Accepted Manuscript Neurosensory effects of transcranial alternating current stimulation Valerio Raco, Robert Bauer, Mark Olenik, Diandra Brkic, Alireza Gharabaghi PII:
S1935-861X(14)00280-0
DOI:
10.1016/j.brs.2014.08.005
Reference:
BRS 586
To appear in:
Brain Stimulation
Received Date: 26 April 2014 Revised Date:
16 July 2014
Accepted Date: 8 August 2014
Please cite this article as: Raco V, Bauer R, Olenik M, Brkic D, Gharabaghi A, Neurosensory effects of transcranial alternating current stimulation, Brain Stimulation (2014), doi: 10.1016/j.brs.2014.08.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title
2 3
Neurosensory effects of transcranial alternating current stimulation
4
Authors
5 6
Valerio Raco#, Robert Bauer#, Mark Olenik#, Diandra Brkic, Alireza Gharabaghi*
7
Institutions
8
Division of Functional and Restorative Neurosurgery & Division of Translational
9
Neurosurgery, Department of Neurosurgery, Eberhard Karls University Tuebingen,
10
Germany, and Neuroprosthetics Research Group, Werner Reichardt Centre for Integrative
11
Neuroscience, Eberhard Karls University Tuebingen, Germany
13
#
These authors contributed equally to this work
M AN U
14
SC
12
RI PT
1
15
*Correspondence
16
Prof. Dr. Alireza Gharabaghi, [email protected], Division of Functional
17
and Restorative Neurosurgery & Division of Translational Neurosurgery, Department of
18
Neurosurgery, Eberhard Karls University, Otfried-Mueller-Str.45, 72076 Tuebingen,
19
Germany, Tel: +49 7071 29 83550, Fax: +49 7071 29 25104.
AC C
EP
TE D
20
ACCEPTED MANUSCRIPT
Abstract
23
BACKGROUND:
24
Electrical brain stimulation can elicit neurosensory side effects that are unrelated to the
25
intended stimulation effects. This presents a challenge when designing studies with
26
blinded control conditions.
27
OBJECTIVE:
28
The aim of this research was to investigate the role of different transcranial alternating
29
current stimulation (tACS) parameters, i.e. intensity, frequency, and electrode montage, on
30
the probability, duration and intensity of elicited neurosensory side effects.
31
METHODS:
32
In a first study, we examined the influence of tACS on sensations of phosphenes,
33
dizziness, pressure, and skin sensation in fifteen healthy subjects, during eight seconds of
34
stimulation with different amplitudes (1500 A, 1000 A, 500 A, 250 A), frequencies (2
35
Hz, 4 Hz, 8 Hz, 16 Hz, 32 Hz, 64 Hz), and montages (F3/F4, F3/C4, F3/P4, P3/F4, P3/C4,
36
P3/P4). In a second study, ten healthy subjects were exposed to 60 seconds of tACS
37
(1000µA, 2Hz versus 16 Hz, F3/F4 versus P3/P4) and were asked to rate the intensity of
38
sensations every 12 seconds.
39
RESULTS:
40
The first study showed that all stimulation parameters had an influence on the probability
41
and intensity of sensations. Phosphenes were most likely and strongest for frontal
42
montages and higher frequencies. Dizziness was most likely and strongest for parietal
43
montages and at stimulation frequency of 4 Hz. Skin sensations and pressure was more
44
likely when stimulation was performed across central regions and at posterior montages,
45
respectively. The second study also revealed that the probability and the intensity of
46
sensations were neither modified during more extended periods of stimulation nor affected
47
by carry-over effects.
48
CONCLUSION:
49
We demonstrated that the strength and the likelihood of sensations elicited by tACS were
50
specifically modulated by the stimulation parameters. The present work may therefore be
51
instrumental in establishing effective blinding conditions for studies with tACS.
