Clinical Neurophysiology 125 (2014) 1912–1922

Contents lists available at ScienceDirect

Clinical Neurophysiology

journal homepage: www.elsevier .com/locate /c l inph

Neurophysiologic markers in laryngeal muscles indicate functionalanatomy of laryngeal primary motor cortex and premotor cortexin the caudal opercular part of inferior frontal gyrus

http://dx.doi.org/10.1016/j.clinph.2014.01.0231388-2457/� 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved

⇑ Corresponding author at: Laboratory for Human and Experimental Neurophysiology (LAHEN), School of Medicine, University of Split, Šoltanska 2, 21000 SplitTel.: +385 91 754 1385.

E-mail address: vdeletis@chpnet.org (V. Deletis).1 These authors equally contributed.

Vedran Deletis a,b,⇑,1, Maja Rogic a,1, Isabel Fernández-Conejero c,1, Andreu Gabarrós c, Ana Jeroncic a,d

a Laboratory for Human and Experimental Neurophysiology (LAHEN), School of Medicine, University of Split, Split, Croatiab Department for Intraoperative Neurophysiology, Roosevelt Hospital, New York, NY, USAc University Hospital Bellvitge, Barcelona, Spaind Department for Research in Biomedicine and Health, School of Medicine, University of Split, Split, Croatia

a r t i c l e i n f o h i g h l i g h t s

Article history:Accepted 24 January 2014Available online 11 February 2014

Keywords:Premotor cortexInferior frontal gyrusBroca areaPrimary motor cortex for laryngeal musclesMotor speech areasNavigated transcranial magneticstimulationElectrical stimulation of motor speech areasNeurophysiologic markers

� This is a unique study describing methods for eliciting neurophysiologic markers of primary motorcortex (M1) for laryngeal muscles and premotor cortex of inferior frontal gyrus.

� The neurophysiologic markers were elicited by: (a) navigated transcranial magnetic stimulation(nTMS) in a group of healthy subjects, (b) direct cortical stimulation (DCS) of exposed cortex duringcraniotomy.

� These neurophysiologic markers indicate functional anatomy of M1 for laryngeal muscles and premo-tor cortex in the caudal opercular part of inferior frontal gyrus.

a b s t r a c t

Objective: The aim of this study was to identify neurophysiologic markers generated by primary motorand premotor cortex for laryngeal muscles, recorded from laryngeal muscle.Methods: Ten right-handed healthy subjects underwent navigated transcranial magnetic stimulation(nTMS) and 18 patients underwent direct cortical stimulation (DCS) over the left hemisphere, whilerecording neurophysiologic markers, short latency response (SLR) and long latency response (LLR) fromcricothyroid muscle. Both healthy subjects and patients were engaged in the visual object-naming task. Inhealthy subjects, the stimulation was time-locked at 10–300 ms after picture presentation while in thepatients it was at zero time.Results: The latency of SLR in healthy subjects was 12.66 ± 1.09 ms and in patients 12.67 ± 1.23 ms. Thelatency of LLR in healthy subjects was 58.5 ± 5.9 ms, while in patients 54.25 ± 3.69 ms. SLR elicited by thestimulation of M1 for laryngeal muscles corresponded to induced dysarthria, while LLR elicited by stim-ulation of the premotor cortex in the caudal opercular part of inferior frontal gyrus, recorded from laryn-geal muscle, corresponded to speech arrest in patients and speech arrest and/or language disturbances inhealthy subjects.Conclusion: In both groups, SLR indicated location of M1 for laryngeal muscles, and LLR location of pre-motor cortex in the caudal opercular part of inferior frontal gyrus, recorded from laryngeal muscle, whilestimulation of these areas in the dominant hemisphere induced transient speech disruptions.Significance: Described methodology can be used in preoperative mapping, and it is expected to facilitatesurgical planning and intraoperative mapping, preserving these areas from injuries.� 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights

reserved

, Croatia.

V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922 1913

1. Introduction

Mapping of primary motor cortex and Broca’s area duringawake craniotomy is not always the best option, especially inchildren and uncooperative patients; it would be important tofind a neurophysiologic methodology for intraoperative mappingand monitoring anatomic and functional integrity of these corti-cal areas. The development of a methodology for preoperativeidentification using navigated transcranial magnetic stimulation(nTMS) would facilitate surgical planning and intraoperativemapping.

The primary motor cortex (M1) for laryngeal muscles has a rolein execution and motor speech control, while the posterior inferiorfrontal gyrus, namely Broca’s area, is regarded as an important mo-tor speech cortical area having a role in all stages of word encodingand their unification (Sahin et al., 2009), as well as sending coded‘‘commands’’ to M1.

After electrical stimulation of Broca’s area, postsynaptic poten-tials of high amplitudes are recorded in the lateral part of the M1(Greenlee et al., 2004). These results indirectly indicate a functionalconnection of Broca’s area and M1 for face, mouth, pharynx, andlarynx. It has been proposed that the critical parts of M1 neededto control vocalization are closely and uniquely associated withthe laryngeal muscles (Corballis, 2003). The direct functional con-nectivity of M1 for laryngeal muscles was demonstrated in ourstudies (Deletis et al., 2008, 2009, 2011; Espadaler et al. 2012). Inthese studies, we have developed methodologies for stimulatingM1 for laryngeal muscles and recording corticobulbar motor-evoked potentials (MEPs) from vocal and cricothyroid muscles.Corticobulbar MEPs can be regarded as a synonym for short latencyresponse (SLR) representing a neurophysiologic marker of M1 forlaryngeal muscles. It has been also reported that long latency re-sponse (LLR) can be recorded from laryngeal muscles after magneticstimulation of the frontal cortex (Amassian et al., 1988; Ertekinet al., 2001). We also recorded LLRs intraoperatively but we neitherstudied this systematically nor investigated or speculated about itsexact origin (Amassian et al., 1988; Ertekin et al., 2001; Deletiset al., 2008, 2011).

