Journal of Neurolinguistics 38 (2016) 1e13

Contents lists available at ScienceDirect

Journal of Neurolinguistics
journal homepage: www.elsevier.com/locate/jneuroling

Research paper

Transcranial direct current stimulation to improve naming

abilities of persons with chronic aphasia: A preliminary study
using individualized based protocol
Adi Lifshitz Ben Basat a, *, Aviah Gvion b, c, Jean-Jacques Vatine d, e,
Nira Mashal f, g
a

Department of Communication Disorders, School of Health Sciences, Ariel University, Israel

Department of Communication Disorders, Reuth Rehabilitation Hospital, Tel Aviv, Israel
Department of Communication Sciences and Disorders, Ono Academic College, Israel
d
Sackler Faculty of Medicine, Tel Aviv University, Israel
e
Research and Development Institute, Reuth Rehabilitation Hospital, Tel Aviv, Israel
f
School of Education, Faculty of Social Sciences, Bar Ilan University, Israel
g
Gonda Multidisciplinary Brain Research Center, Bar Ilan University, Israel
b
c

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 10 March 2015
Received in revised form 20 September 2015
Accepted 30 September 2015
Available online xxx

Background and purpose: Transcranial direct current stimulation (tDCS) is a noninvasive

tool to facilitate brain plasticity and enhance language recovery after stroke. Our study
aims to develop an efcient protocol for individualizing tDCS to treat naming decits
within chronic aphasia patients.
Method: Seven patients with chronic aphasia participated in this preliminary study. All
participants performed a baseline naming assessment. Next, the best stimulation area
(either Broca or Wernicke), best side (either left or right hemisphere), and best type of
stimulation (either anodal or cathodal) were assessed with tDCS during four individualized
pre-intervention sessions. The location and type of stimulation that produced the greatest
improvement for each patient were used in subsequent treatments. Treatment included
six stimulation sessions (2 mA, 10 min), three treatments per week, two weeks in a raw.
Naming abilities were assessed immediately after treatment, as well as one month and
three months after treatment. Sham (placebo like) tDCS was administered to all participants three months post treatment.
Results: Treatment led to signicant improvement in percentage of correct responses
compared to baseline, whereas sham led to no equivalent improvement. Improvement was
still present three months after treatment.
Conclusions: An individually-tailored protocol of 2 mA, 10 min tDCS was found to improve
naming abilities of individuals with chronic aphasia. If proved to be effective in larger
studies, our ndings may have important clinical implications for the use of tDCS in
enhancing language abilities after stroke.
2015 Elsevier Ltd. All rights reserved.

Keywords:
tDCS
Brain stimulation
Aphasia
Anomia
Speech therapy

* Corresponding author. Department of Communication Disorders, School of Health Sciences, Ariel University, Ariel 40700, Israel.
E-mail address: adilb@ariel.ac.il (A. Lifshitz Ben Basat).
http://dx.doi.org/10.1016/j.jneuroling.2015.09.004
0911-6044/ 2015 Elsevier Ltd. All rights reserved.

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

1. Introduction
A relatively new area of research is associated with the clinical implications of neuromodulation of the cortex by transcranial direct current stimulation (tDCS) to facilitate brain recovery after stroke. tDCS is a noninvasive, painless cortical
neuromodulation technique that modies spontaneous neuronal excitability through a tonic depolarization or hyperpolarization of resting membrane potential (Nitsche & Paulus, 2000). Depending on the polarity of the current ow, brain
excitability can either be increased (anodal tDCS) or decreased (cathodal tDCS). The benecial implications of tDCS within
stroke patients were found in the motor domain as well as in the cognitive and language aspects. In the motor cortex, anodal
tDCS over the affected hemisphere and cathodal tDCS over the unaffected hemisphere have been shown to improve motor
function after stroke (Boggio et al., 2007; Fregni et al., 2005). In the cognitive domain, anodal tDCS over the left dorsolateral
prefrontal cortex has been shown to enhance working memory (Jo et al., 2009) and to improve attention after stroke (Kang,
Baek, Kim, & Paik, 2009). In the language domain, several studies attempted to examine the use of tDCS to enhance language
skills. In healthy subjects using tDCS was found to enhance proper naming retrieval (Ross, McCoy, Wolk, Coslett, & Olson,
2010), matching picture to invented trained words (Fiori et al., 2011; Liuzzi et al., 2010), to increase the number of words
generated by healthy persons during a verbal uency task (Iyer et al., 2005), and to improve both accuracy and response time
on a picture naming task (Fertonani, Rosini, Cotelli, Rossini, & Miniussi, 2010; Sparing, Dafotakis, Meister,
Thirugnanasambandam, & Fink, 2008).
Within stroke aphasic patients, one of the frequent language decits is Anomia-the crucial impairment to name pictures or
objects. It signicantly reduces a person's ability to communicate and deliver a message and therefore causes frustration to
both patient and caregivers. In the last decade, several studies examined the use of tDCS with or without concurrent speech
therapy, in order to improve the impaired naming abilities of aphasic patients after stroke. For instance, Monti et al. (2008)
administered tDCS (2 mA, 10 min) over the left inferior frontal gyrus (IFG) to eight chronic non-uent aphasic patients.
Patients were assigned to either an anodal-tDCS or a cathodal-tDCS group. Active (AtDCS or CtDCS) and sham tDCS were
tested in random order, allowing at least 1 week to elapse between sessions. The subjects and the examiner were blinded to
the type of stimulation. Results suggested that cathodal-tDCS signicantly improved performance accuracy on a picture
naming task (33.6% improvement relative to baseline), whereas anodal stimulation had no such effect. Monti et al. (2008)
argued that the improvement in naming abilities was due to tDCS-induced depression of cortical inhibitory inter-neurons
that led to disinhibition and consequently to improved function of the damaged language areas. Baker, Rorden, and
Fridriksson (2010) reported that it was anodal-tDCS rather than cathodal-tDCS that improved naming abilities in ten
chronic aphasia patients. The authors administered anodal-tDCS (1 mA, 20 min) for ve days and sham tDCS for another ve
days, with concurrent computerized naming treatment. tDCS positioning was guided using a priori functional MRI results
during an overt naming task. Naming accuracy of treated items improved following anodal-tDCS as compared to sham-tDCS
(36 vs. 15 correctly named items, respectively), and lasted one week post treatment. Recently, Vestito, Rosellini, Mantero, and
Bandini (2014) found long lasting effects after anodal tDCS administered. The participants all who had post stroke chronic
aphasia received 10 days of anodal tDCS on the left frontal (perilesional) region. During each tDCS treatment the patients had
to complete a picture naming task. The results showed signicant improvement in naming ability compared to baseline up to
5 months after the end of treatment. Moreover, Fiori et al. (2011, 2013) found improvement in naming abilities following
anodal tDCS that was applied to posterior cortical regions such as Wernicke's area by the superior temporal gyrus (STG) (Fiori
et al., 2011; 2013). It appears that the few studies that looked at tDCS in aphasia have examined the effects of treatment after
differing periods of times, have stimulated various brain regions in the left hemisphere, and have used different stimulation
type (cathodal vs. anodal tDCS). Some of this inconsistency might be resolved if treatment takes into account the different
possibilities for language reorganization within the damaged brain, as well as the connectivity between the two cerebral
hemispheres during language processing.
In recent years a growing body of evidence suggests that not only is the left hemisphere involved in language recovery, but
that the right is involved as well. This idea is not new (Barlow, 1877; Basso, Gardelli, Grassi, & Mariotti, 1989; Gowers & Barker,
1886; Kinsbourne, 1971). Imaging studies have also suggested that recovery involves the right hemisphere as well (Blasi et al.,
2002; Buckner, Corbetta, Schatz, Raichle, & Petersen, 1996; Cappa et al., 1997; Musso et al., 1999; Thulborn, Carpenter, & Just,
1999; Weiller et al., 1995). This is further supported by studies that examined right hemisphere activation post therapy
(Crosson et al., 2005; Peck et al., 2004; Raboyeau et al., 2008). A recent study examined whether applying tDCS over the
el et al., 2011). The results showed that
homolog right IFG rather than the left IFG could enhance naming abilities as well (Flo
application of anodal tDCS over the right IFG signicantly improved naming accuracy, and the effect lasted two weeks post
treatment. However, the mechanisms underlying right hemisphere involvement in language recovery are not yet fully understood. Heiss and Thiel (2006) describe paths of brain plasticity during aphasia recovery after stroke using hierarchical
model. According to this model, best recovery is obtained when the original activation within the left hemisphere is restored.
However, such activation will be found only when brain damage is relatively circumscribed. Incomplete, but often satisfactory
improvement of language function can occur when primary functional centers in the left dominant hemisphere are damaged,
but there is still preserved activation of areas surrounding the lesion. At the bottom of this hierarchy model lays only limited
recovery that can be achieved when ipsilateral network components are severely damaged. Then, an activation of contralateral homologous regions occurs, enhancing partial interhemispheric compensational language processing in these areas.
However, this involvement of the contralesional areas contributes to some improvement in language functioning but usually
it is not as efcient as intrahemispheric compensation (Heiss & Thiel, 2006). Consistent with the hierarchical model Schlaug,