AC C
EP
TE D
M AN U
SC
RI PT
21 22
ACCEPTED MANUSCRIPT
Introduction
53
Electrical brain stimulation can elicit side effects that are unrelated to the intended
54
stimulation effects. For transcranial direct current stimulation (tDCS), the feasibility of
55
blinding the participants with regard to the experimental conditions has been successfully
56
introduced, particularly for stimulation amplitudes below 2 mA [1]. There is conclusive
57
evidence that sensations elicited by tDCS fade with stimulation time [2]. Study designs
58
based on a fade-in/stimulation/fade-out approach have thus been proven effective in
59
blinding subjects [3]. In many clinical trials, subjects are therefore not able to distinguish
60
between active and sham tDCS [4,5].
61
However, designing studies with blinded control conditions for transcranial alternating
62
current stimulation (tACS), still poses a challenge. This emerging technique modulates the
63
ongoing oscillatory brain activity, rendering it a valuable tool for investigating brain function
64
in healthy and diseased conditions [6-9]. For tACS, entrainment effects have been found to
65
be linked to ongoing oscillatory activity [10,11], and strongly depend on the current state of
66
the central nervous system [12,13]. Furthermore, tACS might affect switching between
67
states [14], rendering it particularly suitable for closed-loop stimulation protocols which
68
detect specific brain states and trigger short-term stimulation to specifically modulate
69
neurophysiological activity [15,16]. However, tACS may elicit sensations [6] that are
70
unrelated to the stimulation effects, thus possibly inducing unwanted side effects. For
71
closed-loop applications of tACS, effective blinding of short-term stimulation will therefore
72
be important. Consequently, this research direction cannot rely on the established fade-
73
in/stimulation/fade-out approach [3], but will require a better understanding of the
74
neurosensory effects induced by short-term tAC stimulation, e.g. for several seconds up to
75
one minute. The contribution of different stimulation parameters, such as stimulation
76
amplitude, frequency, montage and duration on the induction of neurosensory side effects
77
therefore requires further research.
78
During earlier studies, three primary sensations were shown to be elicited by noninvasive
79
electrical brain stimulation. These were phosphenes (i.e. flickering perception of a light)
80
during tACS [6], skin sensations (e.g. perception of itching or tingling) during transcranial
81
direct current stimulation (tDCS) [17], and sensations of dizziness during alternating
82
current galvanic vestibular stimulation (AC-GVS) [18]. Earlier studies indicated that the
83
strongest phosphenes were perceived when stimulation was applied in the low beta band
84
[16,19], and the sensitivity to phosphenes decreased as the distance between the
85
stimulation electrodes and the retina increased [20]. With regard to skin sensations,
AC C
EP
TE D
M AN U
SC
RI PT
52
ACCEPTED MANUSCRIPT
studies on the electrical stimulation of skin revealed a linear dependency between the
87
sensation ratings and stimulation intensity and frequency [21]. In other words, high
88
stimulation intensities/frequencies resulted in more perceptible skin sensations. When it
89
came to sensations of dizziness, Stephan et al. [18] showed a frequency-dependency in
90
an AC-GVS study, with an inverse relation between the strength of the sensations and the
91
stimulation frequency, indicating that the strongest sensations for stimulation frequencies
92
lie in the low frequency range (< 4 Hz).
93
The aim of this research project was to provide a deeper insight into the neurosensory side
94
effects of tACS. In a first study, the acute effects of different stimulation parameters were
95
explored to identify those settings resulting in the highest likelihood and intensity of
96
sensations. The second study was based on those settings with the highest probability for
97
side effects and explored the influence of stimulation duration.
SC
RI PT
86
M AN U
98 Methods
100
Subjects
101
We recruited eighteen healthy subjects (mean age 23.9 years, SD 2.25, eleven females)
102
for the first study, and ten healthy subjects (mean age 26.2 years, SD 2.28, five females)
103
for the second study. Subjects had no history of psychiatric or neurological conditions, and
104
participated after giving written informed consent. The study was approved by the local
105
ethics committee. For the first study, three subjects could not be included in the analysis,
106
as one subject reported a persistent headache, while two others were not following the
107
instructions adequately.