The standard intraoperative neurophysiologic method for map-ping of Broca’s area and M1 for orofacial, pharyngeal, and laryn-geal muscles consists of electrical stimulation of these areas andinducing transient speech disruptions: (a) speech arrest and/orlanguage disturbances while stimulating Broca’s area and (b) dys-arthria while stimulating M1 for orofacial, pharyngeal, and laryn-geal muscles. Electrical stimulation with 50–60 Hz is used duringsurgery or through the subdural grid electrodes. In addition to 50-Hz stimulation for 1–3 s, speech arrest was also elicited with ashort train of stimuli technique (5 pulses, 3-ms interstimulusinterval) (Axelson et al., 2009). Furthermore, high-frequencyrepetitive transcranial magnetic stimulation with a repetition rateof 2–25 Hz was also used, but so far, only three groups of authorsreported induced speech arrest (Epstein et al., 1996, 1999;Pascual-Leone et al., 1991). Recently, Picht et al. (2013) showeda good overall correlation between repetitive navigated (nTMS)and direct cortical stimulation (DCS) during awake surgeryfor the identification of language-related areas in patients withleft-hemisphere lesions.

We hypothesized that time-locked electrical activity is recordedin laryngeal muscles after the stimulation of M1 for laryngeal mus-cles and of premotor cortex in the caudal opercular part of inferiorfrontal gyrus.

Therefore, the goal of this study was to investigate the relation-ship between cortical spots generating neurophysiologic markersrecorded in laryngeal muscle and the functional role of these cor-tical areas by clinically producing transient speech disruptions.

2. Methodology

2.1. Healthy subjects/patients

Ten right-handed healthy subjects, three male and sevenfemale, average age 31 ± 13.32 (range 22–66 years), and 18 right-handed patients, 10 male and eight female, average age46.2 ± 13.69, range 27–68 years, were included in the study. TheEdinburgh Inventory Questionnaire test (Oldfield, 1971) was usedfor assessment of handedness. All healthy subjects and patientssigned informed consent forms to participate in the study. Healthysubjects received a small honorarium. The study was approved bythe Ethical Committees of School of Medicine, University of Split,Croatia, and University Hospital Bellvitge, Barcelona, Spain.

2.2. Healthy subjects group

2.2.1. Magnetic resonance imaging (MRI) and navigated transcranialmagnetic stimulation (nTMS)

Magnetic resonance imaging (MRI) of the head for each subjectwas performed with Siemens Magnetom Avanto, Tim (76 � 18)strength 1.5 T. MRI images were recorded by specific MRI require-ments for nTMS-NBS (Navigated Brain Stimulation), Nexstim Sys-tem 4 (Helsinki, Finland), including the head, visible ears, andnose. MRI images are integrated in the nTMS machine with athree-dimensional navigation system display of the subjects’ brain.Sophisticated real-time data processing allows the precise displayof the induced electric field (E-field) within the brain tissue. Tar-geting tools available on-screen are the following: grid for system-atic brain mapping, targeting tool for optimal coil placement,aiming tool for precise repetition of given stimuli, and automatedstimulation (location controlled). The subject wears an opticalhead tracker and by using a pointer, 12 points are registered onthe subject’s scalp. An air-cooled eight-shaped figure coil was used,generating a biphasic pulse of 289 ls pulse length. The maximumE-field is 172 V/m below the Nexstim Focal coil in the sphericalconductor model representing the human head.

The nTMS mapping procedure is as follows:

1. Eliciting MEP resting threshold for hand muscle, abductorpollicis brevis (APB);

2. Mapping of the very lateral part of M1 and recording SLR incricothyroid muscle during the visual object-naming task;

3. Mapping of the inferior frontal gyrus and recording LLR incricothyroid muscle during the visual object-naming task;

4. Stimulation of cortical spot which elicited SLR to producetransient speech disruption during the visual object-namingtask;

5. Stimulation of cortical spot which elicited LLR to producetransient speech disruptions during the visual object-nam-ing task.

Mapping of the M1 for hand muscle is the standard method innTMS studies dealing with mapping of M1 (Epstein et al., 1999;Schmidt et al., 2009; Julkunen et al., 2011) and its vicinity (Jennumet al., 1994; Michelucci et al., 1994; Epstein et al., 1999; Stewartet al., 2001), and as a standard intraoperative procedure precedingmapping of primary motor cortex and Broca’s area (Duffau, 2008;Kim et al., 2009; Lubrano et al., 2010). Mapping over the left M1for APB was determined by the ‘‘omega knob’’ on axial MRI images,or a ‘‘hook structure’’ at the sagittal MRI (Yousry et al., 1997). Thecentral sulcus was also used as a landmark while moving the coil inthe anterior–posterior direction in order to map the hot spot forM1 for APB. The MEP resting threshold was defined as the loweststimulus intensity for eliciting at least five MEPs in the APB muscle

1914 V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922

with an amplitude of at least 50 lV in a series of 10 consecutivetrails. Mapping over the left M1 for laryngeal muscle wasperformed after mapping of M1 for APB muscle. The thresholdand location of M1 for laryngeal muscle was determined by record-ing corticobulbar MEPs from the cricothyroid muscle. Mappingover the left M1 for cricothyroid muscle was performed using intoaccount our recently reported data on Euclidian distance of25.19 ± 6.51 mm between the cortical representation of M1 forAPB and M1 for cricothyroid muscle, with cricothyroid muscle lat-eral to APB (Espadaler et al., 2012). The resting threshold of corti-cobulbar MEP was defined as the lowest stimulus intensity foreliciting at least five corticobulbar MEPs in cricothyroid musclewith an amplitude of 50 lV in a series of 10 consecutive trails ofTMS.

For mapping of the inferior part of frontal gyrus, the guidelinetopography of Broca’s area was used according to Ebeling et al.(1989). We also used as a guideline the distance between M1 forcricothyroid muscle and of inferior frontal gyrus generating tran-sient speech disruptions area of 17.23 ± 4.73 mm (Rogic et al.,2014). For mapping of the inferior frontal gyrus, the stimulationintensity was adjusted using the resting threshold for corticobul-bar MEPs in cricothyroid muscle. If this intensity was unpleasantto the subject, the stimulus intensity was reduced in steps of 1–2% of maximum stimulator output.

2.2.2. Stimulation parameters and speech task procedurePatterned bursts of magnetic stimuli were used, described in

our previous study (Rogic et al., 2014). Briefly, two paradigms ofpatterned repetitive stimulation were used:

The ‘‘first paradigm’’ consisted of four bursts of five stimulieach, 6 ms apart, burst repetition rate of 4 Hz.

The ‘‘second paradigm’’ consisted of 16 bursts of four stimulieach, 6 ms apart, burst repetition rate of 12 Hz.

All healthy subjects underwent two testing sessions separatedby 4 months.