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Marchina, and Wan (2011) describe two paths of language recovery after stroke. First, recovery is assumed to follow
recruitment of perilesional areas in the impaired left hemisphere, with variable involvement of the contralateral homologous
structures of the right hemisphere. This path is more likely to characterize cases of relatively small lesions in the left
hemisphere. The other path occurs when there are larger lesions of language areas within the fronto-temporal lobes. In this
case, recovery is more likely to involve recruitment of homologous language and speech-motor areas in the right hemisphere
rather than recruitment of left perilesional regions.
Additional studies conrm the involvement of both hemispheres focusing on either phases of recovery or premorbid
language lateralization. For instance, Saur et al. (2006) studied the dynamics of language processing reorganization by
repeated functional MRI (fMRI) examinations with parallel language testing from the acute to the chronic stage of 14 patients
with aphasia caused by left hemisphere stroke. An age-matched control group was similarly tested. Results showed minor
early activation of non-infarcted left hemispheric language structures in the acute phase, while the subacute phase showed a
signicant increase of activation in the bilateral language network with peak activation in the right Broca-homolog area. This
bilateral increased activation also showed the strongest correlation with improved language function. In the chronic phase a
normalization of activation with a re-shift of peak activation to left-hemispheric language areas was observed, associated
with further language improvement.
Saur et al. (2006) suggest that brain reorganization during language recovery proceeds in three phases: At rst, a
signicantly reduced activation of remaining left language areas in the acute phase takes place. Second, there is a recruitment
of homolog language brain areas, which correlates with language improvement. Finally, a normalization of language activation in the left hemisphere occurs. The authors argue that this normalization may reect an extent of consolidation in the
language system. Saur and Hartwigsen (2012) suggest that homolog right hemisphere involvement in the chronic phase after
stroke depends on the amount of individual premorbid language lateralization. Thus, patients with a more bilateral premorbid language representation use more homolog right areas. In addition, the recruitment of the right hemisphere depends
also on the lateralization of the language function of interest (i.e., bilaterally organized functions like language comprehension might cause more involvement of right hemisphere than left-lateralized functions, like language production or
syntax). Thus, there is strong evidence of the involvement of both hemispheres in language recovery depending on the extent
and site of lesion, phases of recovery (with more benecial involvement of the right homologous language regions in the prechronic stages) and individual's language lateralization.
Most of the tDCS studies in aphasia set the left hemisphere as their target for stimulation (Baker et al., 2010; Fiori et al.,
2011; Monti et al., 2008). However given the role of the right hemisphere in language recovery it is crucial to consider
stimulation of the right hemisphere as well. Nevertheless, additional factors related to the dynamics that exist between both
hemispheres should be considered. Schlaug et al. (2011) suggest that the right hemisphere assumes speech motor output
functions through the homolog arcuate fasciculus which connects the homologous right language areas, i.e., connects the
right IFG and the right STG regions. However, the increased right hemisphere activity following stroke is not always benecial
(Lefaucheur, 2006; Martin et al., 2005; Naeser et al., 2005; Rosen et al. 2000; Saur et al. 2006). In some cases involvement of
the right hemisphere in language production after stroke reects increased transcallosal inhibition from right hemisphere to
the damaged left hemisphere. This interhemispheric inhibition may exacerbate language difculties and impede recovery
from aphasia (Hamilton, Chrysikou, & Coslett, 2011; Martin et al., 2009; Naeser et al., 2005). Thus, there is great variability in
the language reorganization after brain damage. The right hemisphere involvement in language recovery may be either
through inhibitory mechanism (Hamilton et al., 2011) or through compensating mechanism (Hamilton et al., 2011; Schlaug
et al., 2011) that may be depended on the site, size and characteristic of the individual brain lesion (Crinion, Holland, Copland,
Thompson, & Hillis, 2013; Heiss & Thiel, 2006; Martin et al., 2009; Parkinson, Raymer, Chang, FitzGerald, & Crosson., 2009).
Given the various paths to recovery from aphasia, it is plausible to assume that some patients will benet more from anodal
tDCS over left IFG, whereas others will benet more from cathodal-tDCS over the right IFG. The former treatment will
enhance excitation of perilesional areas, whereas the latter treatment will decrease right hemisphere inhibition.
When considering language recovery after stroke, other brain areas and tracts should be taken into account. In a recent
review, Catani and Mesulam (2008) showed that the anatomy of the arcuate fasciculus is more complex than previously
assumed (Catani, Jones, & Fytche, 2005). The arcuate fasciculus includes a direct path connecting Wernicke's area with Broca's
area. A lesion to the arcuate fasciculus causes the disconnection syndrome (Geschwind, 1965, 1970; Geschwind & Levitsky,
1968). The arcuate fasciculus also includes an indirect pathway consisting of two segments, an anterior segment linking
Broca's territory with the inferior parietal lobule, and a posterior segment linking the inferior parietal lobule with Wernicke's
area. The authors argue that this arrangement supports a more exible architecture of parallel processing (Mesulam, 1990).
Thus, it is important to consider several brain areas as potential target sites of stimulation. In the current study we tested the
left IFG and the left STG, as well as their right homolog regions.
Given the complexity of brain reorganization after stroke, recent studies have examined the most appropriate, individualized site for stimulation, which would then be targeted to help optimize response to neurostimulation (e.g. Datta, Baker,
Bikson, & Fridriksson, 2011; Garcia, Norise, Faseyitan, Naeser, & Hamilton, 2013; Naeser et al. 2012; Shah-Basak et al., 2015).
Naeser et al. (2012) administered naming tests before and after 10 min of 1 HZ rTMS sessions in four sub-regions within right
Broca's homolog region. Similarly, Garcia et al. (2013) tested naming performances before and after six rTMS stimulations
delivered to six different sites in the right inferior frontal lobe. In both studies the best response location was dened as the
site associated with a naming score that is at least 2 SD above baseline. However, these pioneer studies, although taking into
account individual differences in locating the stimulation target, tested only the suppression of right hemisphere areas and