EP
TE D
99
AC C
108 109
Experimental Setup
110
Experiments
111
(NeuroConn, Ilmenau, Germany). We used rubber electrodes (size of reference electrode
112
5 x 7 cm² and size of stimulation electrode 4 x 4 cm²) inserted into sponges soaked with
113
tap water. Please note that tap water has been shown to smoothen the current density
114
distribution [10]. Whenever montages were switched, the wetness of the electrodes was
115
inspected and if necessary, additional tap water was added. The electrodes were attached
116
to the scalp by a non-conductive elastic band and a cap which is usually used for
117
electroencephalography. Stimulation and reference electrodes were positioned based on
were
conducted
using
a multi-channel transcranial AC
stimulator
ACCEPTED MANUSCRIPT
the international 10-20 EEG system, with stimulation electrodes always being placed on
119
the right hemisphere (F4,C4 or P4) and reference electrode on the left hemisphere (F3 or
120
P3). Impedance was checked at the beginning of each stimulation block and kept below 17
121
kΩ. Stimulation waveform was sinusoidal and without DC offset. Subjects were seated on
122
a comfortable chair with their eyes open in a dimly lit room, facing a dark wall or a
123
computer screen in the first and the second study, respectively. Subjects were kept
124
blinded towards the stimulation condition.
RI PT
118
125 Sensation Rating System
127
We had initially planned to instruct subjects to pay special attention to three specific
128
sensations, but as the additional sensation of pressure was reported by the first subject in
129
the first study, we decided to include it in the list of possible sensations for all subsequent
130
measurements. Subjects were therefore informed before the task that four sensations
131
were most likely: phosphenes, pressure, dizziness and skin sensations. In the first study,
132
subjects were instructed to pay attention to any feelings induced by the stimulation and to
133
provide a qualitative description of any sensation. In the second study, subjects were
134
instructed to concentrate on only one sensation which was indicated by a text presented
135
on a computer screen. Following their qualitative description, subjects rated the intensity of
136
the perceived sensation on a 6-point scale, with 0 and 6 indicating no sensations and
137
strong sensations, respectively. The probability of a sensation was later calculated by
138
comparing intensity ratings to zero, resulting in binary values.
EP
TE D
M AN U
SC
126
139
All reports of visual sensations, such as flashes, lights, flickering, foggy and blurred vision,
141
were grouped together as phosphenes. All reports of increased weight or pressure, either
142
focal at the electrodes sites or generalized across the whole head were grouped together
143
as pressure. All reports of changes in spatial perception, such as vertigo or being off-
144
balance as well as lightheadedness, were grouped together as dizziness. All reports of
145
tingling, itching or heat causing a desire to scratch or withdraw were grouped together as
146
skin sensations.
147 148 149
AC C
140
ACCEPTED MANUSCRIPT
First Study
151
In a three-factorial design, we researched the effect of amplitude, frequency and montage
152
on probability and intensity of sensations. We explored six different bipolar montages (F3-
153
F4, F3-C4, F3-P4, P3-F4, P3-C4, P3-P4). For each electrode montage, six stimulation
154
frequencies (2 Hz, 4 Hz, 8 Hz, 16 Hz, 32 Hz, and 64 Hz) at four different stimulation
155
intensities (1500 µA, 1000 µA, 500 µA, and 250 µA) were explored, resulting in 24 trials
156
per montage. At the start of each trial, stimulation was ramped up for three seconds;
157
subsequently stimulation was kept constant for four seconds and followed by one second
158
of ramping down. In the following interval without stimulation, subjects were asked to
159
describe and rate any sensation experienced. After four seconds, the next trial started.
