Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Results on the spatial resolution of repetitive transcranial magnetic stimulation for cortical language mapping during object naming in healthy subjects

Results on the spatial resolution of repetitive transcranial magnetic stimulation for cortical... Background: The spatial resolution of repetitive navigated transcranial magnetic stimulation (rTMS) for language mapping is largely unknown. Thus, to determine a minimum spatial resolution of rTMS for language mapping, we evaluated the mapping sessions derived from 19 healthy volunteers for cortical hotspots of no‑ response errors. Then, the distances between hotspots (stimulation points with a high error rate) and adjacent mapping points (stimulation points with low error rates) were evaluated. Results: Mean distance values of 13.8 ± 6.4 mm (from hotspots to ventral points, range 0.7–30.7 mm), 10.8 ± 4.8 mm (from hotspots to dorsal points, range 2.0–26.5 mm), 16.6 ± 4.8 mm (from hotspots to apical points, range 0.9– 27.5 mm), and 13.8 ± 4.3 mm (from hotspots to caudal points, range 2.0–24.2 mm) were measured. Conclusions: According to the results, the minimum spatial resolution of rTMS should principally allow for the iden‑ tification of a particular gyrus, and according to the literature, it is in good accordance with the spatial resolution of direct cortical stimulation (DCS). Since measurement was performed between hotspots and adjacent mapping points and not on a finer ‑ grained basis, we only refer to a minimum spatial resolution. Furthermore, refinement of our results within the scope of a prospective study combining rTMS and DCS for resolution measurement during language map‑ ping should be the next step. Keywords: Cortical stimulation, Language function, Navigated transcranial magnetic stimulation, Preoperative language mapping, Spatial resolution Background Although language mapping based on rTMS is fre- Navigated transcranial magnetic stimulation (nTMS) quently used for preoperative neurosurgical diagnostics nowadays plays a crucial role in presurgical planning, as well as neuroscientific trials, the spatial resolution as it can be used to map cortical areas associated with of this method is largely unknown [1, 4–6, 10]. In this motor or language function [1–5]. When applied with context, the spatial resolution of rTMS is regarded as high frequency during an object-naming task (repetitive the average of distances between language-positive and nTMS = rTMS), this technique is able to elicit a transient language-negative stimulation points. Since direct com- impairment of language or speech performance within parison to intraoperative stimulation by direct cortical the scope of cortical mapping [6–9]. stimulation (DCS), which is able to differentiate between essential, language-positive and language-negative areas, is not possible in healthy subjects, the present study *Correspondence: Sandro.Krieg@tum.de Department of Neurosurgery, Klinikum rechts der Isar, Technische was designed to assess the minimum spatial resolu- Universität München, Ismaninger Str. 22, 81675 Munich, Germany tion of rTMS for language mapping. This was achieved Full list of author information is available at the end of the article © The Author(s) 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Sollmann et al. BMC Neurosci (2016) 17:67 Page 2 of 10 by differentiating between rTMS hotspots (stimulation For baseline testing, the volunteers had to name all points with high naming error rates due to rTMS) and objects, and any delayed or misnamed objects were dis- adjacent rTMS mapping points (stimulation points with carded. Baseline testing was carried out twice, meaning low naming error rates), followed by discussion of results that the second run was performed with the remain- with regard to previous data derived from studies that ing object stack immediately after the first testing, and used pre- or intraoperative mapping techniques. Regard- the total number of naming errors was documented in ing the stimulation approach, rTMS was carried out over the end. Consequently, only objects that were correctly predefined spots distributed across the left hemisphere, named twice were used during language mapping. Thus, which follows the most common approach of rTMS- baseline testing was performed to be able to remove based language mapping to date. objects the individual participants were not familiar with, resulting in an individualized stack of images. During Methods language mapping, incorporation of objects that the par- Volunteers ticipants are not acquainted with could cause incorrect All mapping sessions were originally performed in 19 attribution of naming errors to stimulation effects rather healthy, right-handed (as indicated by the Edinburgh than to object unfamiliarity, leading to imprecise results. Handedness Inventory = EHI) subjects in 2013 to inves- tigate different research questions of rTMS-based lan - Procedure guage mapping [11, 12]. The enrolled volunteers were The examination started with the determination of the German native speakers without general transcranial individual resting motor threshold (RMT) by motor map- magnetic stimulation (TMS) exclusion criteria (e.g., ping of the cortical representation of the hand area [2, metal implants). 6, 7, 10]. The volunteers sat in a comfortable chair with armrests, and electrodes (Neuroline 720, Ambu, Bal- Magnetic resonance imaging lerup, Denmark) were placed over the abductor pollicis All imaging was performed on a magnetic resonance brevis (APB) and abductor digiti minimi (ADM) mus- scanner (Achieva 3T, Philips Medical Systems, The cle of the right hand to be able to detect motor evoked Netherlands B.V.) by the use of an eight-channel phased- potentials (MEPs) during continuous electromyography array head coil. Our scanning protocol consisted of a (EMG) recording. Furthermore, the reference electrode three-dimensional (3-D) gradient echo sequence (TR/ was placed at the elbow. Then, a coarse round of single- TE 9/4  ms, 1  mm isovoxel covering the whole head, pulse stimulations was conducted to localize the spot 6  min and 58  s acquisition time) without intravenous with the highest MEP amplitude, the motor hotspot, contrast administration. Subsequent to imaging, the which is usually found within the area of the hand knob 3-D magnetic resonance imaging (MRI) dataset of each [14]. During pulse application, the induced electrical field subject was transferred to the rTMS system via DICOM was oriented perpendicular to the precentral gyrus, and standard. the RMT was then determined at the motor hotspot. In this context, the RMT was defined as the lowest stimula - Transcranial magnetic stimulation tion intensity that elicits MEPs over 50  µV in amplitude In all volunteers, language mapping by rTMS was per- in at least 50% of stimulation trials in a relaxed muscle formed using the Nexstim eXimia NBS, version 4.3, [15]. According to the stimulation protocol, the exact combined with a NexSpeech module and a biphasic stimulation intensity for later language mapping was figure-of-eight stimulation coil (Nexstim Oy, Helsinki, adjusted with respect to the RMT. Finland). During stimulation, the subjects had to name their individualized sets of objects according to previous base- Task line testing. Thus, all objects that were correctly named During baseline testing and rTMS-based language twice according to baseline testing were displayed in ran- mapping, volunteers participated in an object-naming domized order while rTMS was applied in a time-locked task, which has been frequently used in previous rTMS fashion. Overall, 46 left-hemispheric cortical spots, investigations [2, 6–8, 10–12]. The task consisted of a which were manually tagged on the 3-D MRI scan of the total amount of 100 colored photographs, similar to the volunteer prior to mapping, were stimulated three times objects of the Snodgrass and Vanderwart picture set each in a row without any particular order (Fig.  1), as (1980) [11–13]. The photographs portrayed familiar liv - reported previously [11, 12]. Thus, mapping was carried ing as well as non-living objects (e.g., banana, chair, out over predefined spots. snake), and had to be named in German as quickly and Due to unacceptable stimulation-related discom- precisely as possible. fort, none of the 46 points were located within the Sollmann et al. BMC Neurosci (2016) 17:67 Page 3 of 10 Table 1 Cortical parcellation system (CPS) Number Abbreviation Anatomy 1 anG Angular gyrus 2 aSMG Anterior supramarginal gyrus 3 aSTG Anterior superior temporal gyrus 4 dPoG Dorsal postcentral gyrus 5 dPrG Dorsal precentral gyrus 6 mMFG Middle middle frontal gyrus 7 mMTG Middle middle temporal gyrus 8 mPoG Middle postcentral gyrus 9 mPrG Middle precentral gyrus Fig. 1 Cortical parcellation system (CPS). This figure visualizes the 10 mSFG Middle superior frontal gyrus cortical stimulation targets (white spots, n = 46) within a template of 11 mSTG Middle superior