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

did not use electric stimulation as tDCS. Thus, these studies point to the potential of individual tailored tDCS treatment to
enhance language recovery after stroke.
Additional variable to consider is that there is no obvious way to assess quality of brain activation in language areas (e.g.,
Broca and Wernicke's area). Different individuals can benet from different type of stimulation (either anodal or cathodal
tDCS) over different targeted regions within the two cerebral hemispheres depending among other factors on the individual's
language lateralization and on inter-individual variability of anatomical location (Andoh & Martinot, 2008; Fama &
Turkeltaub, 2014). For instance, it has been found that 1 Hz-rTMS over the left hemisphere slows down language processing in subjects with left-side language dominance, but not in subjects with right-side language dominance (Knecht et al.
(2002), in Andoh & Martinot, 2008). The authors argue that these ndings demonstrate the susceptibility of the neuromodulation technique (TMS in that case) to the degree of language lateralization. Furthermore, the anatomical location of the
targeted areas varies across individuals. For instance, it has been shown that the inter-individual variability of the peak voxel
in Wernicke's area varies substantially between participants (Andoh et al., 2006).
Given that naming is a task that is comprised of multiple representations another variable that might have an effect on
naming recovery post stroke, is the functional loci of the naming impairment. Some researches refer to selective impairment
of naming as a result of damage to selective regions (left temporal pole, left anterior temporal, left posterior temporal regions
out of the classical language areas) leading to category specic naming decits (e.g., persons, animals, tools) (Damasio,
Grabowski, Tranel, Hichwa, & Damasio, 1996; Damasio, Tranel, Grabowski, Adolphs, & Damasio, 2004). Current psycholinguistic models of speech production conceive word production as a multi-staged process, with distinct components
(Butterworth, 1989; Caramazza & Hillis, 1990; Dell, 1986; Ellis & Young, 1996; Friedmann, Biran, & Dotan, 2013; Levelt, 1989,
1992; Nickels, 1997; Patterson & Shewell, 1987), each can be selectively impaired as a result of focal brain damage. These
models generally suggest that the retrieval of names starts with a non-verbally formulated conceptual representation that
activates the lexical representation of the word's meaning, which is stored in the semantic system. These representations do
not include the spoken names of the concepts, only their meaning. At this stage the speaker can provide detailed semantic
information about the item without retrieving yet its' phonological representation, namely the actual sound of the words. The
latter is only being activated at the next stage at the phonological output lexicon where the phonological structure of words is
stored. The phonological output lexicon then transmits the activation down to the phonological output buffer, a short-term
memory storage, responsible for the short-term maintenance of phonological units until their articulation (Franklin, Buerk, &
Howard, 2002; Nickels, 1997; Shallice, Rumiati, & Zadini, 2000), as well as for metrical, segmental and morphological
composition (Biran & Friedmann, 2005). Importantly each of these components can be selectively impaired following brain
damage resulting in different types of errors and different sensitivity to different word parameters such as the frequency of
the target words and their syllabic or phonemic length.
In the current study both the complexity of brain reorganization in language recovery, and the multi-level model of the
naming task, were considered. First, we examined a new protocol to nd the best stimulation type and the most appropriate
target site for each individual patient, including only right handed individuals. Initially we assessed naming performance,
then we applied anodal and cathodal-tDCS over the left and right STG and over the left and right IFG, and then we assessed
naming again. For each patient, we calculated the normalized improvement by dividing comparison of pre and post stimulation naming scores, by the patient SD of the items correctly named across the eight conditions. The patient's normalized
improvement indicated which site and type of simulation led to his greatest improvement. This site and type of stimulation
was then used in the patient's sequence of six treatments.
Another innovative aspect of our study focused on the functional loci of the naming impairment. Based on the fact that
individuals with anomia, do not form a unied group of patients we chose to include in the study only individuals with
naming decits as a result of a post semantic-lexical decit, at the phonological output lexicon or at the phonological output
buffer. To the best of our knowledge, no previous studies used this protocol for individually applying the site and type of tDCS
and for choosing the participants according to functional criteria that is based on such theoretical grounds. We aimed to nd
whether such individually-tailored protocol for a relatively functionally more unied group of patients could lead to more
consistent recommendations as to the best treatment approach, and whether it would lead to lasting effects. Due to the
different mechanisms underlying language processing after stroke, we expected to nd large variability in terms of the
appropriate type and site of stimulation within our study population.
2. Methods
2.1. Participants
Seven right handed individuals (5 men and 2 women, mean age 70 years, SD 16.07) participated in the study at the
Reuth Medical and Rehabilitation Center in Tel Aviv, Israel. All participants had chronic aphasia following left hemisphere
stroke. Two participants (LAH and CHA) had agrammatic (Broca) aphasia and ve (FO, ZUS, HOL, AZ, AR) suffered from anomic
aphasia. Data regarding exact localization of lesion were acquired from medical records (see Table 1). Only two patients (FO
and ZUS) also provided us with their CT scans. Time post stroke onset was 1e14 years (mean 5.9, SD 4.17). Recruitment
and procedure were approved by a Helsinki Committee at the Reuth Medical and Rehabilitation Center and participants
signed a consent form that explained study aims. All participants spoke uent Hebrew prior to their stroke. Patients did not
receive any additional naming or other speech therapy treatment during the period of their participation in the study.

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Table 1
Patients' information and type and site of stimulation.
Patient

Sex

Age

Years post
stroke

Lesion location

Type of
stimulation

Location of
stimulation

Normalized
improvement

FO
ZUS
HOL
AZ
AR
LAH
CHA

M
M
M
F
M
M
F

57
80
84
78
67
60
43

1
6
6
1
4
1.5
14

Left
Left
Left
Left
Left
Left
Left

Anodal
Cathodal
Anodal
Cathodal
Cathodal
Anodal
Cathodal

R-IFG
R-STG
L-IFG
R-STG
L-STG
R-STG
L-STG

1.52
2.73
2.26
3.28
2.10
10.63
1.80

MCA
ICA
corona radiata
basal nuclei
fronto-temporaleoccipital lobes
MCA
fronto-temporal lobes and left basal nuclei

Note: MCA middle cerebral artery; ICA internal carotid artery; R-IFG right inferior frontal gyrus; L-IFG left inferior frontal gyrus; R-STG right
superior temporal gyrus; L-STG left superior temporal gyrus.

Exclusion criteria consisted of apraxia of speech, dysarthria, a history of psychiatric disease, current use of anti-psychotic
drugs, a history of alcohol abuse, pregnancy, presence of a pacemaker, and a seizure within the last 36 months. Tables 1e3
summarize all the background details including the participants' performance on each of the background tasks described
in the forthcoming paragraphs.
Importantly, all participants suffered from a severe naming impairment as was evinced by their performance in the
SHEMESH picture naming test (Biran & Friedmann, 2004) (see Section 2.2.1 for more details). As one of the inclusive criteria
was a post lexical semantic decit, several tasks and analysis of naming errors were done to identify the exact functional locus
of the naming decit of each participant. Accordingly, all seven participants had indeed primary decits post the semantic
lexicon (See Tables 2 and 3).
In the forthcoming sections we will describe the tasks and the naming analysis used for selecting the participants and to
assess their locus of the naming decit.
2.1.1. Identifying the functional loci of the naming decit: description of the tasks and patients' performance
In order to examine the conceptual system and the semantic lexicon we used the following tasks.
PALPA 47. A spoken word-to-picture matching task taken from the Psycholinguistic Assessment of Language Processing in
Aphasia (PALPA 47, Kay, Lesser, & Coltheart, 1992; Hebrew version Gil & Edelstein, 2001). This test consists of 40 groups of ve
pictures including a target word (i.e. a dog) and four close and distant semantic distracters (i.e. a cat, a giraffe, a rocking horse,
and a kite respectively). The task is to match a spoken word to an appropriate picture.
Written Word Association test (Biran & Friedmann, 2007). This task includes 35 triads of written words, respectively, with
one target presented at the top and two word stimuli presented at the bottom. The task is to choose the word semantically
related to the target. For example the word a glove is presented at the top and the patients are requested to choose between
the written words a hand and a foot presented at the bottom. The last 10 triads depict more abstract terms (e.g. honesty,
truth, and lie).
Picture Association test (Biran & Friedmann, 2007). In order to examine the conceptual system we used the pictorial version
of the written word association test described above. This test is identical to the word association task however pictures of
objects are presented instead of written words. In contrast to the written word version, the last 10 triads are pictures of
concrete objects.
Participants' performance in the various tasks. As can be seen in Table 2, all participants received a score of 88% and above
(mean: 94.8%, SD 3.34) on the Picture Association task; a score of 88% and above (mean: 94%, SD 4.7) on the Word Association task; and 80% and above (mean: 91.6%, SD 5.62) on PALPA 47.
2.1.2. Identifying the functional loci of the naming decit: analyzing types of errors and effects on naming
Another way to assess the functional loci of the naming decit is to analyze the error types in the naming task and to see
which if any effects (frequency or length effects) affect the performance. Typically, errors as a result of an impairment to the
conceptual system involve unrelated lexical or nonlexical substitution errors such as producing table instead of orange (Dotan
& Friedmann, 2015; Friedmann et al., 2013). A decit in the semantic lexicon is manifested by semantic paraphasias, namely
replacing the target word with a semantically related word (i.e. producing cat instead of dog) (Caramazza & Hillis, 1990; Dotan
& Friedmann, 2015; Rapp & Goldrick, 2000). Decits in the phonological output lexicon or in the phonological output buffer
would be manifested in various phonological errors, such as formal paraphasias e replacing the target word with a
phonologically similar lexical word (table, stove, pale instead of cable, stone, tape, respectively); phonemic paraphasia e
replacing the target word with a phonological similar nonword (mable instead of table). Similarly, there may be errors
involving the order of the segments resulting in formal paraphasia (breadebeard) or a phonemic paraphasia (grapefruitegrafepruit). Additional possible phonological errors are tip-of-the-tongue (TOT) states and phonological approximations
(uniler uli univerm university). Semantic paraphasias may sometimes be also present in addition to phonological errors
in a decit in the phonological output lexicon (Caramazza & Hillis, 1990; Rapp & Goldrick, 2000).
Effects on naming. Given that the phonological output lexicon is organized by frequency, an impairment at this stage is
manifested in better performance in naming high frequency words compared to lower frequency words (Nickels, 1997). Given
that the phonemic buffer is a phonemic short term memory store, the buffer is affected by the length of the spoken word. As