First testing session. The first paradigm of stimulation was usedfor eliciting: (a) a neurophysiologic marker of M1 for laryngealmuscles obtained by mapping of the lateral part of M1 and record-ing of SLR from cricothyroid muscle and (b) a neurophysiologicmarker of premotor cortex of inferior frontal gyrus, recorded fromlaryngeal muscle, obtained by mapping of the caudal opercularpart of inferior frontal gyrus and recording of LLR from cricothyroidmuscle. Due to the slight jittering in LLR responses, the meanvalues of five consecutive latencies of LLRs have been calculated.

Table 1SLR and LLR parameters in a group of healthy subjects and the type of transient speech d

First testing session Second testing sessio

SLRSubjects Gender Age Latency

(ms)Intensity(%)

Intens(%)

1 F 29 11.57 38 Dysarthria 262 M 66 13.57 45 Dysarthria 373 F 27 12.53 34 Dysarthria 234 M 39 11.65 42 Dysarthria 305 F 27 12.54 34 Dysarthria 416 F 26 14.96 40 Dysarthria 407 M 30 11.9 40 Dysarthria 438 F 21 11.98 34 a a

9 F 23 13.7 37 Dysarthria 3510 F 22 12.2 39 Dysarthria 29Mean ± SD 31 ± 13.32 12.66 ± 1.09 38.3 ± 3.7 33.8 ±Range 22–66 11.57–14.96 34–45 23–43

No = transient language disruption elicited number of times.a Subject did not participate in the second testing session of the study.

Second testing session. For testing the functional role of corticalspots generating neurophysiologic markers in cricothyroid muscle,the second paradigm of stimulation was used in order to elicittransient speech disruptions while stimulating the cortical spotwhich elicited neurophysiologic markers.

Stimulation was applied while subjects were engaged in thevisual object-naming task. The methodology for visually presentedobjects (VPOs)/pictures were taken from a normative study byBrodeur et al. (2010). Fifty pictures were presented during mea-surement in a randomized manner in one session. The pictureswere presented on a computer monitor (LG 2200 ‘‘LCD’’ with1920 � 1080 resolution) using the Presentation program (Neuro-behavioral Systems, Albany, CA, USA). The picture was presentedon a white background while the edges of the screen were blackfor the photo sensor to indicate the start and end of the presentedpicture. VPO was presented for 3000 ms, followed by 2390 ms ofblank screen. The time from VPO until the presentation of the nextVPO was 5390 ms. The Presentation program triggered nTMS 0–350 ms after VPO. In the first testing session in eight subjects,the beginning of stimulation was 300 ms after VPO, in one subject350 ms, and in another 150 ms after VPO. In the second testing ses-sion, the Presentation program triggered nTMS at zero time withthe VPO.

In the first testing session, neurophysiologic markers wererecorded from cricothyroid muscle, while in the second testing ses-sion, stimulation of identical cortical spots generating markerswere repeated to elicit transient speech disruptions, without plac-ing electrodes in cricothyroid muscle. One subject (Table 1, No. 1)was an exception in whom in both testing sessions the electrodeswere placed in cricothyroid muscle to record neurophysiologicmarkers and speech-related activity. In this subject, the secondparadigm of stimulation was used to elicit neurophysiologic mark-ers and transient speech disruptions.

According to Abel et al. (2009), errors can happen incidentally inhealthy subjects during the visual object-naming task; therefore,before the nTMS stimulation session started, each subject partici-pated in a visual object-naming task in two control sessions. Inthe second control session, we discarded ‘‘problematic’’ pictures,and did not use them during the nTMS session.

Control measurement with nTMS was performed in four subjectswith the coil set away from the subject’s head during the visual ob-ject-naming task in the first testing session. This measurement wasperformed to exclude possible influence of micro-reflexes, espe-cially of sonomotor origin (Bickford et al. 1964; Bickford, 1966).

isruption by stimulating cortical spot which elicited SLR in cricothyroid muscle.

n First testing session Second testing session

LLRity Latency

(ms)Intensity(%)

Speecharrest (No)

Languagedisturbances (No)

Intensity(%)

64.0 31 3 8 2661.1 45 0 1 3758.9 36 1 4 2357.6 40 1 1 3051.4 30 1 3 4150.4 40 1 3 4067.7 40 2 0 4351.5 32 a a a

58.7 36 3 3 3563.7 37 0 2 29

7.08 58.5 ± 5.9 36.7 ± 4.7 33.8 ± 7.0850.4–67.7 30–45 23–43

V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922 1915

2.2.3. RecordingsThe recordings of neurophysiologic markers from cricothyroid

muscles were done using a five-channel EMG amplifier integratedwith E-field navigation and TMS stimulation of Nexstim NBS sys-tem (Nexstim NBS System 4, Helsinki, Finland). The channels are:(a) EMG from APB muscle, (b) EMG from cricothyroid muscle, (c)beginning and end of VPO, (d) stimulator trigger, and e) micro-phone recordings.

Surface electrodes Neuroline 720 (Ambu Inc, Glen Burnie, MD,USA) were placed over the right APB muscle. The method for thepercutaneous placement of electrodes into cricothyroid musclehas been described in our previous studies (Deletis et al., 2011;Espadaler et al., 2012). Briefly described, for recording SLR andLLR from right cricothyroid muscle, two hook-wire electrodes wereused, each consisting of a Teflon-coated stainless steel wire, 76 lmin diameter passing through 27-gauge needles (0.4 mm) and of13 mm length (hook-wire electrodes, specially modified, ViasysHealthcare WI, MA, USA). The recording wires have a stripped Tef-lon isolation of 2 mm at their tip and curved to form the hook foranchoring them.

In the second testing session, modified methods of video andaudio recordings (Panasonic HDC-SDT750) (Lioumis et al., 2012)were taken for off-line analysis.

2.3. Patient group

2.3.1. Intraoperative DCSThe patients presented with pathology in the left hemisphere

(Table 2), located in or nearby the eloquent cortical areas. A mono-polar handheld probe was used to stimulate the exposed cortexwith a stimulation paradigm of bursts (monopolar stimulatorinstrument dissector 1.2 mm curved 30�, INOMED, Germany),while a bipolar electrode for 50 Hz stimulation. The electrodeswere placed in the cricothyroid muscle and recordings were madeduring all mapping procedures for M1 for laryngeal muscles andpremotor cortex of inferior frontal gyrus.