temporal gyrus the left hemisphere. In addition, the cortical surface is divided into 12 opIFG Opercular inferior frontal gyrus subregions, and the numbers refer to the anatomical names of the 13 pMFG Posterior middle frontal gyrus stimulated subregions (see Table 1 for anatomical names and abbre‑ viations of the CPS) 14 pMTG Posterior middle temporal gyrus 15 pSFG Posterior superior frontal gyrus 16 pSMG Posterior supramarginal gyrus 17 pSTG Posterior superior temporal gyrus following regions: orbital part of the inferior frontal 18 SPL Superior parietal lobe gyrus (orIFG), polar superior temporal gyrus (polSTG), 19 trIFG Triangular inferior frontal gyrus polar middle temporal gyrus (polMTG), anterior mid- 20 vPoG Ventral postcentral gyrus dle temporal gyrus (aMTG), polar superior frontal 21 vPrG Ventral precentral gyrus gyrus (polSFG), polar middle frontal gyrus (polMFG), and polar inferior frontal gyrus (polIFG). In addition, Anatomical names and abbreviations of the CPS for all regions stimulated. The numbers refer to the individual subregions, which are visualized in Fig. 1 the inferior temporal gyrus (ITG) was also not mapped because of the increased coil-cortex distance. Parcel- lation of the anatomical structures was performed according to the cortical parcellation system (CPS, • Stimulation intensity: 100% of the RMT Table 1; Fig. 1) [2, 6, 16]. • Stimulation frequency: 7 Hz For transcranial stimulation, the magnetic coil was • Number of pulses: 10 placed tangential to the subject’s skull, and the induced • Duration of each stimulation burst (7 Hz/10 pulses): electrical field was oriented in strict anterior–posterior 1430 ms direction during language mapping [2, 6, 8, 17, 18]. Both • Picture-to-trigger interval (=PTI, time between the coil angulation (tangential to the subject’s skull) and the presentation of an object on the screen and the electrical field orientation (anterior–posterior) were dis - beginning of the rTMS pulse): 0 ms played by the help of the neuronavigation unit. Stimula- • Display time (=DT, duration of the screening of an tion trials with incorrect coil angulation or orientation object): 700 ms were repeated and not taken into account during analysis. • Inter-picture interval (=IPI, time between the All points of stimulation and the electrical field strength screening of two objects): 3000 ms at the stimulation spots were automatically saved for later analysis. Moreover, a video of the naming performance Data analysis during baseline testing as well as during stimulation was Error maps recorded for post hoc analysis. All videos were analyzed for no-response errors strictly blinded to the stimulated cortical spots by the same per- Protocol son who had already conducted rTMS [6–8, 19]. A no- For each mapping session, the following stimulation response error was defined as a complete lack of naming parameters were chosen, as they (a) have shown to be response within the duration of the stimulation. Other efficient in terms of eliciting reproducible naming errors error types, like hesitations or performance errors, during object naming, and (b) have been proven to be were not taken into account in the present investiga- well tolerable and safe for the individual subject [1, 2, tion. Although other error types represent stimulation- 6–8, 10–12]: induced disruption of important language subfunctions Sollmann et al. BMC Neurosci (2016) 17:67 Page 4 of 10 as well, we decided to not take these categories into these adjacent spots shows an error rate of less than 11% account since no responses have proven to be among and can clearly be characterized as ventral, dorsal, apical, the most frequent error types whilst being easy to detect or caudal with respect to the hotspot. during video analysis [6, 20]. After identification of the four suitable hotspots and The total numbers of no-response errors as well as the their adjacent points on the template (Fig. 2), these spots numbers of stimulation bursts were pooled across all vol- were exported from each volunteer’s individual rTMS unteers. Then, mapping results of the 46 stimulated corti - mapping session via DICOM standard to an external cal spots were projected into the CPS by putting pooled working station. Then, measurement of a minimum spa - error rates (=number of induced no responses at one of tial resolution for rTMS-based language mapping was the 46 cortical spots divided by the total number of stim- performed using OsiriX imaging software (OsiriX version ulation bursts applied to this spot) on a brain template of 5.8.5, Pixmeo SARL, Bernex, Switzerland). Therefore, the the left hemisphere (Fig. 2). two-dimensional (2-D) distances parallel to the cortical surface between each hotspot and its corresponding adja- Resolution measurement cent points were measured separately on the individual Cortical spots that were prone to comparatively high no- coronal MRI slices for the apical and caudal points to the response error rates (=hotspots) and were surrounded by hotspot (Fig.  3a), and on the sagittal MRI slices for the four spots with lower error rates were identified from the distance between the hotspot and the ventral as well as raw data of mapping results and the error map (Fig.  2). the dorsal spots (Fig.  3b). Since measurement was done Hence, hotspot definition was based on visual inspection between hotspots and adjacent mapping points belong- of error distributions (Fig.  2) without additional statisti- ing to the CPS and not on a finer-grained basis including cal testing between single stimulation points. The four cortical spots lying in-between, we refer to a minimum adjacent spots were separated into one ventral, one dor- spatial resolution. sal, one apical, and one caudal point with respect to the localization of the hotspot. If there were more than four Statistics points matching the inclusion criteria for being adequate Mean values  ±  standard deviation (SD), medians, mini- adjacent spots, the closest one to the hotspot was cho- mum and maximum values of subject-related charac- sen. Applying the described rule to Fig.  2, a number of teristics, mapping parameters, and distances between four stimulated points fulfilled the hotspot criteria (hot - mapping spots were calculated by using GraphPad Prism spots 1, 2, 3, and 4). Hotspot 1, for example, shows a no- (GraphPad Prism 6, La Jolla, CA, USA). Furthermore, a response error rate of 11%, and is surrounded by values one-way analysis of variance (ANOVA) between the indi- of only 7% (ventral spot within the mMFG), 5% (api- vidual electrical field strengths of the enrolled subjects cal spot within the mMFG), 0% (dorsal spot within the at the hotspots was performed, and a one-way ANOVA pMFG), and 5% (caudal spot within the opIFG). Each of followed by Tukey’s multiple comparisons test, including calculation of 95% confidence intervals (CIs), was carried out to compare distance measurements between hot- spots and adjacent mapping points. For all statistical cal- culations, a p value of <0.05 was considered statistically significant. Results Subject and mapping characteristics Mappings were successfully performed in 19 healthy, right-handed volunteers, which were already analyzed for different purposes in previous investigations [11, 12]. Table  2 gives an overview of subject-related characteris- tics and mapping parameters. Regarding the comparison of electrical field strengths at the hotspots (Table  2), there was no statistically significant difference between the Fig. 2 Distribution of no‑response errors. This figure shows the four hotspots (p = 0.3445). no‑response error rates (=number of induced no responses at a certain stimulation spot divided by the total number of stimulation Spatial resolution bursts applied to this spot) as a percentage projected into the cortical Overall, mean distance values of 13.8  ±  6.4  mm (aver- parcellation system (CPS) including all stimulated spots (n = 46). Additionally, the four identified hotspots are marked age of all hotspots to ventral points, range 0.7–30.7 mm), Sollmann et al. BMC Neurosci (2016) 17:67 Page 5 of 10 Fig. 3 Spatial resolution measurement. Illustration of the distance measurement procedure on a subject’s individual coronal magnetic resonance imaging (MRI) slice for the apical and caudal points to the hotspot (a), and on a sagittal MRI slice for the distance between the hotspot and the ven‑ tral and dorsal spots (b). The black lines represent the distance between the hotspot (H) and the corresponding apical (a) or dorsal point (b) parallel to the cortical surface Table 2 Subject and mapping characteristics minimum, and maximum distance values of the meas- urements between each particular hotspot and the cor- Mean ± SD Range responding adjacent points. Furthermore, the results of Age (years) 24.6 ± 1.7 21.8–29.4 measurements between all hotspots taken together and EHI score 84.3 ± 13.2 57–100 adjacent points are illustrated in Fig. 4. Pain ( VAS) According to ANOVA, statistically significant differ - Convexity 2.2 ± 1.6 0–6 ences regarding the measurements between the hotspots Temporal 5.0 ± 2.0 2–10 and adjacent points were revealed with respect to hotspot Correct baseline objects 93.7 ± 3.3 87–98 1, 2, and 3 (F = 16.5, F = 11.7, F = 8.8, p < 0.0001, Table 3), RMT (% output) 33.5 ± 5.1 24–43 whereas no statistically significant difference was revealed Electric field strength ( V/m) concerning hotspot 4 (F  =  2.1, p  =  0.1111, Table  3). Hotspot 1 62.0 ± 9.8 48–84 Regarding hotspot 1, there was a statistically significant Hotspot 2 58.0 ± 11.5 40–83 difference in measurements for the hotspot to ventral ver - Hotspot 3 63.0 ± 10.3 42–87 sus apical points (CI −9.6 to −3.7), to dorsal versus api- Hotspot 4 65.0 ± 14.1 42–90 cal points (CI −10.0 to −4.1), and to apical versus caudal points (CI 1.2  to  7.2), whereas statistically significant dif - Overview about subject-related characteristics including age, Edinburg Handedness Inventory (EHI) scores, and discomfort during stimulation according ferences were revealed for the hotspot to ventral versus to the visual analogue scale (VAS). Moreover, the individual amount of correctly dorsal (CI 1.0  to  6.9), to dorsal versus apical (CI −9.5 to named objects during baseline testing, the resting motor threshold (RMT ), and the electrical field strength that was applied to the four hotspots during −3.6), and to dorsal versus caudal points (CI −6.6 to −0.7) stimulation are shown. In this context, the electrical field strength for each for hotspot 2. Concerning hotspot 3, there was a statisti- mapping point was automatically calculated and saved by the system cally significant difference in measurements for the hot - spot to dorsal versus apical points (CI −11.8 to −3.6) and 10.8  ±  4.8  mm (average of all hotspots to dorsal points, to dorsal versus caudal points (CI −9.5 to −1.3). range 2.0–26.5  mm), 16.6  ±  4.8  mm (average of all hotspots to apical points, range 0.9–27.5  mm), and Discussion 13.8 ± 4.3 mm (average of all hotspots to caudal points, Current knowledge range 2.0–24.2  mm) were measured. More detailed, Only a few publications investigated the resolution of Table  3 provides information about the mean  ±  SD, TMS up to now [21–24]. The publications of Opitz et al. Sollmann et al. BMC Neurosci (2016) 17:67 Page 6 of 10 Table 3 Measurement results Hotspot Adjacent spots Mean SD MIN MAX p value 1 Ventral 11.9 2.2 8.6 16.8 <0.0001 Dorsal 11.5 5.4 6.0 26.5 Apical 18.6 2.5 14.7 22.1 Caudal 14.4 2.8 9.0 17.8 2 Ventral 18.0 3.9 11.6 25.8 <0.0001 Dorsal 14.1 3.1 9.3 20.1 Apical 20.6 3.7 14.1 27.5 Caudal 17.7 2.9 13.8 24.2 3 Ventral 10.7 8.3 0.7 30.7 <0.0001 Dorsal 6.9 2.2 2.0 11.0 Apical 14.6 2.8 7.2 20.9 Caudal 12.3 3.0 6.5 18.2 4 Ventral 14.5 6.8 1.9 27.5 0.1111 Dorsal 10.7 4.9 3.8 18.4 Apical 12.6 5.1 0.9 21.0 Caudal 10.9 4.7 2.0 17.8 All directions 1 14.1 4.4 6.0 26.5 – 2 17.6 4.1 9.3 27.5 3 11.1 5.5 0.7 30.7 4 12.2 5.5 0.9 27.5 All hotspots and directions 13.8 5.5 0.7 30.7 – This table provides information about the spatial resolution measurements (in mm) by showing the distances from the four different hotspots to the four corresponding adjacent points (hotspot to ventral, dorsal, apical, and caudal spots), presented as mean ± standard deviation (SD), minimum (MIN), and maximum (MAX) values. In addition, one-way analysis of variance (ANOVA) was performed to compare distance measurements between hotspots and adjacent mapping points, leading to the p values presented in the table. In this context, a p value of <0.05 was considered statistically significant means that the focality of the electrical field is increased [22]. The authors Bijsterbosch et  al. [23] systematically investigated the impact of various TMS coil positions on electrical field shaping, finding that, at least for most coil positions, the induced field includes the target region of stimulation but is not distinctly restricted to it. Accord- ing to their publication, the distribution of subarachnoid cerebral fluid and gyral geometry predominantly influ - ence the modelling of the electrical field in the human brain [23]. Although the described publications represent exten- sive and valuable contributions to the current knowledge about the spatial resolution of TMS, they neither explic- Fig. 4 Measurement results. Boxplots with median, min‑, and max ‑ whiskers, and quartile‑boxes for the hotspot to ventral, dorsal, apical, itly focused on cortical language mapping, nor did they and caudal distance measurements (in mm) examine the spatial resolution from a functional point of view, meaning that they did not particularly aim for a dif- ferentiation of functionally relevant from irrelevant corti- [21] and Thielscher et  al. [22] calculated the electri - cal spots [21–23]. Under these premises, examination of cal field induced by TMS in the human brain and came the spatial resolution of rTMS-based language mapping to the conclusion that the field strength is significantly seemed mandatory. enhanced when the currents run perpendicular to the stimulated gyrus. Moreover, this effect was demonstrated Gyrus identification to be primarily restricted to the gyral lips and crowns, The average width of a cortical gyrus in the adult human but it did not extend into the walls of the sulci, which brain varies between 10 and 20  mm, although there can Sollmann et al. BMC Neurosci (2016) 17:67 Page 7 of 10 be distinct inter-individual variations. Therefore, when language-related task performance to the signal obtained the central point of a figure-of-eight TMS coil is placed during resting-state measurement within the same sub- perpendicular to the subject’s skull and in the middle of a ject [25, 26]. Then, the systematic comparison of both certain gyrus, the center of stimulation should primarily signal datasets allows drawing conclusions about the hit the stimulated gyrus, as our mean spatial resolution localization of individual language-related brain areas. measurements for each of the four directions accounted Overall, fMRI is characterized by comparatively good for values clearly under 20  mm (Table  3; Fig.  4). As an spatial resolution, and a recent study on nTMS-based important result, we can state that specific gyri can be motor mapping has demonstrated that the spatial differ - targeted and identified by rTMS, at least when stimula - ences between nTMS- and DCS-positive cortical spots tion is applied to the gyrus center with the protocol used and fMRI- and DCS-positive points are both in the range in the present study. of millimeters, although better results were observed Furthermore, because the spatial resolution of rTMS for nTMS- versus DCS-positive spots (10.5  ±  5.7 vs. does not exceed the average gyrus width and a particu- 15.0 ± 7.6 mm) [27]. lar gyrus can be targeted by this modality, the common Another neuroimaging modality for the investigation of practice of parcellating the cortical surface into subre- cortical language representation is magnetoencephalog- gions according to its gyral structure (e.g., as is done for raphy (MEG), which has been particularly used for deter- the CPS) seems to be a reasonable approach for rTMS mination of language lateralization over the last years research. [28, 29]. In this context, cortical activation in response to a language-related task performance goes along with Electrical field strength a local rise in neuronal signaling, which is characterized In general, a higher electrical field strength induced by by an increased flow of intracellular ions mediating asso - rTMS should principally cause a lower spatial resolution ciated magnetic fields that can be measured at the scalp and vice versa. The mean electrical field strengths at the surface in the form of event-related potentials [30]. How- four hotspots were comparable, and there was no statisti- ever MEG has failed to show sufficient correlation with cally significant difference revealed. Thus, our approach language maps generated by rTMS or DCS, at least partly including a comparison of hotspots being located within due to its lower spatial resolution, which is typically lim- different remote cortical gyri seems reasonable because ited to several centimeters [1, 31]. the applied electrical field does not significantly  vary When it comes to DCS, which is regarded as the cur- between these spots, and therefore, minimum measured rent gold standard for mapping human cortex function, distances seem comparable. However, statistical analysis language performance has been repeatedly tested dur- of the four hotspots revealed a statistically significant dif - ing awake surgery [32–34]. Interestingly, Haglund et  al. ference of the mean distances between the hotspots and [33] showed that radical brain tumor resection without adjacent stimulation spots (Table  3), highlighting that postoperative permanent language deficits can only be the mean differences in distance measurements to the achieved reliably when the resection border is at least hotspot were different for the