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Table 2
Patients' language prole.
Aphasia type

Conceptual semantics
(correct %)

FO
ZUS
HOL
AZ

Anomic
Anomic
Anomic
Anomic

AR
LAH
CHA

Anomic
Agrammatic (Broca)
Agrammatic (Broca)

Lexical semantics (correct %)

Written word
association

PALPA 47

Number and types of errors within PALPA 47

97%
97%
88%
97%

97%

97%
95%
90%
80%

97%
94%
100%

100%
88%

1
1
3
2
2
2
2
3

91%
94%

95%
92%
92%

semantic related
semantic related; 1 visual
semantic related; 1 don't know
semantic related; 3 semantic distant;
don't know; 1 visual
semantic related
semantic related; 1 visual
semantic related

Table 3
Rates of errors of different types and word effects in confrontation naming task.
Formal
Phonemic Pho.
Paraphrase No verbal Latency Perseveration Frequency Length
Correct/total Semantic Semantic
response
effect (rpb) effect (rpb)
non-related paraphasias paraphasia approx
related
errors
errors
FO
ZUS
HOL
AZ
AR
LAH
CHA

47/100*
64/100*
14/100*
41/100*
11/100*
31/100*
48/96*

5
8
14
16
12
13
11

3
1
1
8
1

1
2
18
4

11
3
3
26
21
3
11

22
3
1
7
38
12
16

1
5
37
2
6
8
3

2
12
31
7
7
31
6

16
5
3
2
3
6

2
3

.2**
n.s
.3**
n.s.
.3**
.2**
.4**

.4**
.5**
n.s
n.s.
n.s
n.s
n.s

* Signicantly lower score than age-matched control group (p < .05).

**p  .02.
Pho. approx phonological approximations.
Note. The total number of types of errors exceeds the total number of errors given that in some words the participants produced more than one error.

such, a decit within the phonemic buffer impairs lexical retrieval as the target words get longer. Thus frequency effect was
calculated as the point biserial correlation between the word frequency and the success in producing it. Length effect was
calculated as the point biserial correlation between the word length in phonemes and the success in producing each phoneme
(see Table 3).
Analysis of the participants' naming errors and effects. As demonstrated by Table 3, the prominent errors of all participants
were errors that typically result from post semantic lexical decits. These errors include formal and phonemic paraphasias,
phonological approximations and TOT states. Additional errors suggest semantic paraphasias that are also typical to post
lexical decits in addition to the phonological errors. Similarly, the effects on naming, as shown in Table 3, were those
typically evinced following post lexical semantic decits, namely frequency and length effects.
2.2. Procedure
2.2.1. Assessing naming abilities
Picture naming was assessed with the SHEMESH stimuli (Biran & Friedmann, 2005) as well as additional 100 colorful
pictures of nouns that were matched for frequency of use (t(198) .366, ns) (as estimated according to Linzen, 2009) for
number of syllables (t(198) .548, ns) and number of letters (t(198) .801, ns) with the SHEMESH stimuli. Each of the 200
pictures was presented in the same size (10.05 cm in width and 9 cm in height) at the center of a personal computer screen for
10 s, and participants were asked to name them. Participants who named correctly more than 75% of the pictures were not
included in the tDCS study so as to avoid ceiling effects.
The 200 pictures were divided into eight blocks of 25 pictures each to be used with the eight tDCS assessment conditions
(anodal/cathodal  left/right IFG, anodal/cathodal  left/right STG). Blocks were matched for grammatical class (nouns), size
(M 10.05  9 cm; width and height respectively) and did not differ from each other statistically in terms of frequency
(F(7,192) .62, ns), number of syllables (F(7,192) .56, ns) and number of letters (F(7,192) 1.06, ns). Pictures in each of the eight
blocks were reordered so as not to repeat the order of the original full list.
2.2.2. Identifying stimulation type and site
The rst four tDCS assessment sessions were devoted to identify the optimal brain area and stimulation type. In each
session, two stimulations were tested, each at the intensity of 2 mA, applied for 10 min. A 40 min break was given between the
two stimulation sessions as a wash over period (as recommended by Vines, Nair, & Schlaug, 2008). A total of eight stimulations were applied at this stage, half anodal and half cathodal, each type of stimulation applied in four different locations:
the left IFG, the right IFG, the left STG and the right STG.

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

After each of the eight tDCS conditions, the participant was asked to name a block of 25 pictures. The number of pictures
named correctly post tDCS stimulation was compared to the number of pictures (of the same block) named correctly prior to
tDCS. The average number of items named correctly by each person across all eight blocks was computed. A normalized
improvement was then calculated by dividing the delta of correct answers (post vs. pre tDCS) by each participant's standard
deviation of correct responses across the eight blocks. The specic type and site of stimulation that produced the largest
improvement was used in subsequent treatments. Table 1 shows the type and site of stimulation determined for each patient.
Fig. 1 depicts the procedure of the baseline session.
2.2.3. tDCS
Six treatment sessions were applied, three treatments per week for two weeks. tDCS (2 mA) was delivered for 10-min per
session through two saline-soaked sponge electrodes (5  7 cm), using a constant current stimulator (neuroConn DC
stimulator plus, Incl GmbH). The active electrode was placed over the brain area that was found to induce the greatest
improvement in naming. The left IFG was dened as the crossing point between T3-Fz and F7-Cz. The right IFG was dened as
the crossing point between T4-Fz and F8-Cz. The left and right STG were identied as T3 and T4, respectively, using the 10e20
EEG system. The reference electrode was placed over the contralateral supraorbital area. In order to test for a possible placebo
effect, six sham sessions were also conducted. On these sessions the same montage of the real stimulation session was used
but the stimulator was turned off after 30 s. This procedure ensured that participants felt the initial itching sensation at the
beginning of tDCS, but prevented any effective modulation of cortical excitability by tDCS. Individuals were not informed
about the type of stimulation they received. Sham sessions were administered three times a week for two weeks, beginning
three months after the end of treatment sessions. Fig. 2 presents the full study design.
2.2.4. Assessment of treatment
Three post-treatment assessments were conducted. Following the six treatment sessions, participants were asked to name
100 pictures out of the 200 that had been presented at baseline. One month after the end of treatment participants were
asked to name the remaining 100 pictures. Three months after the end of treatment participants were asked to name the rst
100 pictures again. This assessment was used to examine lasting treatment effects, and it also served as the pre-sham baseline
measure. Once sham sessions were over, participants were asked to name the remaining 100 pictures.
Altogether, naming was tested in nine sessions (baseline, four sessions to dene type and location of stimulation,
immediately post treatment, one month post treatment, three months post treatment, post sham). tDCS was administered in
16 sessions (four sessions to dene type and location of stimulation, six treatment sessions, and six sham sessions).
3. Statistical analysis
The main outcome parameter was percent of pictures named correctly. First, the site and type of stimulation that best ts
each of the patients was detected. Next, a repeated-measures analysis of variance (ANOVA) was conducted to determine
improvement in treatment as compared with sham. This analysis used the improvement in percent of correct responses on
the naming assessments as the dependent variable, with four conditions: difference between baseline and immediately post
treatment naming performance, difference between baseline performance and performance one month after treatment,
difference between baseline performance and performance three months after treatment, and difference between performance before and after sham. Next, a repeated-measures ANOVA was conducted to determine short and long-term training
outcome. This analysis used percent of correct responses as the dependent variable, with four conditions: baseline, immediately after treatment, one month after treatment, and three months after treatment. Planned comparisons were conducted
following the ANOVAs to compare improvement from baseline to each of the post-treatment conditions (immediately after
treatment, one month after treatment, and three months after treatment).

