forked from wmvanvliet/viswordrec-baseline
-
Notifications
You must be signed in to change notification settings - Fork 0
/
plot_model_brain_comparison_mne.py
132 lines (116 loc) · 4.43 KB
/
plot_model_brain_comparison_mne.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
"""
Plot correlations between grand-average source level activity and model activity.
"""
import mne
import numpy as np
import pickle
import pandas as pd
from scipy.stats import zscore
import os
# Path to the OSF downloaded data
data_path = './data'
stimuli = pd.read_csv(f'{data_path}/stimuli.csv')
stimuli['type'] = stimuli['type'].astype('category')
## Load model layer activations
with open(f'{data_path}/model_layer_activity.pkl', 'rb') as f:
d = pickle.load(f)
layer_activity = zscore(d['mean_activity'], axis=1)
layer_names = d['layer_names']
layer_acts = pd.DataFrame(layer_activity.T, columns=layer_names)
# The brain will be plotted from two angles.
views = ['lateral', 'ventral']
map_surface = 'white'
##
# Read in the dipoles and morph their positions to the fsaverage brain.
# The resulting positions will be placed on the brain figure as "foci": little
# spheres.
foci1 = []
foci2 = []
foci3 = []
subjects = list(range(1, 16))
for subject in subjects:
dips_tal = mne.read_dipole(f'{data_path}/dipoles/sub-{subject:02d}/meg/sub-{subject:02d}_task-epasana_dipoles_talairach.bdip')
dip_selection = pd.read_csv(f'{data_path}/dipoles/sub-{subject:02d}/meg/sub-{subject:02d}_task-epasana_dipole_selection.tsv', sep='\t', index_col=0)
pos_tal = dips_tal.pos * 1000
if 'LeftOcci1' in dip_selection.index:
foci1.append(pos_tal[dip_selection.loc['LeftOcci1'].dipole])
if 'LeftOcciTemp2' in dip_selection.index:
foci2.append(pos_tal[dip_selection.loc['LeftOcciTemp2'].dipole])
if 'LeftTemp3' in dip_selection.index:
foci3.append(pos_tal[dip_selection.loc['LeftTemp3'].dipole])
foci1 = np.array(foci1)
foci2 = np.array(foci2)
foci3 = np.array(foci3)
os.makedirs('figures', exist_ok=True)
## Layers 1-5 vs dipole group 1 (Early visual response)
for layer_name in layer_names[:5]:
c = mne.read_source_estimate(f'{data_path}/brain_model_comparison/brain_model_comparison_mne_{layer_name}')
c = c.copy().crop(0.065, 0.115)
c.data = np.maximum(c.data, 0)
brain = c.plot(
'fsaverage',
subjects_dir=data_path,
hemi='lh',
background='white',
cortex='low_contrast',
surface='inflated',
initial_time = c.get_peak()[1],
clim=dict(kind='value', lims=(0.2, 0.55, 0.9)),
)
# Save images without dipoles
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_no_dipoles.png')
# Save images with dipoles
brain.add_foci(foci1, map_surface=map_surface, hemi='lh', scale_factor=0.2)
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_with_dipoles.png')
## Layers 6-7 vs dipole group 2 (Letter string response)
for layer_name in layer_names[5:7]:
c = mne.read_source_estimate(f'{data_path}/brain_model_comparison/brain_model_comparison_mne_{layer_name}')
c = c.copy().crop(0.14, 0.2)
c.data = np.maximum(c.data, 0)
brain = c.plot(
'fsaverage',
subjects_dir=data_path,
hemi='lh',
background='white',
cortex='low_contrast',
surface='inflated',
initial_time = c.get_peak()[1],
clim=dict(kind='value', lims=(0.1, 0.225, 0.35)),
)
# Save images without dipoles
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_no_dipoles.png')
# Save images with dipoles
brain.add_foci(foci2, map_surface=map_surface, hemi='lh', scale_factor=0.2)
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_with_dipoles.png')
## Layer 8 vs dipole group 3 (N400m response)
layer_name = layer_names[7]
c = mne.read_source_estimate(f'{data_path}/brain_model_comparison/brain_model_comparison_mne_{layer_name}')
c = c.copy().crop(0.3, 0.5)
c.data = np.maximum(c.data, 0)
brain = c.plot(
'fsaverage',
subjects_dir=data_path,
hemi='lh',
background='white',
cortex='low_contrast',
surface='inflated',
initial_time = c.get_peak()[1],
clim=dict(kind='value', lims=(0.05, 0.125, 0.20)),
)
# Save images without dipoles
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_no_dipoles.png')
# Save images with dipoles
brain.add_foci(foci3, map_surface=map_surface, hemi='lh', scale_factor=0.2)
for view in views:
brain.show_view(view)
brain.save_image(f'figures/{layer_name}_{view}_mne_with_dipoles.png')