160
The approximate duration of the whole experiment was 90 min for each subject.
SC
RI PT
150
161 Second Study
163
In a three-factorial design, we examined the effect of duration, frequency and montage on
164
probability and intensity of sensations. Stimulation intensity was kept constant at 1000 A
165
while we explored two different bipolar montages (F3-F4, P3-P4) and two stimulation
166
frequencies (2 Hz versus 16 Hz). Every possible combination was explored four times; on
167
each occasion, the subject was instructed to focus on one specific sensation (phosphenes,
168
pressure, dizziness or skin sensations), thus resulting in 16 trials. To avoid potential carry-
169
over effects, the inter-stimulation break was extended to 60 seconds in this study. Prior to
170
each trial, the subject was informed about the sensation of interest within this trial. At the
171
onset of each trial, stimulation was ramped up for one second; subsequently stimulation
172
was kept constant for 58 seconds, followed by one second of ramping down. At the onset
173
of the trial and every 12 seconds from then on, an auditory cue was given, and subjects
174
rated the intensity of the sensation of interest by pressing the respective number on a
175
keyboard (0 to 6). The approximate duration of the whole experiment was 50 min for each
176
subject.
AC C
EP
TE D
M AN U
162
177 178
Data analysis
179
The statistical analysis of the results was performed using R and Matlab. Given the factors
180
amplitude, montage and frequency for the first study and time, montage and frequency for
181
the second study, we performed a 3-way repeated measures ANOVA on the probability of
ACCEPTED MANUSCRIPT
182
a sensation being experienced. We included only cases in which a sensation was
183
perceived, and subsequently performed a 3-way repeated measures ANOVA on the
184
intensity of each sensation. For the second study, we also performed a 4-way repeated
185
measures ANOVA with the factors sensation, stimulation duration, frequency and montage
186
on the response time of ratings and the probability of a sensation.
RI PT
187 Results
189
First Study
190
Phosphenes were reported by all fifteen subjects. The probability of experiencing
191
phosphenes depended on the frequency (F (5, 2132) = 69.1, p
S1935-861X(14)00280-0
DOI:
10.1016/j.brs.2014.08.005
Reference:
BRS 586
To appear in:
Brain Stimulation
Received Date: 26 April 2014 Revised Date:
16 July 2014
Accepted Date: 8 August 2014
Please cite this article as: Raco V, Bauer R, Olenik M, Brkic D, Gharabaghi A, Neurosensory effects of transcranial alternating current stimulation, Brain Stimulation (2014), doi: 10.1016/j.brs.2014.08.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title
2 3
Neurosensory effects of transcranial alternating current stimulation
4
Authors
5 6
Valerio Raco#, Robert Bauer#, Mark Olenik#, Diandra Brkic, Alireza Gharabaghi*
7
Institutions
8
Division of Functional and Restorative Neurosurgery & Division of Translational
9
Neurosurgery, Department of Neurosurgery, Eberhard Karls University Tuebingen,
10
Germany, and Neuroprosthetics Research Group, Werner Reichardt Centre for Integrative
11
Neuroscience, Eberhard Karls University Tuebingen, Germany
13
#
These authors contributed equally to this work
M AN U
14
SC
12
RI PT
1
15
*Correspondence
16
Prof. Dr. Alireza Gharabaghi, [email protected], Division of Functional
17
and Restorative Neurosurgery & Division of Translational Neurosurgery, Department of
18
Neurosurgery, Eberhard Karls University, Otfried-Mueller-Str.45, 72076 Tuebingen,
19
Germany, Tel: +49 7071 29 83550, Fax: +49 7071 29 25104.
AC C
EP
TE D
20
ACCEPTED MANUSCRIPT
Abstract
23
BACKGROUND:
24
Electrical brain stimulation can elicit neurosensory side effects that are unrelated to the
25
intended stimulation effects. This presents a challenge when designing studies with
26
blinded control conditions.