Fig. 1. Outline of a specic pre-treatment session to assess stimulation type and site. In four pre-treatment sessions (two stimulation in each session), eight
different stimulation combinations were administered, (anodal/cathodal over left IFG; anodal/cathodal over right IFG; anodal/cathodal over left STG; anodal/
cathodal over right STG).

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Fig. 2. Study design.

Finally, we used Fisher's exact test to examine whether age and time since stroke differed signicantly between participants who showed a signicant improvement and participants who showed no signicant improvement (using the Wilcoxon
signed rank test).
4. Results
All participants tolerated tDCS well and no adverse effects related to the application of tDCS were demonstrated. All
patients completed both treatment and sham phases and all accompanying testing sessions.
4.1. Heterogeneity of site and type of stimulation among patients
In order to detect the site and type of stimulation that best ts each of the patients, we computed for each tDCS condition
at the pre-treatments assessment sessions, the patient's normalized improvement by dividing the delta of his naming performances (post minus pre tDCS) by the patient SD of the items correctly named across the eight blocks. Relative to baseline,
post tDCS naming performance improved in all participants in more than 1.5 standard deviations. The range of this
normalized improvement varied among patients from 1.52 to 10.63 standard deviations (see Table 1). As expected, there was
large variability of stimulation types and locations across participants. As can be seen in Table 1, ve different combinations of
stimulation type and location were used. Only two pairs of participants were found to benet from the same type of stimulation (ZUS and AZ: Cathodal-tDCS over right STG; AR and CHA: cathodal-tDCS over left STG).
4.2. Comparing treatment to sham
The results of the individual percentage of improvement in correct responses across time are presented in Table 4.
To determine whether the patients' ability to name pictures improved due to tDCS treatment, a one-way repeatedmeasures ANOVA was conducted with delta percentage of correct responses (as compared with baseline) as the dependent
variable and time as a within subject factor (before and immediately after treatment, before and one month after treatment,
before and three month after treatment, before and after sham). A signicant difference was found across the four relevant
conditions (F(3,18) 4.57, p < .05) (see Fig. 3). Planned comparisons showed that the improvement from baseline to immediately after treatment (M 30.95%, SD 23.51) was signicantly greater than the change following sham (M .53%,
SD 23.14), t(6) 3.84, p < .01. Improvement from baseline to three months post treatment (M 28.00%, SD 20.57) was
signicantly greater than change following sham, t(6) 3.77, p < .01. Improvement from baseline to one month after
treatment (M 8.01%, SD 11.16) did not differ signicantly from improvement following sham, t(6) 1.68, p .14.
4.3. Follow up measurements
To examine whether the tDCS effect lasts beyond the day of treatment, we conducted a one-way repeated-measures
ANOVA with percentage of correct responses as the dependent variable. This analysis showed a signicant difference in
performance across sessions (F(3,18) 5.844, p < .01). Planned comparisons revealed a signicant difference between baseline
performance (M 37.71%, SD 18.99) and performance immediately after treatment (M 46.86%, SD 21.5), t(6) 4.79,
p < .01, as well as between baseline performance and performance three months after treatment (M 47.14%, SD 22.68),

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Table 4
Individual naming results. Percentage of improvement in correct responses in each of the three time points compared to baseline. Signicant improvement
was tested by Wilcoxon signed rank test.
Participant

Lesion location

Baseline vs. immediately

post treatment

Baseline vs. one month

post treatment

Baseline vs. three months

post treatment

Before vs.
post sham

FO
LAH
ZUS
HOL
AZ
AR

Left MCA
Left MCA
Left ICA
Left corona radiata
Left basal nuclei
Left fronto-temporal-occipital
lobes
Left fronto-temporal lobes and
left basal nuclei

28.85*
3.13
20.31*
71.43*
32.50
50.00

21.43*
6.90
28.30*
21.05
13.46
60.00**

42.31**
9.38
7.81
14.29
37.50*
64.29*

4.05
37.14
5.80
0
30.9*
21.74

10.42

12.28

22.92

18.64

30.95
8.89

8.01
11.16

28.00
7.77

10.69
20.02

CHA
Mean
SD

Note: *p < .05, **p < .01.

MCA middle cerebral artery; ICA internal carotid artery.

Fig. 3. Mean percentage of improvement in correct responses (and standard errors) in each of the three time points compared with sham.

t(6) 3.53, p < .05. The difference between baseline performance and performance one month after treatment (M 43.57%,
SD 20.26) was not signicant t(6) 1.87, p .11.

4.4. Individual performances

Next we examined for each participant whether naming performance improved signicantly from baseline at each time
point post treatment using the Wilcoxon signed rank test. Thus, naming of the list of 100 pictures at baseline was compared
with the corresponding list of pictures at each time point post treatment (immediately, one month, three month after
treatment, and pre vs. post sham). Signicant comparisons are marked with an asterisk in Table 4. As is shown in that table,
ve participants (FO, ZUS, HOL, AZ, AR) out of the seven, improved signicantly from baseline at least at one point post
treatment: One participant (FO) improved at each of the three post treatment points; two participants (ZUS, AR) improved at
two post treatment points while the naming of one participant (AZ) improved signicantly only at one point post treatment