2.3.2. DCS procedureThe procedure includes the following:

1. Mapping of somatosensory cortex by using phase reversal med-ian somatosensory-evoked potentials

2. Mapping of M1 for hand, leg, and facial muscles to producemuscle contractions

Table 2Latencies of SLR and LLR recorded in cricothyroid muscle in a group of patients.

Patients Gender Age Pathology

1 M 59 Left frontal GB2 M 67 Left frontal M13 F 68 Left temporal G4 F 32 Left frontal glio5 M 43 Left temporal G6 M 35 Left frontal cav7 M 60 Left frontal M18 M 38 Left frontal cav9 F 26 Left parietal gl

10 F 39 Left fronto-tem11 F 55 Left frontal olig12 F 27 Left frontal MA13 M 59 Left frontal GB14 M 43 Left fronto-tem15 F 34 Left temporal c16 M 61 Left front-oper17 F 38 Left temporal-i18 M 48 Left parietal caMean ± SD 46.22 ± 13.69Range 27–68

3. Mapping of the inferior frontal gyrus in order to produce speecharrest

4. Positive spots from mapping step 2 were repeated for elicitingMEPs in hand, leg, and facial muscles – very lateral part of M1was also mapped to elicit SLR in cricothyroid muscle

5. Positive spots from mapping step 3 were repeated for elicitingLLR of premotor cortex in the caudal opercular part of inferiorfrontal gyrus, recorded from laryngeal muscle.

2.3.3. Stimulation parameters and speech task procedureAll patients underwent electrical stimulation of the exposed

cortex during awake craniotomy using the following stimulatingparameters:

a) Stimulation paradigm of 50 Hz, biphasic pulses of 4 msduration, 3 s of train, intensity up to 15 mA for mappingof M1 for upper (hand muscles) and lower extremities(leg muscles) and facial muscles in order to produce musclecontractions were applied. The same paradigm of stimula-tion was applied in order to elicit speech arrest whilemapping the inferior frontal gyrus. During surgery afteridentification of the Rolandic cortex, a series of line-draw-ing images were presented on the computer screen to thepatients as a visual object-naming task. Images weretime-locked at zero time with the electrical stimulation.Before the surgery, patients were acquainted with thevisual object-naming task and problematic pictures wereexcluded from the intraoperative testing. The patient wasinstructed to name the object in combination with thebeginning sentence: ‘‘This is a...’’ during cortical stimula-tion. All cortical sites were stimulated three times as astandard procedure during intraoperative mapping of thesecortical areas. The cortical spot was considered ‘‘positive’’ ifDCS elicited speech arrest in at least two of three timesafter testing of identical point.

b) Stimulation paradigm of bursts consists of four monophasicpulses of 0.5 ms duration each, 4 ms apart, burst repetitionrate of 4 Hz, intensity up to 15 mA for eliciting SLR, andLLR recorded from cricothyroid muscle. Stimulation wasapplied while subjects were in rest (not engaged in thevisual object-naming task). Due to the slight jittering inLLR responses, the mean values of five consecutive latenciesof LLRs have been calculated.

SLR latency (ms) LLR latency (ms)

M 10.49 55.212.2 53

BM 12.6 56.76ma 12.1 60.35BM 11 51.45

ernoma 12.35 58.712.95 59.05

ernoma 14.23 58.7ioma 12.15 50.15poral astrocitoma 15.7 50.05odendroglioma 14.35 52.45V 13.45 55.1

M 12.9 50.05poral astrocitoma 13.45 50.15avernoma 12.9 53.95cular glioma 11.6 50.15nsular astrocitoma 11.65 58.95vernoma 11.6 52.45

12.67 ± 1.23 54.25 ± 3.6910.49–15.7 50.05–60.35

1916 V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922

In the patient group, the electrodes were placed in the cricothy-roid muscle and recordings were made during all mappingprocedures.

2.3.4. RecordingsA pair of twisted subdermal needles (Neuroline-Ambu,

0.500 � 27 G) were placed in: orbicularis oris, extensor digitorumcommunis and APB, tibialis anterior, and abductor hallucis. Themethod for the percutaneous placement of electrodes into crico-thyroid muscle was identical to that described in the group ofhealthy subjects. In the patient group, we used a modified Synergy(Viasys Healthcare) 10-channel EMG-evoked potential unit.

The team consisted of a neurophysiologist and neuropsychol-ogist observing and documenting elicited transient speechdisruptions.

Fig. 1. Upper: Subject during naming visually presented object (VPO) and examiner hpremotor cortex in the caudal opercular part of inferior frontal gyrus. (Legend: 1 = microsubject with the attached photo sensor, 3 = microphone connected with the video camstimulation site, 6 = cloned monitor 5 for video shooting, and 7 = cloned monitor 2 for vprimary motor cortex for laryngeal muscles and premotor cortex in the caudal opercular plaryngeal muscle; (A) repeatability of SLRs recorded from cricothyroid muscle, (B) superrecorded from cricothyroid muscle, and (D) superimposed LLRs. The beginning of recormiddle: 3D MRI of subject’s brain with the localization of the stimulation.

2.4. Anesthesia

All patients underwent a standard technique of asleep–awake–asleep anesthesia, with conscious sedation using propofol andremifentanil, and assisted ventilation with laryngeal mask (LMA).After craniotomy was performed, the patients were graduallyawakened by weaning their sedation and LMA was removed, andreinserted after completion of the mapping procedure.

3. Results

3.1. Short latency response

In all healthy subjects and patients, SLR was successfully re-corded from cricothyroid muscle while stimulating the very lateral

olding the coil on the dominant hemisphere with stimulation localized over thephone connected to EMG amplifier, 2 = monitor with visual object presented to theera, 4 = magnetic coil for nTMS, 5 = monitor with MRI for precise determination ofideo shooting). Lower: Neurophysiologic markers indicating functional anatomy ofart of inferior frontal gyrus shown for one healthy subject (No. 3). Left: SLR of M1 forimposed SLRs. Right: LLR of opercular part of Broca’s area, (C) repeatability of LLRs

ding is set on the zero time which corresponds to time of 300 ms after VPO. In the

Fig. 2. Upper: Intraoperative mapping during awake craniotomy. The patient and the surgeon are separated with surgical drapes and can communicate with each other. Thepatient is engaged in the visual object-naming task. Lower: The results of cortical mapping. Stimulation of the cortical spot marked with red marker 4 elicited SLR, while spotmarked with yellow marker 1 elicited LLR in cricothyroid muscle. From the same spot where LLR was elicited, speech arrest was induced for Spanish. (T = tumor location; Bluemarkers 1–7 = somatosensory cortex; Red markers 1–4 = primary motor cortex; English, Catalan, and Spanish flags = cortical spots for speech arrest for these languages; fromthese cortical spots LLR was not elicited in the cricothyroid muscle). (For interpretation of the references to colour in this figure legend, the reader is referred to the webversion of this article.)