respective adjacent map - 10  mm away from the nearest language site determined ping points. This might show that spatial resolution of by DCS. Thus, 10 mm seems to be the spatial resolution rTMS for language mapping—at least within a certain of DCS-based language mapping during awake surgery. range—depends on the localization of cortical stimula- As our findings show, the spatial resolution of rTMS- tion, although the applied electrical field strength does based language mapping is slightly above 10 mm (Table 3; not significantly change during mapping. Yet, this effect Fig. 4), which is comparable to the results of DCS [33]. might also be due to the chosen distribution of our corti- cal stimulation spots, and we can state that the minimally Further considerations measured spatial resolution of rTMS was 11.1 ± 5.5 mm Regarding the distribution of no responses across the (Table 3). hemisphere, the opIFG and trIFG, which should roughly overlap with the classical Broca’s area, were not charac- Comparison to other modalities terized by a high rate of errors during rTMS. However, Overall, cortical language distribution has been under these regions showed error rates to a higher extent when extensive investigation during the last decades. In this it comes to other error categories, which were incorpo- context, one frequently used technique for the identi- rated into analysis of the same dataset in other publica- fication of cortical sites related to language function is tions focusing on different purposes [11, 12]. Hence, functional magnetic resonance imaging (fMRI). In its considerable error rates were achieved during mapping of most common form, this approach compares the blood the opIFG and trIFG, but belonging to other types than oxygenation level dependent (BOLD) signal during a no responses. As a possible explanation for the low rate Sollmann et al. BMC Neurosci (2016) 17:67 Page 8 of 10 of no responses in these regions when compared to pre- mapping parameters, like the stimulation frequency or vious investigations [6–8, 20], the stimulation frequency the angulation and shape of the magnetic coil, for exam- of 7 Hz (instead of 5 Hz) has to be considered, but the set ple [1, 11, 17, 36]. As a consequence, it cannot be the of objects, which was different when compared to these claim of the present study to define the definite spatial previous publications, might have also played a role. resolution of rTMS-based language mapping indepen- However, the distinct cause for the specific pattern of no dently from setup factors. Instead, this study at least pro- responses observed in the present study cannot be fully vides some practical orientation on the minimum spatial assessed within the scope of our approach. resolution for a commonly used mapping protocol, which Although the resolution measurements of the pre- has revealed to be reliable in terms of inducing transient sent study can be discussed in the light of  previous lit- errors during object naming while being well tolerable erature on DCS-based language mapping, there is no for the individual subject. opportunity to directly compare rTMS to DCS results Another potential limitation could be the fact that on an intra-individual basis in healthy subjects due to distance measurement was only performed for prede- the highly invasive character of DCS. As a consequence, fined cortical points. In this context, the rTMS resolu - the language-positive areas mapped in the present study tion measurement was limited by the distance between cannot be verified by the gold standard method, and a single stimulation points of the CPS, meaning that areas final decision about the possible essential character of a in-between the different stimulation spots were not sys - particular language-positive site cannot be made due to tematically examined. Thus, the results presented in the the comparatively high sensitivity and low specificity of current approach reflect a minimum resolution measure - rTMS [10]. However, this limitation can at least partly ment. Consequently, it might be possible that the defi - be overcome by the results of recent studies that showed nite resolution of rTMS is finer-grained. Yet, all cortical a much higher specificity and positive predictive value points of stimulation were placed manually at the center between language maps generated by both mapping tech- of the gyri, meaning that finer-grained mapping might niques using the CPS, which again provides spatial reso- also have targeted sulci instead of cortical tissue, which lution data on a gyri level [1, 2]. To draw more definite is generally not regarded as adequate for the rTMS-based conclusions concerning resolution, a prospective study language mapping approach per se [6, 8]. We decided to including preoperative rTMS and intraoperative DCS follow the presented kind of approach because one major mapping among patients should be the next step. objective was to investigate the required cortical distance In the recent years, the differentiation of naming errors to distinguish between hotspots (stimulation points with into various predefined error categories has been dem - a high error rate) and adjacent mapping points (stimu- onstrated repeatedly within the scope of rTMS and DCS lation points with low error rates) in a practical setup. trials [4, 6, 8, 16, 20, 34]. Although other error categories Basically, this method is similar to the measurement pre- except no responses are based on stimulation-induced sented in Haglund et al., which represents a key reference disruption of important language subfunctions as well, for the accuracy of DCS-based language mapping during we decided to not take these error types into account awake surgery [33]. since no responses have proven to be among the most frequent error types [6, 20]. Incorporation of other error Conclusions types could have led to different hotspots, but poten - The present study examined the spatial resolution of tially on the basis of more unclear language disruption. rTMS-based language mapping via distance meas- Therefore, our current approach seems to be reasonable urement. In this context, rTMS is able to differenti - within the scope of one of the first studies regarding the ate between hotspots (stimulation points with a high exploration of rTMS-based language mapping resolution. error rate) and adjacent mapping points (stimulation However, we have to be aware of the fact that upcoming points with low error rates) at a minimum distance studies systematically analyzing other categories might of 11.1  ±  5.5  mm as measured for a parietal hotspot. further refine our minimum resolution measurements. According to these distance measurement results, the Additionally, the spatial resolution of rTMS should also spatial resolution of rTMS should principally allow for be evaluated for mapping of other cognitive functions the identification of a particular gyrus as it is the case for since such approaches are currently emerging. In this DCS-based language mapping during awake surgery as context, non-invasive assessment of calculation functions the current gold standard. by rTMS has been successfully performed recently, but spatial resolution for this purpose is vastly unknown [35]. Abbreviations Furthermore, it is already known that rTMS mapping 2‑D: two ‑ dimensional; 3‑D: three ‑ dimensional; aMTG: anterior middle results and electrical field shaping depend on various temporal gyrus; ANOVA: analysis of variance; ADM: abductor digiti minimi; Sollmann et al. BMC Neurosci (2016) 17:67 Page 9 of 10 APB: abductor pollicis brevis; BOLD: blood oxygenation level dependent; CI: 2. Krieg SM, Tarapore PE, Picht T, Tanigawa N, Houde J, Sollmann N, Meyer confidence interval; CPS: cortical parcellation system; DT: display time; DCS: B, Vajkoczy P, Berger MS, Ringel F, et al. Optimal timing of pulse onset direct cortical stimulation; EHI: Edinburgh Handedness Inventory; EMG: elec‑ for language mapping with navigated repetitive transcranial magnetic tromyography; fMRI: functional magnetic resonance imaging; IPI: inter‑picture stimulation. NeuroImage. 2014;100:219–36. interval; ITG: inferior temporal gyrus; MEG: magnetoencephalography; MEP: 3. Picht T, Schulz J, Vajkoczy P. The preoperative use of navigated motor evoked potential; MRI: magnetic resonance imaging; nTMS: navigated transcranial magnetic stimulation facilitates early resection of sus‑ transcranial magnetic stimulation; orIFG: orbital part of the inferior frontal pected low‑ grade gliomas in the motor cortex. Acta Neurochir ( Wien). gyrus; polIFG: polar inferior frontal gyrus; polMFG: polar middle frontal gyrus; 2013;155(10):1813–21. polMTG: polar middle temporal gyrus; polSFG: polar superior frontal gyrus; 4. Rosler J, Niraula B, Strack V, Zdunczyk A, Schilt S, Savolainen P, Lioumis polSTG: polar superior temporal gyrus; PTI: picture‑to ‑trigger interval; RMT: P, Makela J, Vajkoczy P, Frey D, et al. Language mapping in healthy resting motor threshold; rTMS: repetitive navigated transcranial magnetic volunteers and brain tumor patients with a novel navigated TMS stimulation; SD: standard deviation; TMS: transcranial magnetic stimulation; system: evidence of tumor‑induced plasticity. Clin Neurophysiol. VAS: visual analogue scale. 