10

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

compared to baseline. In general the improvement, looking individually at each patient, was across all points of post treatment sessions given that at each point three patients manifested a signicant naming improvement compared to baseline.
As for possible improvement following sham, Table 4 shows that none of the patients improved their naming abilities due
to sham. On the contrary, one patient (AZ) even deteriorated signicantly in her naming abilities after sham (p .037).
4.5. Effects of age and time since stroke on naming improvement
The participants were divided into two groups according to age (below or above average age), as well as according to time
since stroke (below or above average time since stroke). A Fisher's exact test was used to test whether there is a signicant
relationship between age and signicant improvement as well as between years post stroke and signicant improvement.
There was no signicant relation between signicant improvement in naming immediately after treatment and age (p .37),
one month after treatment and age (p .62), and three months after treatment and age (p .15). We also found no signicant
relation between time since stroke and improvement in naming immediately after treatment (p .37), one month after
treatment (p .62), and three months after treatment (p .62).
5. Discussion
In the present study we found that six tDCS sessions signicantly improved naming accuracy among patients with chronic
aphasia, whereas sham stimulation had no such effect. Furthermore, improvement was found to last three months post
treatment.
Heterogeneity was found in terms of both stimulation type and location as well as in individual improvement in naming
performances though the participants shared similar functional decit and in particular, all the patients demonstrated post
lexical semantic decits. At each time point, three participants demonstrated signicant improvement but participants who
improved immediately post treatment did not necessarily improve after one month. As our last analysis shows, it is unlikely
that age and time after stroke led to this diversity. Furthermore, as we will discuss later, we tested patients in their chronic
phase of their aphasia, hence, it is also less likely that spontaneous recovery is the main cause for the diversity in naming
improvement. Instead, we suggest that the core of these diversities lay in the individual differences associated with lesion site
and size and, in particular, with individual differences in language reorganization after stroke. In our study, all participants
suffered from lesions within their left hemisphere, but as can be seen in Table 4, there were differences in the extent of the
lesions and in their exact locations within the left hemisphere. Current theories on brain organization suggest that language is
organized in the brain in two widespread, segregated, and overlapping networks, a dorsal stream which involves mapping
sound to articulation, and a ventral stream which involves mapping sound to meaning (Saur et al., 2008). The persistence of
these separate pathways was conrmed using imaging techniques of fMRI and diffusion tensor imaging (DTI). Thus, in the
current study, differences in lesion location might affect processing within the dorsal stream, the ventral stream, or both,
depending on lesion extent and exact location. However, looking at Table 4, no clear trend in naming improvement is found
among patients with similar lesion location. Our patients provided limited data about their lesions which may be the reason
that a trend could not be determined. Holland and Crinion (2012) had emphasized the challenge of choosing the stimulation
target among patients with aphasia. The authors also pointed out that in chronic post-stroke aphasia the interhemispheric
relationship may be altered as a consequence of time and brain-language recovery after stroke and therefore the extended
speech production network is differentially engaged (Holland & Crinion, 2012).
In our study, ve combinations of stimulation type and location were found to be benecial for the seven patients suggesting that different pathways were constructed during brain reorganization post their stroke. For instance, whereas data
from patient HOL's medical record showed that he suffered a lesion which was relatively centered around his left corona
radiata, patient ZUS suffered from a much wider lesion around his left internal carotid artery (Tables 1 and 4). Patient HOL
with the relatively centered damage was found to best benet from anadoal-tDCS over his left IFG. Taking into account the
hierarchical model that was described by Heiss and Thiel (2006), it might be conceivable that in that case, parts of the original
language networks within the left hemisphere are restored. Anodal-tDCS could enhance the naming abilities of this patient
through induced increase in activation of peri-lesional preserved language networks. On contrary, patient ZUS who suffered a
wider injury within his left hemisphere beneted best from cathodal-tDCS over his right STG. In this case, the right hemisphere involvement in language recovery might have reected increased transcallosal inhibition from right hemisphere to the
damaged left hemisphere (Hamilton et al., 2011). Therefore, it's plausible that cathodal-tDCS over the right STG led to
improvement in naming abilities due to tDCS-induced depression of cortical inhibitory inter-neurones, that led to disinhibition and, consequently, to improved function of the damaged language areas. More detailed information about the patients'
brain lesion could have helped us draw a clearer picture of such assumption for brain re-organization. Hence, the variety of
stimulation types that was found in the current study may reect heterogeneity in lesion characteristics and in language
reorganization mechanisms post stroke. The fact that it is hard to detect the exact language recovery pathway that serves each
patient and the large heterogeneity that was found among our patient stimulation types, emphasize the importance of using
an individualized treatment protocol.
In addition to nding effects of treatment, we also found that they lasted up to three months post treatment (at group level
analysis and in 3 individual cases). Previous studies measured lasting effects at a much closer time point to the end of
el et al., 2011) and three weeks
treatment such as one week post treatment (Baker et al., 2010), two weeks post treatment (Flo

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

11

post treatment (Fiori et al., 2013; Fridriksson, Richardson, Baker, & Rorden, 2011). However, interestingly, despite the signicant improvement that we measured three months post treatment, no signicance improvement was found after one
month, at least for the whole group. As we included individuals with chronic aphasia, it is unlikely that improvement was due
to spontaneous recovery. Moreover, improvement cannot be ascribed to traditional speech therapy treatment since no such
treatment was conducted during the study. Furthermore, according to these ndings it is less likely to assume that the
improvement in naming was induced by repeated exposure to the pictures or by priming effect. If that was the case, we would
have expected to nd improvement at a closer time point to end of treatment such as a month later. The fact that lasting effect
was found three month but not one month post treatment combined with the lack of improvement after sham stimulation,
strengthen the assumption that it was the stimulation (and not spontaneous recovery or priming effect) that enhanced the
improvement. One possible explanation for this type of lasting effect can be revealed while looking in a recent study of
DaSilva et al. (2012). The authors studied the implications of tDCS over the left motor cortex among patients with chronic
migraine and found a delayed effect. Hence, signicant improvement in terms of decrease in pain intensity and length of
migraine was found only in the long term evaluation four months post treatment, but not immediately or two months post
treatment. One of the authors' explanations for this delayed effect was their method of stimulation. On contrary to other pain
studies, they used a protocol in which stimulation was delivered every other day. The authors mentioned that this might have
reduced the initial efciency of tDCS as compared to daily sessions and therefore cumulative changes developed more
gradually (DaSilva et al., 2012). In our study we also delivered the stimulation every other day. It is notable to mention that the
mean of performances one month post treatment was still found to be higher than the pre treatment mean, but this trend did
not reach signicance. Delivering daily stimulations sessions might have enhanced a more stabled effect and prevented the
decrease in signicance that we found a month post treatment. Furthermore, even when considering the lasting effect that
we found, as DaSilva et al. (2012) mentioned, repeated exposures maybe needed in order to induce lasting plastic changes
promoting synaptic strengthening of the structures targeted specially after stroke.
In our study we used unilateral stimulation in order to improve naming abilities of patients with Aphasia. However, recent
evidence points to the potential of dual-hemisphere stimulation to enhance language improvement after stroke (Khedr et al.,
2014). For instance, Khedr et al. (2014) studied 30 patients with subacute post-stroke nonuent aphasia using real or sham
rTMS. The researchers delivered low frequency r-TMS over the right unaffected Broca's area, combined with high frequency
rTMS over the left affected Broca's area for 10 consecutive days, followed by speech/language training. The real stimulation,
but not the sham, signicantly improved scores in the language section of the Hemispheric Stroke Scale (HSS), as well as
scores in Stroke Aphasic Depression Questionnaire-Hospital Version (SADQ-H). The improvement lasted two months post
stimulation. Further investigation is required in order to determine the optimal frequency of treatment and the most
appropriate type (uni/dual) of stimulation to induce long term improvement.
The current study has some limitations. First, another group in which the sham and treatment will be given in the opposite
order is needed. Optimally, one group of patients would participate in the Active-Sham and the second group in the ShamActive stimulation. This optimal design would have claried whether the lack of improvement following sham is the result of
ceiling effect or the result of the non-therapeutic effects of the sham. Future studies should randomized and counterbalanced
order of stimulation condition. Secondly, the current study lacks functional imaging data before and after the treatment. This
data might provide valuable evidence regarding the specic brain networks that are involved in post stroke lexical retrieval
reorganization. However, using an individual e based protocol and using a functional criteria for choosing our patients, seem
to offer an efcient way to cope with the later challenge. Third, although our Broca's aphasics and the anomic (posterior)
aphasics had common cognitive mechanisms that were related to naming errors, it is possible that their improvements in
tDCS treatment protocol may have been mediated by other cognitive mechanisms. It is possible that the patients may rely on
executive functions modulated by their frontal lobes or use some degrees of preserved attention capacities. Future studies
with a bigger sample size should use measures of connectivity or anatomical variables on patient sub-groups in order to
better dene the target regions.
6. Conclusions
Though having few limitations, this preliminary individualized study seems to be promising and shed a light on the
complexity of the mechanisms that underlie language recovery after stroke. An individualized protocol that contains six
sessions of tDCS at the intensity of 2 mA for 10 min can improve naming abilities in individuals with chronic aphasia.
Improvement was found to last three months post treatment. Our ndings may have important clinical implications for the
use of tDCS as a potential tool to enhance language abilities after stroke. Future studies that combine the use of tDCS and EEG
monitoring or other functional imaging techniques might help us to cope with the challenge of nding the best stimulation
montage for each stroke patient. Moreover, further investigation with a larger sample size is needed as well as further
assessment of the effect of tDCS when applied in combination with other traditional speech and language therapy.

References
Andoh, J., Artiges, E., Pallier, C., Riviere, D., Mangin, J. F., Cachia, A., et al. (2006). Modulation of language areas with functional MR image-guided magnetic
stimulation. Neuroimage, 29, 619e627.