V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922 1917

part of M1. The latency of SLR elicited by magnetic stimulation inhealthy subjects was 12.66 ± 1.09 ms (Table 1 and Fig. 1), and byelectrical stimulation in patients 12.67 ± 1.23 ms (Table 2, Fig. 2).The functional role of this area was determined by stimulation ofcortical spot which elicited SLR and clinically producing dysarthria(Table 1 and Table 2), accompanied by contractions of facial andmasticatory muscles.

Long Latency Response. In all healthy subjects and patients,LLR was successfully recorded from cricothyroid muscle whilemapping the caudal opercular part of inferior frontal gyrus. Thelatency of LLR elicited by magnetic stimulation in healthy subjectswas 58.5 ± 5.9 ms (Table 1, Fig. 1), while by electrical stimulationin patients it was 54.25 ± 3.69 ms (Table 2, Fig. 2).

The functional role of this area was tested by stimulation of cor-tical spot that elicited LLR and clinically produced transient speechdisruptions (Table 1 and Table 2). Intraoperatively, speech arrestwas elicited in all patients, while in healthy subjects speech arrestand/or language disturbances were elicited (Table 1).

Speech arrest was elicited in seven healthy subjects (77.8%). Inthose seven healthy subjects in whom speech arrest was inducedonce, twice, or more times, language disturbances (semantic errors)were also induced stimulating the same cortical spot. In two sub-jects (22.2%) (No. 2 and No. 10), language disturbances (semantic er-rors) (Table 1) were elicited (kruška (pear)/instead/groz -de (grapes)/,etc.). In healthy subjects, speech arrest was characterized as a com-plete cessation of speech, confirmed by absent microphone record-ings together with an off-line analysis of video recordings showingno facial and masticatory muscle contractions. Semantic errors(subjects gave semantically similar word for the object) were alsoconfirmed by an off-line analysis of video recordings.

In healthy subjects, when M1 for cricothyroid muscle and infe-rior frontal gyrus were mapped, the electrical activity was recordedin cricothyroid muscle in the pre-speech (pre-EMG period) andspeech execution period (contractions related to the speech). TheSLRs and LLRs are elicited in cricothyroid muscle and can be de-tected in the period before the pre-speech phase. By stimulatingthe same LLR cortical spot which elicited speech arrest, semanticparaphasias could also be elicited with LLR recordings beforepre-speech, and with electrical activity in pre-speech and speechexecution period. When semantic paraphasias were elicited fromother parts of inferior frontal gyrus (not LLR cortical spot), LLRwas not recorded in cricothyroid muscle; only electrical activityin the pre-speech and speech execution phase was recorded incricothyroid muscle.

In one healthy subject (Table 1, No. 1), in whom speech arrestwas elicited while simultaneously recording activity from cricothy-roid muscle, LLRs were present in EMG, but voice was absent inmicrophone recordings (Fig. 3). The subject did not give any re-sponse for object presenting/vješalica (coat hanger)/which wasconfirmed with absent activity recorded with the microphone.The off-line video analysis also showed absent activity in facialand masticatory muscles. The latency of LLRs, indicated with thered stars in Fig. 3 was 57.6 ± 0.71 ms.

LLR was elicited in all healthy subjects 423.50 ± 115.31 ms fromthe beginning of the visual object presentation. The speech onsetreaction time was 684.07 ± 91.48 ms. The repeatability of elicitingLLR when mapping the caudal opercular part of inferior frontalgyrus is shown for one healthy subject No. 3 in Fig. 4.

In the patient group, M1 was mapped first applying electricalstimulation with 50 Hz in order to elicit tonic contractions in the

Fig. 3. Eliciting LLRs in cricothyroid muscle during speech arrest. Upper: Absent activity in the microphone recordings during speech arrest, while LLRs were elicited incricothyroid muscle (within the frame). Lower: Recordings in square (indicated with the red star) represent repeatability of four consecutive LLR, superimposed at the bottom.Scale in seconds, while amplitude in mV. Activity at the beginning of each trace represents the tail from the previous response. CTHY = cricothyroid muscle; VON = visualobject naming; arrow indicates beginning of visual presented object; MIC = microphone recordings; STIM = stimulation; artifacts from 16 bursts of stimuli are shown acrossall channels.

1918 V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922

orofacial muscles during rest and dysarthric speech. With the samestimulation paradigm, inferior frontal cortex was mapped, andonly a specific area in the caudal opercular part of inferior frontalgyrus, which elicited speech arrest, also elicited LLR from cricothy-roid muscle (Fig. 2). Stimulation of larger and more rostral parts ofthe opercular inferior frontal gyrus elicited speech arrest orparaphasia without eliciting LLR in cricothyroid muscle (Figs. 2and 5).

4. Discussion

For the first time, we have shown that it is possible to elicit neu-rophysiologic markers of M1 and premotor cortex in the caudal

opercular part of inferior frontal gyrus for laryngeal muscles,applying a specific paradigm of stimulation. To elicit SLR and LLR,a patterned burst of magnetic stimuli were used in healthy sub-jects (Rogic et al., 2014), while a short train of electrical stimuliwas used in patients. The functional role of these areas was testedby stimulation of cortical spot that elicited SLR and LLR and clini-cally producing transient speech disruptions, respectively.

SLR can be regarded as a neurophysiologic marker of M1 for lar-yngeal muscles, according to the following evidence:

1) SLRs are elicited from the vocal and cricothyroid musclesduring electrical and magnetic stimulation of the M1 for lar-yngeal muscles in a group of patients and healthy subjects

Fig. 4. Eliciting LLR when mapping premotor cortex in the caudal opercular part of inferior frontal in healthy subject No. 3. Pink star indicates LLR elicited in cricothyroidmuscle before speech execution. The green star indicates the beginning of the microphone recordings of speech. The first channel: stimulation packages; the second channel:recordings of electrical activity from cricothyroid muscle; and the third channel: microphone recordings.