2014;125(3):526–36. 5. Picht T, Mularski S, Kuehn B, Vajkoczy P, Kombos T, Suess O. Navigated Authors’ contributions transcranial magnetic stimulation for preoperative functional diagnostics NS: manuscript preparation, data acquisition, data handling, data analysis, in brain tumor surgery. Neurosurgery. 2009;65(6 Suppl):93–8 (discussion statistics, literature review. TH, LT, SI, SM, TBB: data acquisition, data handling, 98–9). data analysis. FR, BM, SK: data acquisition, data handling, data analysis, study 6. Sollmann N, Tanigawa N, Ringel F, Zimmer C, Meyer B, Krieg SM. supervision. All authors read and approved the final manuscript. Language and its right‑hemispheric distribution in healthy brains: an investigation by repetitive transcranial magnetic stimulation. NeuroIm‑ Authors’ information age. 2014;102P2:776–88. Nico Sollmann is a Ph.D. student at the Department of Neurosurgery. 7. Hernandez‑Pavon JC, Makela N, Lehtinen H, Lioumis P, Makela JP. Eec ff ts Theresa Hauck is a medical doctor and lab intern. Lorena Tussis is a master of navigated TMS on object and action naming. Front Hum Neurosci. student and lab intern. Sebastian Ille and Stefanie Maurer are residents 2014;8:660. at the Department of Neurosurgery. Tobias Boeckh‑Behrens is an attend‑ 8. Lioumis P, Zhdanov A, Makela N, Lehtinen H, Wilenius J, Neuvonen T, Han‑ ing neuroradiologist, and Sandro M. Krieg is an attending neurosurgeon. nula H, Deletis V, Picht T, Makela JP. A novel approach for documenting Florian Ringel is vice chief, and Bernhard Meyer is chief of the Department of naming errors induced by navigated transcranial magnetic stimulation. J Neurosurgery. Neurosci Methods. 2012;204(2):349–54. 9. Pascual‑Leone A, Gates JR, Dhuna A. Induction of speech arrest and Author details counting errors with rapid‑rate transcranial magnetic stimulation. Neurol‑ Department of Neurosurgery, Klinikum rechts der Isar, Technische Universität ogy. 1991;41(5):697–702. München, Ismaninger Str. 22, 81675 Munich, Germany. TUM‑Neuroimaging 10. Picht T, Krieg SM, Sollmann N, Rosler J, Niraula B, Neuvonen T, Savol‑ Center, Klinikum rechts der Isar, Technische Universität München, Ismaninger ainen P, Lioumis P, Makela JP, Deletis V, et al. A comparison of language Str. 22, 81675 Munich, Germany. Section of Neuroradiology, Department mapping by preoperative navigated transcranial magnetic stimulation of Radiology, Klinikum rechts der Isar, Technische Universität München, Isma‑ and direct cortical stimulation during awake surgery. Neurosurgery. ninger Str. 22, 81675 Munich, Germany. 2013;72(5):808–19. 11. Hauck T, Tanigawa N, Probst M, Wohlschlaeger A, Ille S, Sollmann N, Mau‑ Acknowledgements rer S, Zimmer C, Ringel F, Meyer B, et al. Stimulation frequency determines Nico Sollmann gratefully acknowledges the support of the graduate school of the distribution of language positive cortical regions during navigated our university. transcranial magnetic brain stimulation. BMC Neurosci. 2015;16(1):5. 12. Hauck T, Tanigawa N, Probst M, Wohlschlaeger A, Ille S, Sollmann N, Competing interests Maurer S, Zimmer C, Ringel F, Meyer B, et al. Task type affects location of FR and SK are consultants for BrainLAB AG (Feldkirchen, Germany). SK is con‑ language‑positive cortical regions by repetitive navigated transcranial sultant for Nexstim Oy (Helsinki, Finland). The authors declare that they have magnetic stimulation mapping. PLoS ONE. 2015;10(4):e0125298. no competing interests. 13. Snodgrass JG, Vanderwart M. A standardized set of 260 pictures: norms for name agreement, image agreement, familiarity, and visual complexity. Availability of data and materials J Exp Psychol Hum Learn. 1980;6(2):174–215. All data used for analysis are presented in the manuscript. The discussion 14. Niskanen E, Julkunen P, Saisanen L, Vanninen R, Karjalainen P, Kononen M. and conclusions only rely on the data presented. Raw mapping data can be Group‑level variations in motor representation areas of thenar and ante ‑ provided upon request. rior tibial muscles: Navigated Transcranial Magnetic Stimulation Study. Hum Brain Mapp. 2010;31(8):1272–80. Ethics approval and consent to participate 15. Rossini PM, Barker AT, Berardelli A, Caramia MD, Caruso G, Cracco RQ, The local ethical committee of the Technische Universität München approved Dimitrijevic MR, Hallett M, Katayama Y, Lucking CH, et al. Non‑invasive the experimental procedures (Registration Number: 2793/10) in accordance electrical and magnetic stimulation of the brain, spinal cord and with the Declaration of Helsinki. Written informed consent was obtained from roots: basic principles and procedures for routine clinical application. all volunteers. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol. 1994;91(2):79–92. Funding 16. Corina DP, Loudermilk BC, Detwiler L, Martin RF, Brinkley JF, Ojemann The study was financed by institutional grants from the Department of Neuro ‑ G. Analysis of naming errors during cortical stimulation mapping: surgery and the Section of Neuroradiology. implications for models of language representation. Brain Lang. 2010;115(2):101–12. Received: 17 June 2016 Accepted: 19 October 2016 17. Epstein CM, Lah JJ, Meador K, Weissman JD, Gaitan LE, Dihenia B. Opti‑ mum stimulus parameters for lateralized suppression of speech with magnetic brain stimulation. Neurology. 1996;47(6):1590–3. 18. Ruohonen J, Karhu J. Navigated transcranial magnetic stimulation. Neuro‑ physiol Clin. 2010;40(1):7–17. 19. Picht T, Schmidt S, Brandt S, Frey D, Hannula H, Neuvonen T, Karhu References J, Vajkoczy P, Suess O. Preoperative functional mapping for rolandic 1. Tarapore PE, Findlay AM, Honma SM, Mizuiri D, Houde JF, Berger MS, brain tumor surgery: comparison of navigated transcranial magnetic Nagarajan SS. Language mapping with navigated repetitive TMS: proof of stimulation to direct cortical stimulation. Neurosurgery. 2011;69(3):581–8 technique and validation. NeuroImage. 2013;82:260–72. (discussion 588). Sollmann et al. BMC Neurosci (2016) 17:67 Page 10 of 10 20. Krieg SM, Sollmann N, Tanigawa N, Foerschler A, Meyer B, Ringel F: Corti‑ dominance for language assessed by magnetoencephalographic imag‑ cal distribution of speech and language errors investigated by visual ing. Ann Neurol. 2012;71(5):668–86. object naming and navigated transcranial magnetic stimulation. Brain 29. Hirata M, Goto T, Barnes G, Umekawa Y, Yanagisawa T, Kato A, Oshino Struct Funct. 2016;221(4):2259–86. S, Kishima H, Hashimoto N, Saitoh Y, et al. Language dominance and 21. Opitz A, Windhoff M, Heidemann RM, Turner R, Thielscher A. How the mapping based on neuromagnetic oscillatory changes: comparison with brain tissue shapes the electric field induced by transcranial magnetic invasive procedures. J Neurosurg. 2010;112(3):528–38. stimulation. NeuroImage. 2011;58(3):849–59. 30. Frye RE, Rezaie R, Papanicolaou AC. Functional neuroimaging of language 22. Thielscher A, Opitz A, Windhoff M. Impact of the gyral geometry on the using magnetoencephalography. Phys Life Rev. 2009;6(1):1–10. electric field induced by transcranial magnetic stimulation. NeuroImage. 31. Lev MH, Grant PE. MEG versus BOLD MR imaging: functional imaging, the 2011;54(1):234–43. next generation? AJNR Am J Neuroradiol. 2000;21(8):1369–70. 23. Bijsterbosch JD, Barker AT, Lee KH, Woodruff PW. Where does transcranial 32. Tate MC, Herbet G, Moritz‑ Gasser S, Tate JE, Duffau H. Probabilistic map magnetic stimulation ( TMS) stimulate? Modelling of induced field maps of critical functional regions of the human cerebral cortex: Broca’s area for some common cortical and cerebellar targets. Med Biol Eng Comput. revisited. Brain. 2014;137(Pt 10):2773–82. 2012;50(7):671–81. 33. Haglund MM, Berger MS, Shamseldin M, Lettich E, Ojemann GA. Cortical 24. Walsh V, Cowey A. Transcranial magnetic stimulation and cognitive neu‑ localization of temporal lobe language sites in patients with gliomas. roscience. Nat Rev Neurosci. 2000;1(1):73–9. Neurosurgery. 1994;34(4):567–76 (discussion 576). 25. Binder JR, Frost JA, Hammeke TA, Cox RW, Rao SM, Prieto T. Human brain 34. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after language map‑ language areas identified by functional magnetic resonance imaging. J ping for glioma resection. N Engl J Med. 2008;358(1):18–27. Neurosci. 1997;17(1):353–62. 35. Maurer S, Tanigawa N, Sollmann N, Hauck T, Ille S, Boeckh‑Behrens T, 26. Binder JR. Functional MRI is a valid noninvasive alternative to Wada test‑ Meyer B, Krieg SM: Non‑invasive mapping of calculation function by ing. Epilepsy Behav. 2011;20(2):214–22. repetitive navigated transcranial magnetic stimulation. Brain Struct Funct. 27. Forster MT, Hattingen E, Senft C, Gasser T, Seifert V, Szelenyi A. Navigated 2016;221(8):3927–3947. transcranial magnetic stimulation and functional magnetic resonance 36. Bolognini N, Ro T. Transcranial magnetic stimulation: disrupt‑ imaging: advanced adjuncts in preoperative planning for central region ing neural activity to alter and assess brain function. J Neurosci. tumors. Neurosurgery. 2011;68(5):1317–24 (discussion 1324–5). 2010;30(29):9647–50. 28. Findlay AM, Ambrose JB, Cahn‑ Weiner DA, Houde JF, Honma S, Hinkley LB, Berger MS, Nagarajan SS, Kirsch HE. Dynamics of hemispheric Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Neuroscience Springer Journals