12

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

Andoh, J., & Martinot, J. L. (2008). Interhemispheric compensation: a hypothesis of TMS-induced effects on language-related areas. European Psychiatry,
23(4), 281e288.
Baker, J., Rorden, C., & Fridriksson, J. (2010). Using transcranial direct current stimulation (tDCS) to treat stroke patients with aphasia. Stroke, 41(6),
1229e1236.
Barlow, T. (1877). On a case of double hemiplegia, with cerebral symmetrical lesions. British Medical Journal, 2(865), 103.
Basso, A., Gardelli, M., Grassi, M. P., & Mariotti, M. (1989). The role of the right hemisphere in recovery from aphasia. Two case studies. Cortex, 25(4),
555e566.
Biran, M., & Friedmann, N. (2004). SHEMESH: Naming a hundred objects. Tel Aviv: Tel Aviv University.
Biran, M., & Friedmann, N. (2005). From phonological paraphasias to the structure of the phonological output lexicon. Language and Cognitive Processes,
20(4), 589e616.
Biran, M., & Friedmann, N. (2007). Ma kashur: Picture association test. Tel Aviv University.
Blasi, V., Young, A. C., Tansy, A. P., Petersen, S. E., Snyder, A. Z., & Corbetta, M. (2002). Word retrieval learning modulates right frontal cortex in patients with
left frontal damage. Neuron, 36(1), 159e170.
Boggio, P. S., Nunes, A., Rigonatti, S. P., Nitsche, M. A., Pascual-Leone, A., & Fregni, F. (2007). Repeated sessions of noninvasive brain DC stimulation is
associated with motor function improvement in stroke patients. Restorative Neurology and Neuroscience, 25(2), 123e129.
Buckner, R. L., Corbetta, M., Schatz, J., Raichle, M. E., & Petersen, S. E. (1996). Preserved speech abilities and compensation following prefrontal damage.
Proceedings of the National Academy of Sciences, 93(3), 1249e1253.
Butterworth, B. (1989). Lexical access in speech production. In W. Marslen-Wilson (Ed.), Lexical representation and process. Cambridge, MA: MIT Press.
Cappa, S. F., Perani, D., Grassi, F., Bressi, F., Alberoni, M., Franceschi, M. Fazio, M. (1997). A PET follow-up study of recovery after stroke in acute aphasics.
Brain and Language, 56(1), 55e67.
Caramazza, A., & Hillis, A. E. (1990). Where do semantic errors come from? Cortex, 26, 95e122.
Catani, M., Jones, D. K., & Fytche, D. H. (2005). Perisylvian language networks of the human brain. Annals of Neurology, 57, 8e16.
Catani, M., & Mesulam, M. (2008). The arcuate fasciculus and the disconnection theme in language and aphasia: history and current state. Cortex, 44(8),
953e961.
Crinion, J., Holland, A. L., Copland, D. A., Thompson, C. K., & Hillis, A. E. (2013). Neuroimaging in aphasia treatment research: quantifying brain lesions after
stroke. Neuroimage, 73, 208e214.
Crosson, B., Moore, A. B., Gopinath, K., White, K. D., Wierenga, C. E., Gaiefsky, M. E., et al. (2005). Role of the right and left hemispheres in recovery of
function during treatment of intention in aphasia. Journal of Cognitive Neuroscience, 17(3), 392e406.
Damasio, H., Grabowski, T. J., Tranel, D., Hichwa, R. D., & Damasio, A. R. (1996). A neural basis for lexical retrieval. Nature, 380, 499e505.
Damasio, H., Tranel, D., Grabowski, T., Adolphs, R., & Damasio, A. (2004). Neural systems behind word and concept retrieval. Cognition, 92(1), 179e229.
DaSilva, A. F., Mendonca, M. E., Zaghi, S., Lopes, M., DosSantos, M. F., Spierings, E. L., et al. (2012). tDCS induced analgesia and electrical elds in pain related
neural networks in chronic migraine. Headache: The Journal of Head and Face Pain, 52(8), 1282e1295.
Datta, A., Baker, J. M., Bikson, M., & Fridriksson, J. (2011). Individualized model predicts brain current ow during transcranial direct-current stimulation
treatment in responsive stroke patient. Brain Stimulation, 4(3), 169e174.
Dell, G. S. (1986). A spreading-activation theory of retrieval in sentence production. Psychological Review, 93, 283e321.
Dotan, D., & Friedmann, N. (2015). Steps toward understanding the phonological output buffer and its role in the production of numbers, morphemes, and
function words. Cortex, 63, 317e351.
Ellis, A. W., & Young, A. W. (1996). Human cognitive neuropsychology. Hove, UK: Erlbaum.
Fama, M. E., & Turkeltaub, P. E. (2014). Treatment of poststroke aphasia: current practice and new directions. In Seminars in Neurology, 34(5), 504e513.
Fertonani, A., Rosini, S., Cotelli, M., Rossini, P. M., & Miniussi, C. (2010). Naming facilitation induced by transcranial direct current stimulation. Behavioural
Brain Research, 208(2), 311e318.
Fiori, V., Cipollari, S., Di Paola, M., Razzano, C., Caltagirone, C., & Marangolo, P. (2013). tDCS stimulation segregates words in the brain: evidence from
aphasia. Frontiers in Human Neuroscience, 7(269), 1e11.
Fiori, V., Coccia, M., Marinelli, C. V., Vecchi, V., Bonifazi, S., Ceravolo, M. G., et al. (2011). Transcranial direct current stimulation improves word retrieval in
healthy and nonuent aphasic subjects. Journal of Cognitive Neuroscience, 23(9), 2309e2323.
el, A., Meinzer, M., Kirstein, R., Nijhof, S., Deppe, M., Knecht, S., et al. (2011). Short-term anomia training and electrical brain stimulation. Stroke, 42(7),
Flo
2065e2067.
Franklin, S., Buerk, F., & Howard, D. (2002). Generalised improvement in speech production for a subject with reproduction conduction aphasia. Aphasiology,
16, 1087e1114.
Fregni, F., Boggio, P. S., Mansur, C. G., Wagner, T., Ferreira, M. J., Lima, M. C., et al. (2005). Transcranial direct current stimulation of the unaffected hemisphere
in stroke patients. Neuroreport, 16(14), 1551e1555.
Fridriksson, J., Richardson, J. D., Baker, J. M., & Rorden, C. (2011). Transcranial direct current stimulation improves naming reaction time in uent aphasia: a
double-blind, sham-controlled study. Stroke, 42(3), 819e821.
Friedmann, N., Biran, M., & Dotan, D. (2013). Lexical retrieval and breakdown in aphasia and developmental language impairment. In C. Boeckx, & K. K.
Grohmann (Eds.), The Cambridge handbook of biolinguistics (pp. 350e374). Cambridge, UK: Cambridge University Press.
Garcia, G., Norise, C., Faseyitan, O., Naeser, M. A., & Hamilton, R. H. (2013). Utilizing repetitive transcranial magnetic stimulation to improve language
function in stroke patients with chronic non-uent aphasia. JoVE (Journal of Visualized Experiments), 77. e50228ee50228.
Geschwind, N. (1965). Disconnexion syndromes in animals and man. Brain, 88, 237e294.
Geschwind, N. (1970). The organization of language and the brain. Science, 170, 940e944.
Geschwind, N., & Levitsky, W. (1968). Human brain: lefteright asymmetries in temporal speech region. Science, 161, 186e187.
Gil, M., & Edelstein, C. (2001). Hebrew version of the PALPA. Ra'anana, Israel: Loewenstein Hospital Rehabilitation Center.
Gowers, W. R., & Barker, A. E. (1886). On a case of abscess of the temporo-sphenoidal lobe of the brain, due to otitis Media, successfully treated by trephining
and drainage. British Medical Journal, 2(1354), 1154.
Hamilton, R. H., Chrysikou, E. G., & Coslett, B. (2011). Mechanisms of aphasia recovery after stroke and the role of noninvasive brain stimulation. Brain and
Language, 118(1), 40e50.
Heiss, W. D., & Thiel, A. (2006). A proposed regional hierarchy in recovery of post-stroke aphasia. Brain and Language, 98(1), 118e123.
Holland, R., & Crinion, J. (2012). Can tDCS enhance treatment of aphasia after stroke? Aphasiology, 26(9), 1169e1191.
Iyer, M. B., Mattu, U., Grafman, J., Lomarev, M., Sato, S., & Wasserman, E. M. (2005). Safety and cognitive effect of frontal DC brain polarization in healthy
individuals. Neurology, 64(5), 872e875.
Jo, J. M., Kim, Y. H., Ko, M. H., Ohn, S. H., Joen, B., & Lee, K. H. (2009). Enhancing the working memory of stroke patients using tDCS. American Journal of
Physical Medicine and Rehabilitation, 88(5), 404e409.
Kang, E. K., Baek, M. J., Kim, S., & Paik, N. J. (2009). Non-invasive cortical stimulation improves post-stroke attention decline. Restorative Neurology and
Neuroscience, 27(6), 647e652.
Kay, J., Lesser, R., & Coltheart, M. (1992). Psycholinguistic assessment of language processing in aphasia (PALPA). Hove, UK: Lawrence Erlbaum Associates Ltd.
Khedr, E. M., El-Fetoh, N. A., Ali, A. M., El-Hammady, D. H., Khalifa, H., Atta, H., et al. (2014). Dual-hemisphere repetitive transcranial magnetic stimulation for
rehabilitation of poststroke aphasia a randomized, double-Blind clinical trial. Neurorehabilitation and Neural Repair, 28(8), 740e750.
Kinsbourne, M. (1971). The minor cerebral hemisphere. Archives of Neurology, 25(4), 302e306.
el, A., Dr
Knecht, S., Flo
ager, B., Breitenstein, C., Sommer, J., Henningsen, H., et al. (2002). Degree of language lateralization determines susceptibility to
unilateral brain lesions. Nature Neuroscience, 5(7), 695e699.