V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922 1919

(Deletis et al., 2008, 2009, 2011; Espadaler et al., 2012). SLRcorresponds/equals to corticobulbar MEP.

2) The results of this study have shown that in all healthy sub-jects and patients, nTMS and DCS of cortical spot, which elic-ited SLR, clinically produced dysarthria.

3) M1 for laryngeal muscles is anatomically represented at themost lateral part of M1. Recently, we have reported that thedistance between the cortical representation of M1 for APBand M1 for cricothyroid muscle was 25.19 ± 6.51 mm, withcricothyroid muscle lateral to APB (Espadaler et al., 2012).

During speech disruptions induced by stimulation of SLR corti-cal spot, dysarthric speech was observed accompanied with con-tractions of the orofacial and masticatory muscles visible on thevideo recordings. Contractions of orofacial and masticatory mus-cles were presumably produced due to the spread of the current

from primary motor cortex for laryngeal muscles to the nearby pri-mary motor cortical representations for other cranial muscles.While the results demonstrate the SLR spot’s role in speech execu-tion, it probably also has functional roles in other voluntary laryn-geal movements that we did not test, such as coughing, throatclearing, swallowing, crying, moaning, or humming.

The following are evidence for LLR being the most likely neuro-physiologic marker of premotor cortex for laryngeal muscles:

1) The results of this study have shown that DCS and nTMS ofthe stimulated cortical spot, which elicited LLR, clinically pro-duced speech arrest in all patients and speech arrest and/orlanguage disturbances in healthy subjects. Anatomically, thiscortical spot corresponds to a small region of the caudal oper-cular inferior frontal gyrus detectable on an MRI of healthysubjects and in the patients by intraoperative inspection.

Fig. 5. Mapping results for M1 and premotor cortex in the caudal opercular part ofinferior frontal gyrus for single patient. (1) Markers indicate mapping results forpremotor cortex in the caudal opercular part of inferior frontal gyrus. Green marksindicate produced language disturbances (semantic and phonological paraphasias).Spanish flags/yellow marks are spots where speech arrest could be elicited. Onlyspot with Spanish flag 1 (yellow mark 1) elicited LLR in cricothyroid muscle; (2) redmarks 1–6 primary motor cortex; (3) blue marks 1–7 primary somatosensorycortex; (4) Results for mapping temporal area (specific language disturbances); and(5) T = tumor location. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)

1920 V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922

2) After electrical stimulation of Broca’s area, postsynapticpotentials of high amplitude were recorded in the lateralpart of the M1 (Greenlee et al., 2004). These data can beregarded as indirect evidence of an anatomical connectionbetween the premotor cortex in the caudal opercular partof inferior frontal gyrus and M1 for laryngeal muscle. Thereare also possible contributions of other pathways to LLR ori-gin, other than through M1.

LLR was recorded in cricothyroid muscle when speech arrestwas elicited (Fig. 3) in one healthy subject since electrodes wereinserted in the second session only in that subject. By stimulatingthe same LLR cortical spot which elicited speech arrest, semanticparaphasias could also be elicited with LLR recordings beforepre-speech, and with electrical activity in pre-speech and speechexecution period. Stimulus spread to nearby Broca’s area languagecortex likely explains these observations. When semantic parapha-sias were elicited by stimulating other parts of the dominantopercular inferior frontal gyrus (not LLR cortical spot), LLR wasnot recorded in cricothyroid muscle; only electrical activity inpre-speech and speech execution phase was recorded in cricothy-roid muscle. Dysphasia due to Broca’s area language cortex disrup-tion likely explains this kind of speech arrest without elicited LLR.Furthermore, even though we did not test the nondominant hemi-sphere, it is possible that cortical spot eliciting LLR by stimulationof nondominant hemisphere is likely part of the relays for emotiveexpression (e.g., prosody and singing). What functional role the LLRspot of either hemisphere may play in other voluntary laryngealmovements, such as humming, etc., is unknown because we testedonly speech.

In the patient group, electrical stimulation with 50 Hz was firstapplied in order to elicit contractions in the orofacial muscles dur-ing rest and dysarthric speech when mapping M1, and to elicitspeech arrest and/or language disturbances when mapping thedominant inferior frontal gyrus. Dysarthric speech was presumablyaccompanied with dysphonic/aphonic speech as well.

Speech arrest and/or phonological or semantic paraphasiaswere elicited by mapping specific parts of premotor cortex in thecaudal opercular part of inferior frontal gyrus, and tonic

contractions could be elicited from cricothyroid muscle. After that,bursts consisting of four stimuli each, 4 ms apart, burst repetitionrate of 4 Hz, were applied to the cortical spots which have previ-ously elicited speech arrest. Cortical spots which elicit semanticor phonological paraphasias and/or speech arrest without LLRrecording in cricothyroid muscle when mapping the dominantinferior frontal gyrus can be regarded as Broca’s area language cor-tex (Figs. 2 and 5). We suggest that dysphasia/aphasia might belikely contributing mechanism to semantic or phonological para-phasias and/or speech arrest without LLR recording in cricothyroidmuscle.

The spot in the caudal opercular part of dominant inferior fron-tal gyrus generating LLR and speech arrest after stimulation can beconsidered only as premotor cortex for laryngeal muscle, but not asmotor language cortex. This spot cannot be a safe resection margin,but an orientation spot for close proximity to the language pro-ducer Broca’s area.

We believe that we are dealing with small neuronal poolswhere language as a mental process gets transformed into a codedmessage to the M1 in order to activate muscles involved in speech.It is a matter of fact that no methodology, so far, to look at the elec-trical activity of laryngeal muscles during speech arrest has beenpublished.

During mapping procedures for eliciting neurophysiologicmarkers, only a very limited area, approximately 5 � 5 mm ofstimulated cerebral cortices, by either nTMS or DCS (Deletiset al., 2008), generates specific neurophysiologic markers record-able in the laryngeal muscles. Therefore, we consider these mark-ers as anatomical and functional indicators of primary motor andpremotor cortex for laryngeal muscles. In our previous studies(Deletis et al., 2008, 2011), we have also showed the possibilityto elicit SLR and LLR in laryngeal muscles by transcranial electricalstimulation (TES) of the right hemisphere with the TES montage C3versus Cz and C4 versus Cz. TES stimulating parameters were ashort train of electric stimuli consisting of three to five stimuli hav-ing a duration of 0.5 ms each. These stimuli were spaced 2 ms apartwith a train repetition rate of 2 Hz and intensity up to 120 mA.