27
OBJECTIVE:
28
The aim of this research was to investigate the role of different transcranial alternating
29
current stimulation (tACS) parameters, i.e. intensity, frequency, and electrode montage, on
30
the probability, duration and intensity of elicited neurosensory side effects.
31
METHODS:
32
In a first study, we examined the influence of tACS on sensations of phosphenes,
33
dizziness, pressure, and skin sensation in fifteen healthy subjects, during eight seconds of
34
stimulation with different amplitudes (1500 A, 1000 A, 500 A, 250 A), frequencies (2
35
Hz, 4 Hz, 8 Hz, 16 Hz, 32 Hz, 64 Hz), and montages (F3/F4, F3/C4, F3/P4, P3/F4, P3/C4,
36
P3/P4). In a second study, ten healthy subjects were exposed to 60 seconds of tACS
37
(1000µA, 2Hz versus 16 Hz, F3/F4 versus P3/P4) and were asked to rate the intensity of
38
sensations every 12 seconds.
39
RESULTS:
40
The first study showed that all stimulation parameters had an influence on the probability
41
and intensity of sensations. Phosphenes were most likely and strongest for frontal
42
montages and higher frequencies. Dizziness was most likely and strongest for parietal
43
montages and at stimulation frequency of 4 Hz. Skin sensations and pressure was more
44
likely when stimulation was performed across central regions and at posterior montages,
45
respectively. The second study also revealed that the probability and the intensity of
46
sensations were neither modified during more extended periods of stimulation nor affected
47
by carry-over effects.
48
CONCLUSION:
49
We demonstrated that the strength and the likelihood of sensations elicited by tACS were
50
specifically modulated by the stimulation parameters. The present work may therefore be
51
instrumental in establishing effective blinding conditions for studies with tACS.
AC C
EP
TE D
M AN U
SC
RI PT
21 22
ACCEPTED MANUSCRIPT
Introduction
53
Electrical brain stimulation can elicit side effects that are unrelated to the intended
54
stimulation effects. For transcranial direct current stimulation (tDCS), the feasibility of
55
blinding the participants with regard to the experimental conditions has been successfully
56
introduced, particularly for stimulation amplitudes below 2 mA [1]. There is conclusive
57
evidence that sensations elicited by tDCS fade with stimulation time [2]. Study designs
58
based on a fade-in/stimulation/fade-out approach have thus been proven effective in
59
blinding subjects [3]. In many clinical trials, subjects are therefore not able to distinguish
60
between active and sham tDCS [4,5].
61
However, designing studies with blinded control conditions for transcranial alternating
62
current stimulation (tACS), still poses a challenge. This emerging technique modulates the
63
ongoing oscillatory brain activity, rendering it a valuable tool for investigating brain function
64
in healthy and diseased conditions [6-9]. For tACS, entrainment effects have been found to
65
be linked to ongoing oscillatory activity [10,11], and strongly depend on the current state of
66
the central nervous system [12,13]. Furthermore, tACS might affect switching between
67
states [14], rendering it particularly suitable for closed-loop stimulation protocols which
68
detect specific brain states and trigger short-term stimulation to specifically modulate
69
neurophysiological activity [15,16]. However, tACS may elicit sensations [6] that are
70
unrelated to the stimulation effects, thus possibly inducing unwanted side effects. For
71
closed-loop applications of tACS, effective blinding of short-term stimulation will therefore
72
be important. Consequently, this research direction cannot rely on the established fade-
73
in/stimulation/fade-out approach [3], but will require a better understanding of the
74
neurosensory effects induced by short-term tAC stimulation, e.g. for several seconds up to
75
one minute. The contribution of different stimulation parameters, such as stimulation
76
amplitude, frequency, montage and duration on the induction of neurosensory side effects
77
therefore requires further research.
78
During earlier studies, three primary sensations were shown to be elicited by noninvasive
79
electrical brain stimulation. These were phosphenes (i.e. flickering perception of a light)
80
during tACS [6], skin sensations (e.g. perception of itching or tingling) during transcranial
81
direct current stimulation (tDCS) [17], and sensations of dizziness during alternating
82
current galvanic vestibular stimulation (AC-GVS) [18]. Earlier studies indicated that the
83
strongest phosphenes were perceived when stimulation was applied in the low beta band
84
[16,19], and the sensitivity to phosphenes decreased as the distance between the
85
stimulation electrodes and the retina increased [20]. With regard to skin sensations,
AC C
EP
TE D
M AN U
SC
RI PT
52
ACCEPTED MANUSCRIPT
studies on the electrical stimulation of skin revealed a linear dependency between the
87
sensation ratings and stimulation intensity and frequency [21]. In other words, high
88
stimulation intensities/frequencies resulted in more perceptible skin sensations. When it
89
came to sensations of dizziness, Stephan et al. [18] showed a frequency-dependency in
90
an AC-GVS study, with an inverse relation between the strength of the sensations and the
91
stimulation frequency, indicating that the strongest sensations for stimulation frequencies
92
lie in the low frequency range (< 4 Hz).
93
The aim of this research project was to provide a deeper insight into the neurosensory side
94
effects of tACS. In a first study, the acute effects of different stimulation parameters were
95
explored to identify those settings resulting in the highest likelihood and intensity of
96
sensations. The second study was based on those settings with the highest probability for
97
side effects and explored the influence of stimulation duration.
SC
RI PT
86
M AN U
98 Methods
100
Subjects
101
We recruited eighteen healthy subjects (mean age 23.9 years, SD 2.25, eleven females)
102
for the first study, and ten healthy subjects (mean age 26.2 years, SD 2.28, five females)
103
for the second study. Subjects had no history of psychiatric or neurological conditions, and
104
participated after giving written informed consent. The study was approved by the local
105
ethics committee. For the first study, three subjects could not be included in the analysis,
106
as one subject reported a persistent headache, while two others were not following the
107
instructions adequately.
EP
TE D
99
AC C
108 109
Experimental Setup
110
Experiments
111
(NeuroConn, Ilmenau, Germany). We used rubber electrodes (size of reference electrode
112
5 x 7 cm² and size of stimulation electrode 4 x 4 cm²) inserted into sponges soaked with
113
tap water. Please note that tap water has been shown to smoothen the current density
114
distribution [10]. Whenever montages were switched, the wetness of the electrodes was
115
inspected and if necessary, additional tap water was added. The electrodes were attached
116
to the scalp by a non-conductive elastic band and a cap which is usually used for
117
electroencephalography. Stimulation and reference electrodes were positioned based on
were
conducted
using
a multi-channel transcranial AC
stimulator
ACCEPTED MANUSCRIPT
the international 10-20 EEG system, with stimulation electrodes always being placed on
119
the right hemisphere (F4,C4 or P4) and reference electrode on the left hemisphere (F3 or
120
P3). Impedance was checked at the beginning of each stimulation block and kept below 17
121
kΩ. Stimulation waveform was sinusoidal and without DC offset. Subjects were seated on
122
a comfortable chair with their eyes open in a dimly lit room, facing a dark wall or a
123
computer screen in the first and the second study, respectively. Subjects were kept
124
blinded towards the stimulation condition.
RI PT
118
125 Sensation Rating System
127
We had initially planned to instruct subjects to pay special attention to three specific
128
sensations, but as the additional sensation of pressure was reported by the first subject in
129
the first study, we decided to include it in the list of possible sensations for all subsequent
130
measurements. Subjects were therefore informed before the task that four sensations
131
were most likely: phosphenes, pressure, dizziness and skin sensations. In the first study,
132
subjects were instructed to pay attention to any feelings induced by the stimulation and to
133
provide a qualitative description of any sensation. In the second study, subjects were
134
instructed to concentrate on only one sensation which was indicated by a text presented
135
on a computer screen. Following their qualitative description, subjects rated the intensity of
136
the perceived sensation on a 6-point scale, with 0 and 6 indicating no sensations and
137
strong sensations, respectively. The probability of a sensation was later calculated by
138
comparing intensity ratings to zero, resulting in binary values.
