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
|
#!/usr/bin/env python3
# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
# vi: set ft=python sts=4 ts=4 sw=4 et:
__doc__ = """
Example of script to analyse the reproducibility in group studies using a
bootstrap procedure
Needs matplotlib
Author: Bertrand Thirion, 2005-2009
"""
print(__doc__)
import numpy as np
# Scipy stats needed for thresholding
import scipy.stats as st
try:
import matplotlib.pyplot as plt
except ImportError:
raise RuntimeError("This script needs the matplotlib library")
import nipy.labs.utils.simul_multisubject_fmri_dataset as simul
from nipy.labs.spatial_models.discrete_domain import grid_domain_from_binary_array
from nipy.labs.utils.reproducibility_measures import (
cluster_reproducibility,
map_reproducibility,
peak_reproducibility,
voxel_reproducibility,
)
###############################################################################
# Generate the data
n_subj = 105
shape = (60, 60)
pos = np.array([[12, 14],
[20, 20],
[30, 20]])
ampli = np.array([2.5, 3.5, 3])
betas = simul.surrogate_2d_dataset(n_subj=n_subj, shape=shape, pos=pos,
ampli=ampli, width=5.0)
n_vox = np.prod(shape)
# set the variance at 1 everywhere
func = np.reshape(betas, (n_subj, n_vox)).T
var = np.ones((n_vox, n_subj))
domain = grid_domain_from_binary_array(np.ones((shape[0], shape[1], 1)))
###############################################################################
# Run reproducibility analysis
ngroups = 10
thresholds = np.arange(.5, 6., .5)
sigma = 2.0
csize = 10
niter = 10
method = 'crfx'
verbose = 0
# do not use permutations
swap = False
kap = []
clt = []
pk = []
sens = []
for threshold in thresholds:
kwargs={'threshold': threshold, 'csize': csize}
kappa = []
cls = []
sent = []
peaks = []
for i in range(niter):
k = voxel_reproducibility(func, var, domain, ngroups, method, swap,
verbose, **kwargs)
kappa.append(k)
cld = cluster_reproducibility(func, var, domain, ngroups, sigma, method,
swap, verbose, **kwargs)
cls.append(cld)
peak = peak_reproducibility(func, var, domain, ngroups, sigma, method,
swap, verbose, **kwargs)
peaks.append(peak)
seni = map_reproducibility(func, var, domain, ngroups, method, True,
verbose, threshold=threshold,
csize=csize).mean()/ngroups
sent.append(seni)
sens.append(np.array(sent))
kap.append(np.array(kappa))
clt.append(np.array(cls))
pk.append(np.array(peaks))
###############################################################################
# Visualize the results
aux = st.norm.sf(thresholds)
a = plt.figure(figsize=(11, 6))
plt.subplot(1, 3, 1)
plt.boxplot(kap)
plt.title('voxel-level \n reproducibility', fontsize=12)
plt.xticks(range(1, 1 + len(thresholds)), thresholds, fontsize=9)
plt.xlabel('threshold')
plt.subplot(1, 3, 2)
plt.boxplot(clt)
plt.title('cluster-level \n reproducibility', fontsize=12)
plt.xticks(range(1, 1 + len(thresholds)), thresholds, fontsize=9)
plt.xlabel('threshold')
plt.subplot(1, 3, 3)
plt.boxplot(pk, notch=1)
plt.title('peak-level \n reproducibility', fontsize=12)
plt.xticks(range(1, 1 + len(thresholds)), thresholds, fontsize=9)
plt.xlabel('threshold')
plt.figure()
for q, threshold in enumerate(thresholds):
plt.subplot(3, len(thresholds) // 3 + 1, q + 1)
rmap = map_reproducibility(func, var, domain, ngroups, method, verbose,
threshold=threshold, csize=csize)
rmap = np.reshape(rmap, shape)
plt.imshow(rmap, interpolation=None, vmin=0, vmax=ngroups)
plt.title(f'threshold: {threshold:g}', fontsize=10)
plt.axis('off')
plt.suptitle('Map reproducibility for different thresholds')
plt.show()
|