A. Lifshitz Ben Basat et al. / Journal of Neurolinguistics 38 (2016) 1e13

13

Lefaucheur, J. P. (2006). Stroke recovery can be enhanced by using repetitive transcranial magnetic stimulation (rTMS). Neurophysiologie Clinique/Clinical
Europhysiology, 36(3), 105e115.
Levelt, W. J. M. (1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press.
Levelt, W. J. M. (1992). Accessing words in speech production: stages, processes and representations. Cognition, 42, 1e22.
Linzen, T. (2009). Corpus of blog postings collected from the Israblog website. Tel Aviv University.
Liuzzi, G., Freundlieb, N., Ridder, V., Hoppe, J., Heise, K., Zimerman, M., et al. (2010). The involvement of the left motor cortex in learning of a novel action
word lexicon. Current Biology, 20(19), 1745e1751.
Martin, P. I., Naeser, M. A., Doron, K. W., Bogdan, A., Baker, E. H., Kurland, J., et al. (2005). Overt naming in aphasia studied with a functional MRI hemodynamic delay design. Neuroimage, 28(1), 194e204.
Martin, P. I., Naeser, M. A., Ho, M., Doron, K. W., Kurland, J., Kaplan, J., et al. (2009). Overt naming fMRI pre-and post-TMS: two nonuent aphasia patients,
with and without improved naming post-TMS. Brain and Language, 111(1), 20e35.
Mesulam, M. M. (1990). Large-scale neurocognitive networks and distributed processing for attention, language, and memory. Annals of Neurology, 28,
597e613.
Monti, A., Cogiamanian, F., Marceglia, S., Ferrucci, R., Mameli, F., & Mrakic-Sposta, S. (2008). Improved naming after transcranial direct current stimulation in
aphasia. Journal of Neurology, Neurosurgery and Psychiatry, 79(4), 451e453.
Musso, M., Weiller, C., Kiebel, S., Mller, S. P., Blau, P., & Rijntjes, M. (1999). Training-induced brain plasticity in aphasia. Brain, 122(9), 1781e1790.
Naeser, M. A., Martin, P. I., Ho, M., Treglia, E., Kaplan, E., Bashir, S., et al. (2012). Transcranial magnetic stimulation and aphasia rehabilitation. Archives of
Physical Medicine and Rehabilitation, 93(1), S26eS34.
Naeser, M. A., Martin, P. I., Nicholas, M., Baker, E. H., Seekins, H., Kobayashi, M., et al. (2005). Improved picture naming in chronic aphasia after TMS to part of
right Broca's area: an open-protocol study. Brain and Language, 93(1), 95e105.
Nickels, L. (1997). Spoken word production and its breakdown in aphasia. Hove, UK: Psychology Press.
Nitsche, M. A., & Paulus, W. (2000). Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. The Journal of
Physiology, 527(3), 633e639.
Parkinson, R. B., Raymer, A., Chang, Y. L., FitzGerald, D. B., & Crosson, B. (2009). Lesion characteristics related to treatment improvement in object and action
naming for patients with chronic aphasia. Brain and Language, 110(2), 61e70.
Patterson, K., & Shewell, C. (1987). Speak and spell: dissociation and word-class effects. In M. Coltheart, R. Job, & G. Sartori (Eds.), The cognitive neuropsychology of language. Hove, UK: Erlbaum.
Peck, K. K., Moore, A. B., Crosson, B. A., Gaiefsky, M., Gopinath, K. S., White, K., et al. (2004). Functional magnetic resonance imaging before and after aphasia
therapy shifts in hemodynamic time to peak during an overt language task. Stroke, 35(2), 554e559.
Raboyeau, G., De Boissezon, X., Marie, N., Balduyck, S., Puel, M., Bezy, C., et al. (2008). Right hemisphere activation in recovery from aphasia lesion effect or
function recruitment? Neurology, 70(4), 290e298.
Rapp, B., & Goldrick, M. (2000). Discreteness and interactivity in spoken word production. Psychological Review, 107(3), 460e499.
Rosen, H. J., Petersen, S. E., Linenweber, M. R., Snyder, A. Z., White, D. A., Chapman, L., et al. (2000). Neural correlates of recovery from aphasia after damage
to left inferior frontal cortex. Neurology, 55(12), 1883e1894.
Ross, L. A., McCoy, D., Wolk, D. A., Coslett, H. B., & Olson, I. R. (2010). Improved proper name recall by electrical stimulation of the anterior temporal lobes.
Neuropsychologia, 48(12), 3671e3674.
Saur, D., & Hartwigsen, G. (2012). Neurobiology of language recovery after stroke: lessons from neuroimaging studies. Archives of Physical Medicine and
Rehabilitation, 93(1), S15eS25.
Saur, D., Kreher, B. W., Schnell, S., Kmmerer, D., Kellmeyer, P., Vry, M. S., et al. (2008). Ventral and dorsal pathways for language. Proceedings of the National
Academy of Sciences, 105(46), 18035e18040.
Saur, D., Lange, R., Baumgaertner, A., Schraknepper, V., Willmes, K., Rijntjes, M., et al. (2006). Dynamics of language reorganization after stroke. Brain, 129(6),
1371e1384.
Schlaug, G., Marchina, S., & Wan, Y. C. (2011). The use of non-invasive brain stimulation techniques to facilitate recovery from post-stroke aphasia.
Neuropsychology Review, 21(3), 288e301.
Shah-Basak, P. P., Norise, C., Garcia, G., Torres, J., Faseyitan, O., & Hamilton, R. H. (2015). Individualized treatment with transcranial direct current stimulation
in patients with chronic non-uent aphasia due to stroke. Frontiers in Human Neuroscience, 9, 1e12.
Shallice, T., Rumiati, R. I., & Zadini, A. (2000). The selective impairment of the phonological output buffer. Cognitive Neuropsychology, 17, 417e546.
Sparing, R., Dafotakis, M., Meister, I. G., Thirugnanasambandam, N., & Fink, G. R. (2008). Enhancing language performance with non-invasive brain stimulation: a transcranial direct current stimulation study in healthy humans. Neuropsychologia, 46(1), 261e268.
Thulborn, K. R., Carpenter, P. A., & Just, M. A. (1999). Plasticity of language-related brain function during recovery from stroke. Stroke, 30(4), 749e754.
Vestito, L., Rosellini, S., Mantero, M., & Bandini, F. (2014). Long-term effects of transcranial direct-current stimulation in chronic post-stroke aphasia: a pilot
study. Frontiers in Human Neuroscience, 8, 785.
Vines, B. W., Nair, D., & Schlaug, G. (2008). Modulating activity in the motor cortex affects performance for the two hands differently depending upon which
hemisphere is stimulated. European Journal of Neuroscience, 28(8), 1667e1673.
Weiller, C., Isensee, C., Rijntjes, M., Huber, W., Mller, S., Bier, D., et al. (1995). Recovery from Wernicke's aphasia: a positron emission tomographic study.
Annals of Neurology, 37(6), 723e732.