Results on the spatial resolution of repetitive transcranial magnetic stimulation for cortical language mapping during object naming in healthy subjects

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/results-on-the-spatial-resolution-of-repetitive-transcranial-magnetic-0Bx256dRc5

References (36)

Publisher
Springer Journals
Copyright
Copyright © 2016 by The Author(s)
Subject
Biomedicine; Neurosciences; Neurobiology; Animal Models
eISSN
1471-2202
DOI
10.1186/s12868-016-0305-4
pmid
27776478
Publisher site
See Article on Publisher Site

Abstract

Background: The spatial resolution of repetitive navigated transcranial magnetic stimulation (rTMS) for language mapping is largely unknown. Thus, to determine a minimum spatial resolution of rTMS for language mapping, we evaluated the mapping sessions derived from 19 healthy volunteers for cortical hotspots of no‑ response errors. Then, the distances between hotspots (stimulation points with a high error rate) and adjacent mapping points (stimulation points with low error rates) were evaluated. Results: Mean distance values of 13.8 ± 6.4 mm (from hotspots to ventral points, range 0.7–30.7 mm), 10.8 ± 4.8 mm (from hotspots to dorsal points, range 2.0–26.5 mm), 16.6 ± 4.8 mm (from hotspots to apical points, range 0.9– 27.5 mm), and 13.8 ± 4.3 mm (from hotspots to caudal points, range 2.0–24.2 mm) were measured. Conclusions: According to the results, the minimum spatial resolution of rTMS should principally allow for the iden‑ tification of a particular gyrus, and according to the literature, it is in good accordance with the spatial resolution of direct cortical stimulation (DCS). Since measurement was performed between hotspots and adjacent mapping points and not on a finer ‑ grained basis, we only refer to a minimum spatial resolution. Furthermore, refinement of our results within the scope of a prospective study combining rTMS and DCS for resolution measurement during language map‑ ping should be the next step. Keywords: Cortical stimulation, Language function, Navigated transcranial magnetic stimulation, Preoperative language mapping, Spatial resolution Background Although language mapping based on rTMS is fre- Navigated transcranial magnetic stimulation (nTMS) quently used for preoperative neurosurgical diagnostics nowadays plays a crucial role in presurgical planning, as well as neuroscientific trials, the spatial resolution as it can be used to map cortical areas associated with of this method is largely unknown [1, 4–6, 10]. In this motor or language function [1–5]. When applied with context, the spatial resolution of rTMS is regarded as high frequency during an object-naming task (repetitive the average of distances between language-positive and nTMS = rTMS), this technique is able to elicit a transient language-negative stimulation points. Since direct com- impairment of language or speech performance within parison to intraoperative stimulation by direct cortical the scope of cortical mapping [6–9]. stimulation (DCS), which is able to differentiate between essential, language-positive and language-negative areas, is not possible in healthy subjects, the present study *Correspondence: Sandro.Krieg@tum.de Department of Neurosurgery, Klinikum rechts der Isar, Technische was designed to assess the minimum spatial resolu- Universität München, Ismaninger Str. 22, 81675 Munich, Germany tion of rTMS for language mapping. This was achieved Full list of author information is available at the end of the article © The Author(s) 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Sollmann et al. BMC Neurosci (2016) 17:67 Page 2 of 10 by differentiating between rTMS hotspots (stimulation For baseline testing, the volunteers had to name all points with high naming error rates due to rTMS) and objects, and any delayed or misnamed objects were dis- adjacent rTMS mapping points (stimulation points with carded. Baseline testing was carried out twice, meaning low naming error rates), followed by discussion of results that the second run was performed with the remain- with regard to previous data derived from studies that ing object stack immediately after the first testing, and used pre- or intraoperative mapping techniques. Regard- the total number of naming errors was documented in ing the stimulation approach, rTMS was carried out over the end. Consequently, only objects that were correctly predefined spots distributed across the left hemisphere, named twice were used during language mapping. Thus, which follows the most common approach of rTMS- baseline testing was performed to be able to remove based language mapping to date. objects the individual participants were not familiar with, resulting in an individualized stack of images. During Methods language mapping, incorporation of objects that the par- Volunteers ticipants are not acquainted with could cause incorrect All mapping sessions were originally performed in 19 attribution of naming errors to stimulation effects rather healthy, right-handed (as indicated by the Edinburgh than to object unfamiliarity, leading to imprecise results. Handedness Inventory = EHI) subjects in 2013 to inves- tigate different research questions of rTMS-based lan - Procedure guage mapping [11, 12]. The enrolled volunteers were The examination started with the determination of the German native speakers without general transcranial individual resting motor threshold (RMT) by motor map- magnetic stimulation (TMS) exclusion criteria (e.g., ping of the cortical representation of the hand area [2, metal implants). 6, 7, 10]. The volunteers sat in a comfortable chair with armrests, and electrodes (Neuroline 720, Ambu, Bal- Magnetic resonance imaging lerup, Denmark) were placed over the abductor pollicis All imaging was performed on a magnetic resonance brevis (APB) and abductor digiti minimi (ADM) mus- scanner (Achieva 3T, Philips Medical Systems, The cle of the right hand to be able to detect motor evoked Netherlands B.V.) by the use of an eight-channel phased- potentials (MEPs) during continuous electromyography array head coil. Our scanning protocol consisted of a (EMG) recording. Furthermore, the reference electrode three-dimensional (3-D) gradient echo sequence (TR/ was placed at the elbow. Then, a coarse round of single- TE 9/4  ms, 1  mm isovoxel covering the whole head, pulse stimulations was conducted to localize the spot 6  min and 58  s acquisition time) without intravenous with the highest MEP amplitude, the motor hotspot, contrast administration. Subsequent to imaging, the which is usually found within the area of the hand knob 3-D magnetic resonance imaging (MRI) dataset of each [14]. During pulse application, the induced electrical field subject was transferred to the rTMS system via DICOM was oriented perpendicular to the precentral gyrus, and standard. the RMT was then determined at the motor hotspot. In this context, the RMT was defined as the lowest stimula - Transcranial magnetic stimulation tion intensity that elicits MEPs over 50  µV in amplitude In all volunteers, language mapping by rTMS was per- in at least 50% of stimulation trials in a relaxed muscle formed using the Nexstim eXimia NBS, version 4.3, [15]. According to the stimulation protocol, the exact combined with a NexSpeech module and a biphasic stimulation intensity for later language mapping was figure-of-eight stimulation coil (Nexstim Oy, Helsinki, adjusted with respect to the RMT. Finland). During stimulation, the subjects had to name their individualized sets of objects according to previous base- Task line testing. Thus, all objects that were correctly named During baseline testing and rTMS-based language twice according to baseline testing were displayed in ran- mapping, volunteers participated in an object-naming domized order while rTMS was applied in a time-locked task, which has been frequently used in previous rTMS fashion. Overall, 46 left-hemispheric cortical spots, investigations [2, 6–8, 10–12]. The task consisted of a which were manually tagged on the 3-D MRI scan of the total amount of 100 colored photographs, similar to the volunteer prior to mapping, were stimulated three times objects of the Snodgrass and Vanderwart picture set each in a row without any particular order (Fig.  1), as (1980) [11–13]. The photographs portrayed familiar liv - reported previously [11, 12]. Thus, mapping was carried ing as well as non-living objects (e.g., banana, chair, out over