.3 613689 stimulus n/a 123
599.6270 0.0 614019 response n/a 201
600.7539 0.3 615173 stimulus n/a 69
601.1396 0.0 615568 response n/a 201
602.3037 0.3 616760 stimulus n/a 129
602.6992 0.0 617165 response n/a 201
603.8691 0.3 618363 stimulus n/a 116
604.1875 0.0 618689 response n/a 201
605.3857 0.3 619916 stimulus n/a 108
605.6592 0.0 620196 response n/a 201
606.9688 0.3 621537 stimulus n/a 121
607.2275 0.0 621802 response n/a 201
608.5010 0.3 623106 stimulus n/a 43
608.8350 0.0 623448 response n/a 201
609.9668 0.3 624607 stimulus n/a 170
610.3389 0.0 624988 response n/a 201
611.4502 0.3 626126 stimulus n/a 120
611.7549 0.0 626438 response n/a 201
612.9658 0.3 627678 stimulus n/a 102
613.2588 0.0 627978 response n/a 201
614.4492 0.3 629197 stimulus n/a 177
614.7627 0.0 629518 response n/a 201
616.0322 0.3 630818 stimulus n/a 46
616.3633 0.0 631157 response n/a 201
617.5312 0.3 632353 stimulus n/a 50
617.8271 0.0 632656 response n/a 201
618.9639 0.3 633820 stimulus n/a 148
619.3066 0.0 634171 response n/a 201
620.5303 0.3 635424 stimulus n/a 64
620.9072 0.0 635810 response n/a 201
622.0967 0.3 637028 stimulus n/a 125
622.3945 0.0 637333 response n/a 202
623.4961 0.3 638461 stimulus n/a 26
623.8672 0.0 638841 response n/a 201
625.0283 0.3 640030 stimulus n/a 167
625.3711 0.0 640381 response n/a 201
626.5781 0.3 641617 stimulus n/a 104
626.9385 0.0 641986 response n/a 201
627.9775 0.3 643050 stimulus n/a 143
628.3389 0.0 643420 response n/a 201
We can read this file using the pandas package, which offers a DataFrame type for working tabular data (similar to a data.frame or tibble in R).
log_file = files_dict['log_files'][0]
log = pd.read_csv(log_file, sep='\t')
log
| onset | duration | sample | trial_type | stim_file | value | |
|---|---|---|---|---|---|---|
| 0 | 4.4004 | 0.0 | 4507 | response | NaN | 202 |
| 1 | 8.6152 | 0.3 | 8823 | stimulus | NaN | 39 |
| 2 | 9.0400 | 0.0 | 9258 | response | NaN | 201 |
| 3 | 10.1318 | 0.3 | 10376 | stimulus | NaN | 178 |
| 4 | 10.4883 | 0.0 | 10741 | response | NaN | 201 |
| ... | ... | ... | ... | ... | ... | ... |
| 636 | 625.3711 | 0.0 | 640381 | response | NaN | 201 |
| 637 | 626.5781 | 0.3 | 641617 | stimulus | NaN | 104 |
| 638 | 626.9385 | 0.0 | 641986 | response | NaN | 201 |
| 639 | 627.9775 | 0.3 | 643050 | stimulus | NaN | 143 |
| 640 | 628.3389 | 0.0 | 643420 | response | NaN | 201 |
641 rows × 6 columns
From this data frame, we can re-create the 3-column events array (see above), now with the correct timings and event codes.
Note that by converting all columns from floats to integers, we are rounding the duration of the events from 0.3 (the actual length of the stimulus) to 0, as is common practice in ERP analysis.
events = log[['sample', 'duration', 'value']].values.astype(int)
events
array([[ 4507, 0, 202],
[ 8823, 0, 39],
[ 9258, 0, 201],
...,
[641986, 0, 201],
[643050, 0, 143],
[643420, 0, 201]])
However, these many different numerical event codes still have no obvious meaning to us.
We can check what they stand for by looking at the task-N170_events.json file that accompanies the dataset:
%cat data/erpcore/N170/task-N170_events.json
{
"value": {
"LongName": "Event code value",
"Levels": {
"1-40": "Stimulus - faces",
"41-80": "Stimulus - cars",
"101-140": "Stimulus - scrambled faces",
"141-180": "Stimulus - scrambled cars",
"201": "Response - correct",
"202": "Response - error"
}
}
}
We see that triggers 1–40 correspond to face stimuli, and triggers 41–80 correspond to car stimuli.
To make our epoching job easier, we’ll collapse these 80 original event codes into just two event codes: 1 for faces and 2 for cars, using MNE’s merge_events() function.
events = merge_events(events, ids=range(1, 41), new_id=1)
events = merge_events(events, ids=range(41, 81), new_id=2)
events
array([[ 4507, 0, 202],
[ 8823, 0, 1],
[ 9258, 0, 201],
...,
[641986, 0, 201],
[643050, 0, 143],
[643420, 0, 201]])
Note that range() is a built-in Python function that generates a sequence of numbers (integers) between a start and end value.
For example, range(1, 41) generates the sequence 1, 2, …, 40 (the end value is not included).
We also need to update the event_id dictionary accordingly, mapping the new event codes to human-readable condition labels.
event_id = {'face': 1, 'car': 2}
3.4. Epoching#
No we’re ready to segment our raw data into epochs, using the events and event_id that we just created.