Thus, neither spot can be specific to language or speech and inorder to use the LLR as a guide to nearby Broca’s area language cor-tex, one must know through other means that the dominant hemi-sphere is being stimulated

As we have mentioned earlier in the methodological part, wemade an effort to exclude micro-reflexes of Bickford (Bickfordet al., 1964; Bickford, 1966) as a potential source of SLR or LLR inthe nTMS measurement of healthy subjects. The sonomotor mi-cro-reflex can be activated via corticoreticular neurons as has beenalready shown (Fisher et al., 2012). We excluded micro-reflexes bymoving the coil away from the head during the stimulation. By thisapproach, SLR and LLR were not recorded. Furthermore, if SLR andLLR are micro-reflexes due to TMS, or activation of the auditorysystem by click produced by magnetic coil, micro-reflex wouldbe activated by the coil away from the head. We have shown thatis not the case. We have presented further evidence that SLR andLLR are cortical origins and specific for the particular region inour data. SLR can only be elicited if M1 for laryngeal muscles isstimulated and LLR can only be elicited if premotor cortex in thecaudal opercular part of inferior frontal gyrus is stimulated.

Our explanation of the mechanism of SLR and LLR generationduring speech execution is as follows: when the coded signal fromthe premotor cortex in the caudal opercular part of inferior frontalgyrus reaches the M1 motoneurons of the muscles involved inspeech, it gets transmitted further via the corticobulbar pathwayto the motor neurons of the cranial nerves in the brain stem. Fromthe motoneurons of the cranial nerves and via cranial motor nerves,targeted muscles get activated. If M1 for laryngeal muscles andpremotor cortex in the caudal opercular part of inferior frontal

V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922 1921

gyrus are stimulated during their ‘‘working state,’’ when their excit-ability is increased, we can easily induce synchronized activity oftheir neurons, and record this activity in the laryngeal muscles asSLR or LLR, depending on the neural structure being stimulated.Increased excitability of Broca’s area during motor speech process-ing has been shown in a temporal domain of 300 ms during theexecution of specific speech tasks. This was shown by an invasivestudy of recording EPSPs from Broca’s area in epileptic patients(Sahin et al., 2009), and noninvasively (Schuhmann et al., 2009,2012; Rogic et al., 2014). In this study, for most of the healthysubjects the delay of 300 ms for triggering of nTMS after picturepresentation has been shown to be optimal for eliciting SLR and LLR.

Presumably, more synapses are involved when stimulatingpremotor cortex in the caudal opercular part of inferior frontalgyrus, compared to M1 for laryngeal muscles; therefore, the latencyof LLR is longer than for SLR. Not only latency difference but alsojittering is more pronounced in LLR than SLR. After analysis ofresponse latencies of LLR elicited by magnetic stimulation inhealthy subjects and by electric stimulation in patients, we foundout that LLR have longer latencies when elicited by magnetic versuselectric stimulation. The LLR latencies elicited by magnetic versuselectric stimulation were 58.5 ± 5.9 and 54.25 ± 3.69 ms. Further-more, it was easier to elicit both SLR and LLR electrically thanmagnetically. This was the reason why we needed to introduce atime-locked visual object-naming task when applying nTMS inhealthy subjects to facilitate eliciting of SLR and LLR. Intraopera-tively, patients were in the rest state when eliciting SLR and LLR.

4. Conclusion

In all healthy subjects and patients, neurophysiologic markersof specific cortical areas are identified: (a) a neurophysiologic mar-ker of M1 for laryngeal muscles obtained by mapping of the lateralpart of M1 and recording of SLR from cricothyroid muscle and (b) aneurophysiologic marker of premotor cortex for laryngeal musclesobtained by mapping of the caudal opercular part of inferior frontalgyrus and recording of LLR from cricothyroid muscle. Stimulationof the SLR spot clinically produced dysarthria (and probably alsodysphonia). Stimulation of the dominant hemisphere’s LLR spotproduced speech arrest likely by disrupting phonation and, in somehealthy controls, produced language disturbances likely byunavoidably also exciting nearby Broca’s area language cortex.LLR cortical spot in the dominant hemisphere may be a usefulguide to the nearby proximity of Broca’s area language cortex. Itmight also help approximate the likely nearby location of Broca’sarea language cortex for patients unable to tolerate language map-ping under local anesthesia. Further studies are needed on thesame group of patients undergoing nTMS and DCS in order to rep-licate the data obtained in this study. The nondominant hemi-sphere should also be tested on the group of healthy subjectsand patients, since it is possible that nondominant SLR or LLRmight disrupt speech execution, which is not specific to the dom-inant hemisphere.

In addition, the possible functional role of the SLR and LLR spotsin executing voluntary phonations and laryngeal movements otherthan speech could be evaluated by specific testing.

5. Significance

Novel methodology for testing the anatomic and functionalintegrity of primary motor and premotor cortex for laryngeal mus-cles can be used in preoperative and intraoperative mapping. Giventhe close proximity of laryngeal premotor cortex to Broca’s lan-guage cortex in the dominant hemisphere, the methodology could

facilitate or compliment surgical planning and intraoperative map-ping intended to help avoid motor speech deficits.

References

Abel S, Dressel K, Kümmerer D, Saur D, Mader I, Weiller C, et al. Correct anderroneous picture naming responses in healthy subjects. Neurosci. Lett.2009;463:167–71.

Amassian VE, Anziska BJ, Cracco JB, Cracco RQ, Maccabee PJ. Focal magneticexcitation of frontal cortex activates laryngeal muscles in man [Abstract]. J.Physiol. (Lond.) 1988;398:41P.

Axelson HW, Hesselarger G, Flink R. Successful localization of the Broca area withshort-train pulses instead of ‘Penfield’ stimulation. Seizure 2009:374–5.

Bickford RG, Jacobson JL, Cody DTR. Nature of average evoked potentials to soundand other stimuli in man. Ann. N. Y. Acad. Sci. 1964;112:204–18.

Bickford RG. Human ‘‘microreflexes’’ revealed by computer analysis. Neurology1966;16:302.

Brodeur MB, Dionne-Dostie E, Montreuil T, Lepage M. The BOSS, a new set of 538normalized photos of objects to be used as ecological stimuli in vision andmemory paradigms. PLoS One 2010;24(5):e10773.