EP
TE D
M AN U
SC
126
139
All reports of visual sensations, such as flashes, lights, flickering, foggy and blurred vision,
141
were grouped together as phosphenes. All reports of increased weight or pressure, either
142
focal at the electrodes sites or generalized across the whole head were grouped together
143
as pressure. All reports of changes in spatial perception, such as vertigo or being off-
144
balance as well as lightheadedness, were grouped together as dizziness. All reports of
145
tingling, itching or heat causing a desire to scratch or withdraw were grouped together as
146
skin sensations.
147 148 149
AC C
140
ACCEPTED MANUSCRIPT
First Study
151
In a three-factorial design, we researched the effect of amplitude, frequency and montage
152
on probability and intensity of sensations. We explored six different bipolar montages (F3-
153
F4, F3-C4, F3-P4, P3-F4, P3-C4, P3-P4). For each electrode montage, six stimulation
154
frequencies (2 Hz, 4 Hz, 8 Hz, 16 Hz, 32 Hz, and 64 Hz) at four different stimulation
155
intensities (1500 µA, 1000 µA, 500 µA, and 250 µA) were explored, resulting in 24 trials
156
per montage. At the start of each trial, stimulation was ramped up for three seconds;
157
subsequently stimulation was kept constant for four seconds and followed by one second
158
of ramping down. In the following interval without stimulation, subjects were asked to
159
describe and rate any sensation experienced. After four seconds, the next trial started.
160
The approximate duration of the whole experiment was 90 min for each subject.
SC
RI PT
150
161 Second Study
163
In a three-factorial design, we examined the effect of duration, frequency and montage on
164
probability and intensity of sensations. Stimulation intensity was kept constant at 1000 A
165
while we explored two different bipolar montages (F3-F4, P3-P4) and two stimulation
166
frequencies (2 Hz versus 16 Hz). Every possible combination was explored four times; on
167
each occasion, the subject was instructed to focus on one specific sensation (phosphenes,
168
pressure, dizziness or skin sensations), thus resulting in 16 trials. To avoid potential carry-
169
over effects, the inter-stimulation break was extended to 60 seconds in this study. Prior to
170
each trial, the subject was informed about the sensation of interest within this trial. At the
171
onset of each trial, stimulation was ramped up for one second; subsequently stimulation
172
was kept constant for 58 seconds, followed by one second of ramping down. At the onset
173
of the trial and every 12 seconds from then on, an auditory cue was given, and subjects
174
rated the intensity of the sensation of interest by pressing the respective number on a
175
keyboard (0 to 6). The approximate duration of the whole experiment was 50 min for each
176
subject.
AC C
EP
TE D
M AN U
162
177 178
Data analysis
179
The statistical analysis of the results was performed using R and Matlab. Given the factors
180
amplitude, montage and frequency for the first study and time, montage and frequency for
181
the second study, we performed a 3-way repeated measures ANOVA on the probability of
ACCEPTED MANUSCRIPT
182
a sensation being experienced. We included only cases in which a sensation was
183
perceived, and subsequently performed a 3-way repeated measures ANOVA on the
184
intensity of each sensation. For the second study, we also performed a 4-way repeated
185
measures ANOVA with the factors sensation, stimulation duration, frequency and montage
186
on the response time of ratings and the probability of a sensation.
RI PT
187 Results
189
First Study
190
Phosphenes were reported by all fifteen subjects. The probability of experiencing
191
phosphenes depended on the frequency (F (5, 2132) = 69.1, p