We’ll use a time window of 1 s, starting 200 ms before the event onset.
For now, we’ll skip baseline correction (this would be enabled by default) because we want to show the effect of baseline correction later.
epochs = Epochs(raw, events, event_id, tmin=-0.2, tmax=0.8, baseline=None)
Not setting metadata
160 matching events found
No baseline correction applied
0 projection items activated
epochs.get_data().shape
Using data from preloaded Raw for 160 events and 1025 original time points ...
0 bad epochs dropped
(160, 32, 1025)
We see that the dimensions have changed, from EEG channels × timepoints (raw) to epochs × EEG channels × timepoints (epochs).
Just as the raw data, the epochs also have a plot() method to visualize the data.
_ = epochs.plot(events=True)
Using matplotlib as 2D backend.
Using data from preloaded Raw for 20 events and 1025 original time points ...
Another effective way of visualizing all epochs in the dataset is the so-called ERP image plot. This is a heatmap with time on the x-axis, epochs on the y-axis, and the EEG amplitude represented by color.
We’ll create this plot for a single EEG channel ('PO8') that is typically sensitive to the N170 face effect.
_ = epochs.plot_image(picks='PO8')
Using data from preloaded Raw for 160 events and 1025 original time points ...
Not setting metadata
160 matching events found
No baseline correction applied
0 projection items activated
This type of plot could also be used to check for interesting patterns in the data, e.g., by sorting the epochs on the y-axis by some variables of interest (e.g., stimulus condition or reaction time).
3.5. Baseline correction#
Some of the epochs have a large offset (vertical shift) compared to the other epochs, which can happen due to technical or physiological drifts at some channels.
We can remove these offsets by applying a baseline correction. This works by, separately for each channel, subtracting the mean of the baseline activity from each timepoint in the epoch. The baseline activity is typically defined as the 200 ms prior to the event onset.
epochs = epochs.apply_baseline((-0.2, 0.0))
_ = epochs.plot(events=True)
Applying baseline correction (mode: mean)
Using data from preloaded Raw for 20 events and 1025 original time points ...
3.6. Rejecting bad epochs#
Despite all our data cleaning efforts (filtering, ICA, referencing, baseline correction), some epochs will still contain large-amplitude artifacts, e.g., due to movements or technical glitches.
We can “reject” (delete) these epochs from the data by using the drop_bad() method (or, alternatively, the reject=... argument of the Epochs constructor).
This allows us to specify a peak-to-peak-threshold (in volts).
If, at any channel, the difference between the minimum and maximum amplitude in an epoch exceeds this threshold, the epoch will be rejected (deleted).
Depending on how clean the dataset is, this threshold will typically be between 50 and 200 µV. Note that the lower the threshold (that is, the more epochs we’re rejecting), the cleaner the remaining epochs will be, but the fewer epochs we’ll have left, potentially reducing our statistical power.
epochs = epochs.drop_bad({'eeg': 100e-6})
Using data from preloaded Raw for 160 events and 1025 original time points ...
Rejecting epoch based on EEG : ['F8']
Rejecting epoch based on EEG : ['F3', 'C5', 'P9', 'PO7', 'P10', 'PO8', 'O2']
Rejecting epoch based on EEG : ['P9', 'P10', 'PO8', 'O2']
Rejecting epoch based on EEG : ['F3', 'F4']
Rejecting epoch based on EEG : ['F8']
5 bad epochs dropped
We see that the ERP image now has a reduced number of epochs (rows), but the color patterns are much clearer (as some large-amplitude epochs have been removed):
_ = epochs.plot_image(picks='PO8')
Using data from preloaded Raw for 155 events and 1025 original time points ...
Not setting metadata
155 matching events found
No baseline correction applied
0 projection items activated
3.7. Exercises#
Repeat the preprocessing (first code cell) for a different ERP CORE experiment (valid experiment names are
'N170','MMN','N2pc','N400','P3', or'ERN'). Then check thetask-<experiment>_events.jsonfile and construct the correcteventsandevent_idvariables. Finally, create epochs from the data and plot the ERP image for a channel that is typically sensitive to the effect of interest (check Kappenman et al (2021), Table 1, for a suggestion).
%cat data/erpcore/.../task-..._events.json
# Your code goes here
...
3.8. Further reading#
Online chapter Segmentation into ERP epochs in Newman (2020)
Tutorial on epochs on the MNE-Python website
3.9. References#
- Gramfort et al., 2013
Gramfort, A., Luessi, M., Larson, E., Engemann, D. A., Strohmeier, D., Brodbeck, C., … Hämäläinen, M. (2013). MEG and EEG data analysis with MNE-Python. Frontiers in Neuroscience, 7, 267. doi:10.3389/fnins.2013.00267
- Kappenman et al., 2021
Kappenman, E. S., Farrens, J. L., Zhang, W., Stewart, A. X., & Luck, S. J. (2021). ERP CORE: An open resource for human event-related potential research. NeuroImage, 225, 117465. doi:10.1016/j.neuroimage.2020.117465
- Luck, 2022
Luck, S. J. (2022). A very brief introduction to EEG and ERPs. Applied Event-Related Potential Data Analysis. LibreTexts. https://doi.org/10.18115/D5QG92.
- Newman, 2020
Newman, A. J. (2020). Neural Data Science in Python. Dalhousie University. https://neuraldatascience.io.