Corballis MC. From mouth to hand: gesture, speech, and the evolution of right-handedness. Behav. Brain Sci. 2003;26:199–260.

Deletis V, Ulkatan S, Cioni B, Meglio M, Colicchio G, Amassian V, et al. Responseselicited in the vocalis muscles after electrical stimulation of motor speech areas.Riv. Med. 2008;14:159–65.

Deletis V, Fernandez-Conejero I, Ulkatan S, Costantino P. Methodology forintraoperatively eliciting motor evoked potentials in the vocal muscles byelectrical stimulation of the corticobulbar tract. Clin. Neurophysiol.2009;120:336–41.

Deletis V, Fernández-Conejero I, Ulkatan S, Rogic M, Carbó EL, Hiltzik D.Methodology for intraoperative recording of the corticobulbar motor evokedpotentials from cricothyroid muscles. Clin. Neurophysiol. 2011;122:1883–9.

Duffau H. Intraoperative neurophysiology during surgery for cerebral tumors. In:Nuwer MR, editor. Handbook of clinical neurophysiology. Intraoperativemonitoring of neural function 2008;vol. 8. Amsterdam: Elsevier; 2008. p.491–507.

Ebeling U, Steinmetz H, Huang Y, Kahn T. Topography and identification of theinferior precentral sulcus in MR imaging. Am. J. Roentgenol. 1989;153:1051–6.

Epstein CM, Lah JJ, Meador K, Weissman JD, Gaitan LE, Dihenia B. Optimum stimulusparameters for lateralized suppression of speech with magnetic brainstimulation. Neurology 1996;47:1590–3.

Epstein CM, Meador KJ, Loring DW, Wright RJ, Weissman JD, Sheppard S, et al.Localization and characterization of speech arrest during transcranial magneticstimulation. Clin. Neurophysiol. 1999;110:1073–9.

Ertekin C, Turman B, Tarlaci S, Celik M, Aydogdu I, Secil Y, et al. Cricopharyngealsphincter muscle responses to transcranial magnetic stimulation in normalsubjects and in patients with dysphagia. Clin. Neurophysiol. 2001;112:86–94.

Espadaler J, Rogic M, Deletis V, Leon A, Quijada C, Conesa G. Representation ofcricothyroid muscles at the primary motor cortex (M1) in healthy subjects,mapped by navigated transcranial magnetic stimulation (nTMS). Clin.Neurophysiol. 2012;123:2205–11.

Fisher KM, Zaaimi B, Baker SN. Reticular formation responses to magnetic brainstimulation of primary motor cortex. J. Physiol. 2012;590:4045–60.

Greenlee JD, Oya H, Kawasaki H, Volkov IO, Kaufman OP, Kovach C, et al. Afunctional connection between inferior frontal gyrus and orofacial motor cortexin human. J. Neurophysiol. 2004;92:1153–64.

Jennum P, Friberg L, Fuglsang-Frederiksen A, Dam M. Speech localization usingrepetitive transcranial magnetic stimulation. Neurology 1994;44:269–73.

Julkunen P, Ruohonen J, Sääskilahti S, Säisänen L, Karhu J. Threshold curves fortranscranial magnetic stimulation to improve reliability of motor pathwaystatus assessment. Clin. Neurophysiol. 2011;122:975–83.

Kim SS, McCutcheon IE, Suki D, Weinberg JS, Sawaya R, Lang FF, et al. Awakecraniotomy for brain tumors near eloquent cortex: correlation of intraoperativecortical mapping with neurological outcomes in 309 consecutive patients.Neurosurgery 2009;64:836–45. discussion 345-6.

Lioumis P, Zhdanov A, Mäkelä N, Lehtinen H, Wilenius J, Neuvonen T, et al. A novelapproach for documenting naming errors induced by navigated transcranialmagnetic stimulation. J. Neurosci. Methods 2012;15(204):349–54.

Lubrano V, Draper L, Roux FE. What makes surgical tumor resection feasible inBroca’s area? Insights into intraoperative brain mapping. Neurosurgery2010;66:868–75. discussion 875.

Michelucci R, Valzania F, Passarelli D, Santangelo M, Rizzi R, Buzzi AM, et al. Rapid-rate transcranial magnetic stimulation and hemispheric language dominance.Neurology 1994;44:1697–700.

Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory.Neuropsychologia 1971;9:97–113.

Pascual-Leone A, Gates JR, Dhuana A. Induction of speech arrest and counting errorswith rapid-rate transcranial magnetic stimulation. Neurology1991;41:697–702.

Picht T, Krieg SM, Sollmann N, Rösler J, Niraula B, Neuvonen T, et al. A comparison oflanguage mapping by preoperative navigated transcranial magnetic stimulationand direct cortical stimulation during awake surgery. Neurosurgery2013;72:808–19.

Rogic M, Deletis V, Fernandez-Conejero I. Inducing transient language disruptionsby mapping of Broca’s area with modified patterned repetitive transcranial

1922 V. Deletis et al. / Clinical Neurophysiology 125 (2014) 1912–1922

magnetic stimulation protocol. J. Neurosurg. 2014. http://dx.doi.org/10.3171/2013.11.JNS13952.

Sahin NT, Pinker S, Cash SS, Schomer D, Halgren E. Sequential processing of lexical,grammatical, and phonological information within Broca’s area. Science2009;326:445–9.

Schmidt S, Cichy RM, Kraft A, Brocke J, Irlbacher K, Brandt SA. An initial transient-state and reliable measures of corticospinal excitability in TMS studies. Clin.Neurophysiol. 2009;120:987–93.

Schuhmann T, Schiller NO, Goebel R, Sack AT. The temporal characteristics offunctional activation in Broca’s area during overt picture naming. Cortex2009;45:1111–6.

Schuhmann T, Schiller NO, Goebel R, Sack AT. Speaking of which: dissecting theneurocognitive network of language production in picture naming. Cereb.Cortex 2012;22:701–9.

Stewart L, Walsh V, Frith U, Rothwell JC. TMS produces two dissociable types ofspeech disruption. NeuroImage 2001;13:472–8.

Yousry T, Schmid UD, Alkadhi H, Schmidt D, Peraud A, Buettner A, et al. Localizationof the motor hand area to a knob on the precentral gyrus-A new landmark.Brain 1997;120:141–57.