predefined spots. snake), and had to be named in German as quickly and Due to unacceptable stimulation-related discom- precisely as possible. fort, none of the 46 points were located within the Sollmann et al. BMC Neurosci (2016) 17:67 Page 3 of 10 Table 1 Cortical parcellation system (CPS) Number Abbreviation Anatomy 1 anG Angular gyrus 2 aSMG Anterior supramarginal gyrus 3 aSTG Anterior superior temporal gyrus 4 dPoG Dorsal postcentral gyrus 5 dPrG Dorsal precentral gyrus 6 mMFG Middle middle frontal gyrus 7 mMTG Middle middle temporal gyrus 8 mPoG Middle postcentral gyrus 9 mPrG Middle precentral gyrus Fig. 1 Cortical parcellation system (CPS). This figure visualizes the 10 mSFG Middle superior frontal gyrus cortical stimulation targets (white spots, n = 46) within a template of 11 mSTG Middle superior temporal gyrus the left hemisphere. In addition, the cortical surface is divided into 12 opIFG Opercular inferior frontal gyrus subregions, and the numbers refer to the anatomical names of the 13 pMFG Posterior middle frontal gyrus stimulated subregions (see Table 1 for anatomical names and abbre‑ viations of the CPS) 14 pMTG Posterior middle temporal gyrus 15 pSFG Posterior superior frontal gyrus 16 pSMG Posterior supramarginal gyrus 17 pSTG Posterior superior temporal gyrus following regions: orbital part of the inferior frontal 18 SPL Superior parietal lobe gyrus (orIFG), polar superior temporal gyrus (polSTG), 19 trIFG Triangular inferior frontal gyrus polar middle temporal gyrus (polMTG), anterior mid- 20 vPoG Ventral postcentral gyrus dle temporal gyrus (aMTG), polar superior frontal 21 vPrG Ventral precentral gyrus gyrus (polSFG), polar middle frontal gyrus (polMFG), and polar inferior frontal gyrus (polIFG). In addition, Anatomical names and abbreviations of the CPS for all regions stimulated. The numbers refer to the individual subregions, which are visualized in Fig. 1 the inferior temporal gyrus (ITG) was also not mapped because of the increased coil-cortex distance. Parcel- lation of the anatomical structures was performed according to the cortical parcellation system (CPS, • Stimulation intensity: 100% of the RMT Table 1; Fig. 1) [2, 6, 16]. • Stimulation frequency: 7 Hz For transcranial stimulation, the magnetic coil was • Number of pulses: 10 placed tangential to the subject’s skull, and the induced • Duration of each stimulation burst (7 Hz/10 pulses): electrical field was oriented in strict anterior–posterior 1430 ms direction during language mapping [2, 6, 8, 17, 18]. Both • Picture-to-trigger interval (=PTI, time between the coil angulation (tangential to the subject’s skull) and the presentation of an object on the screen and the electrical field orientation (anterior–posterior) were dis - beginning of the rTMS pulse): 0 ms played by the help of the neuronavigation unit. Stimula- • Display time (=DT, duration of the screening of an tion trials with incorrect coil angulation or orientation object): 700 ms were repeated and not taken into account during analysis. • Inter-picture interval (=IPI, time between the All points of stimulation and the electrical field strength screening of two objects): 3000 ms at the stimulation spots were automatically saved for later analysis. Moreover, a video of the naming performance Data analysis during baseline testing as well as during stimulation was Error maps recorded for post hoc analysis. All videos were analyzed for no-response errors strictly blinded to the stimulated cortical spots by the same per- Protocol son who had already conducted rTMS [6–8, 19]. A no- For each mapping session, the following stimulation response error was defined as a complete lack of naming parameters were chosen, as they (a) have shown to be response within the duration of the stimulation. Other efficient in terms of eliciting reproducible naming errors error types, like hesitations or performance errors, during object naming, and (b) have been proven to be were not taken into account in the present investiga- well tolerable and safe for the individual subject [1, 2, tion. Although other error types represent stimulation- 6–8, 10–12]: induced disruption of important language subfunctions Sollmann et al. BMC Neurosci (2016) 17:67 Page 4 of 10 as well, we decided to not take these categories into these adjacent spots shows an error rate of less than 11% account since no responses have proven to be among and can clearly be characterized as ventral, dorsal, apical, the most frequent error types whilst being easy to detect or caudal with respect to the hotspot. during video analysis [6, 20]. After identification of the four suitable hotspots and The total numbers of no-response errors as well as the their adjacent points on the template (Fig. 2), these spots numbers of stimulation bursts were pooled across all vol- were exported from each volunteer’s individual rTMS unteers. Then, mapping results of the 46 stimulated corti - mapping session via DICOM standard to an external cal spots were projected into the CPS by putting pooled working station. Then, measurement of a minimum spa - error rates (=number of induced no responses at one of tial resolution for rTMS-based language mapping was the 46 cortical spots divided by the total number of stim- performed using OsiriX imaging software (OsiriX version ulation bursts applied to this spot) on a brain template of 5.8.5, Pixmeo SARL, Bernex, Switzerland). Therefore, the the left hemisphere (Fig. 2). two-dimensional (2-D) distances parallel to the cortical surface between each hotspot and its corresponding adja- Resolution measurement cent points were measured separately on the individual Cortical spots that were prone to comparatively high no- coronal MRI slices for the apical and caudal points to the response error rates (=hotspots) and were surrounded by hotspot (Fig.  3a), and on the sagittal MRI slices for the four spots with lower error rates were identified from the distance between the hotspot and the ventral as well as raw data of mapping results and the error map (Fig.  2). the dorsal spots (Fig.  3b). Since measurement was done Hence, hotspot definition was based on visual inspection between hotspots and adjacent mapping points belong- of error distributions (Fig.  2) without additional statisti- ing to the CPS and not on a finer-grained basis including cal testing between single stimulation points. The four cortical spots lying in-between, we refer to a minimum adjacent spots were separated into one ventral, one dor- spatial resolution. sal, one apical, and one caudal point with respect to the localization of the hotspot. If there were more than four Statistics points matching the inclusion criteria for being adequate Mean values  ±  standard deviation (SD), medians, mini- adjacent spots, the closest one to the hotspot was cho- mum and maximum values of subject-related charac- sen. Applying the described rule to Fig.  2, a number of teristics, mapping parameters, and distances between four stimulated points fulfilled the hotspot criteria (hot - mapping spots were calculated by using GraphPad Prism spots 1, 2, 3, and 4). Hotspot 1, for example, shows a no- (GraphPad Prism 6, La Jolla, CA, USA). Furthermore, a response error rate of 11%, and is surrounded by values one-way analysis of variance (ANOVA) between the indi- of only 7% (ventral spot within the mMFG), 5% (api- vidual electrical field strengths of the enrolled subjects cal spot within the mMFG), 0% (dorsal spot within the at the hotspots was performed, and a one-way ANOVA pMFG), and 5% (caudal spot within the opIFG). Each of followed by Tukey’s multiple comparisons test, including calculation of 95% confidence intervals (CIs), was carried out to compare distance measurements between hot- spots and adjacent mapping points. For all statistical cal- culations, a p value of <0.05 was considered statistically significant. Results Subject and mapping characteristics Mappings were successfully performed in 19 healthy, right-handed volunteers, which were already analyzed for different purposes in previous investigations [11, 12]. Table  2 gives an overview of subject-related characteris- tics and mapping parameters. Regarding the comparison of electrical field strengths at the hotspots (Table  2), there was no statistically significant difference between the Fig. 2 Distribution of no‑response errors. This figure shows the four hotspots (p = 0.3445). no‑response error rates (=number of induced no responses at a certain stimulation spot divided by the total number of stimulation Spatial resolution bursts applied to this spot) as a percentage projected into the cortical Overall, mean distance values of 13.8  ±  6.4  mm (aver- parcellation system (CPS) including all stimulated spots (n = 46). Additionally, the four identified hotspots are marked age of all hotspots to ventral points, range 0.7–30.7 mm), Sollmann et al. BMC Neurosci (2016) 17:67 Page 5 of 10 Fig. 3 Spatial resolution measurement. Illustration of the distance measurement procedure on a subject’s individual coronal magnetic resonance imaging (MRI) slice for the apical and caudal points to the hotspot (a), and on a sagittal MRI slice for the distance between the hotspot and the ven‑ tral and dorsal spots (b). The black lines represent the distance between the hotspot (H) and the corresponding apical (a) or dorsal point (b) parallel to the cortical surface Table 2 Subject and mapping characteristics minimum, and maximum distance values of the meas- urements between each particular hotspot and the cor- Mean ± SD Range responding adjacent points. Furthermore, the results of Age (years) 24.6 ± 1.7 21.8–29.4 measurements between all hotspots taken together and EHI score 84.3 ± 13.2 57–100 adjacent points are illustrated in Fig. 4. Pain ( VAS) According to ANOVA, statistically significant differ - Convexity 2.2 ± 1.6 0–6 ences regarding the measurements between the hotspots Temporal 5.0 ± 2.0 2–10 and adjacent points were revealed with respect to hotspot Correct baseline objects 93.7 ± 3.3 87–98 1, 2, and 3 (F = 16.5, F = 11.7, F = 8.8, p < 0.0001, Table 3), RMT (% output) 33.5 ± 5.1 24–43 whereas no statistically significant difference was revealed Electric field strength ( V/m) concerning hotspot 4 (F  =  2.1, p  =  0.1111, Table  3). Hotspot 1 62.0 ± 9.8 48–84 Regarding hotspot 1, there was a statistically significant Hotspot 2 58.0 ± 11.5 40–83 difference in measurements for the hotspot to ventral ver - Hotspot 3 63.0 ± 10.3 42–87 sus apical points (CI −9.6 to −3.7), to dorsal versus api- Hotspot 4 65.0 ± 14.1 42–90 cal points (CI −10.0 to −4.1), and to apical versus caudal points (CI 1.2  to  7.2), whereas statistically significant dif - Overview about subject-related characteristics including age, Edinburg Handedness Inventory (EHI) scores, and discomfort during stimulation according ferences were revealed for the hotspot to ventral versus to the visual analogue scale (VAS). Moreover, the individual amount of correctly dorsal (CI 1.0  to  6.9), to dorsal versus apical (CI −9.5 to named objects during baseline testing, the resting motor threshold (RMT ), and the electrical field strength that was applied to the four hotspots during −3.6), and to dorsal versus caudal points (CI −6.6 to −0.7) stimulation are shown. In this context, the electrical field strength for each for hotspot 2. Concerning hotspot 3, there was a statisti- mapping point was automatically calculated and saved by the system cally significant difference in measurements for the hot - spot to dorsal versus apical points (CI −11.8 to −3.6) and 10.8  ±  4.8  mm (average of all hotspots to dorsal points, to dorsal versus caudal points (CI −9.5 to −1.3). range 2.0–26.5  mm), 16.6  ±  4.8  mm (average of all hotspots to apical points, range 0.9–27.5  mm), and Discussion 13.8 ± 4.3 mm (average of all hotspots to caudal points, Current knowledge range 2.0–24.2  mm) were measured. More detailed, Only a few publications investigated the resolution of Table  3 provides information about the mean  ±  SD, TMS up to now [21–24]. The publications of Opitz et al. Sollmann et al. BMC Neurosci (2016) 17:67 Page 6 of 10 Table 3 Measurement results Hotspot Adjacent spots Mean SD MIN MAX p value 1 Ventral 11.9 2.2 8.6 16.8 <0.0001 Dorsal 11.5 5.4 6.0 26.5 Apical 18.6 2.5 14.7 22.1 Caudal 14.4 2.8 9.0 17.8 2 Ventral 18.0 3.9 11.6 25.8 <0.0001 Dorsal 14.1 3.1 9.3 20.1 Apical 20.6 3.7 14.1 27.5 Caudal 17.7 2.9 13.8 24.2 3 Ventral 10.7 8.3 0.7 30.7 <0.0001 Dorsal 6.9 2.2 2.0 11.0 Apical 14.6 2.8 7.2 20.9 Caudal 12.3 3.0 6.5 18.2 4 Ventral 14.5 6.8 1.9 27.5 0.1111 Dorsal 10.7 4.9 3.8 18.4 Apical 12.6 5.1 0.9 21.0 Caudal 10.9 4.7 2.0 17.8 All directions 1 14.1 4.4 6.0 26.5 – 2 17.6 4.1 9.3 27.5 3 11.1 5.5 0.7 30.7 4 12.2 5.5 0.9 27.5 All hotspots and directions 13.8 5.5 0.7 30.7 – This table provides information about the spatial resolution measurements (in mm) by showing the distances from the four different hotspots to the four corresponding adjacent points (hotspot to ventral, dorsal, apical, and caudal spots), presented as mean ± standard deviation (SD), minimum (MIN), and maximum (MAX) values. In addition, one-way analysis of variance (ANOVA) was performed to compare distance measurements between hotspots and adjacent mapping points, leading to the p values presented in the table. In this context, a p value of <0.05 was considered statistically significant means that the focality of the electrical field is increased [22]. The authors Bijsterbosch et  al. [23] systematically investigated the impact of various TMS coil positions on electrical field shaping, finding that, at least for most coil positions, the induced field includes the target region of stimulation but is not distinctly restricted to it. Accord- ing to their publication, the distribution of subarachnoid cerebral fluid and gyral geometry predominantly influ - ence the modelling of the electrical field in the human brain [23]. Although the described publications represent exten- sive and valuable contributions to the current knowledge about the spatial resolution of TMS, they neither explic- Fig. 4 Measurement results. Boxplots with median, min‑, and max ‑ whiskers, and quartile‑boxes for the hotspot to ventral, dorsal, apical, itly focused on cortical language mapping, nor did they and caudal distance measurements (in mm) examine the spatial resolution from a functional point of view, meaning that they did not particularly aim for a dif- ferentiation of functionally relevant from irrelevant corti- [21] and Thielscher et  al. [22] calculated the electri - cal spots [21–23]. Under these premises, examination of cal field induced by TMS in the human brain and came the spatial resolution of rTMS-based language mapping to the conclusion that the field strength is significantly seemed mandatory. enhanced when the currents run perpendicular to the stimulated gyrus. Moreover, this effect was demonstrated Gyrus identification to be primarily restricted to the gyral lips and crowns, The average width of a cortical gyrus in the adult human but it did not extend into the walls of the sulci, which brain varies between 10 and 20  mm, although there can Sollmann et al. BMC Neurosci (2016) 17:67 Page 7 of 10 be distinct inter-individual variations. Therefore, when language-related task performance to the signal obtained the central point of a figure-of-eight TMS coil is placed during resting-state measurement within the same sub- perpendicular to the subject’s skull and in the middle of a ject [25, 26]. Then, the systematic comparison of both certain gyrus, the center of stimulation should primarily signal datasets allows drawing conclusions about the hit the stimulated gyrus, as our mean spatial resolution localization of individual language-related brain areas. measurements for each of the four directions accounted Overall, fMRI is characterized by comparatively good for values clearly under 20  mm (Table  3; Fig.  4). As an spatial resolution, and a recent study on nTMS-based important result, we can state that specific gyri can be motor mapping has demonstrated that the spatial differ - targeted and identified by rTMS, at least when stimula - ences between nTMS- and DCS-positive cortical spots tion is applied to the gyrus center with the protocol used and fMRI- and DCS-positive points are both in the range in the present study. of millimeters, although better results were observed Furthermore, because the spatial resolution of rTMS for nTMS- versus DCS-positive spots (10.5  ±  5.7 vs. does not exceed the average gyrus width and a particu- 15.0 ± 7.6 mm) [27]. lar gyrus can be targeted by this modality, the common Another neuroimaging modality for the investigation of practice of parcellating the cortical surface into subre- cortical language representation is magnetoencephalog- gions according to its gyral structure (e.g., as is done for raphy (MEG), which has been particularly used for deter- the CPS) seems to be a reasonable approach for rTMS mination of language lateralization over the last years research. [28, 29]. In this context, cortical activation in response to a language-related task performance goes along with Electrical field strength a local rise in neuronal signaling, which is characterized In general, a higher electrical field strength induced by by an increased flow of intracellular ions mediating asso - rTMS should principally cause a lower spatial resolution ciated magnetic fields that can be measured at the scalp and vice versa. The mean electrical field strengths at the surface in the form of event-related potentials [30]. How- four hotspots were comparable, and there was no statisti- ever MEG has failed to show sufficient correlation with cally significant difference revealed. Thus, our approach language maps generated by rTMS or DCS, at least partly including a comparison of hotspots being located within due to its lower spatial resolution, which is typically lim- different remote cortical gyri seems reasonable because ited to several centimeters [1, 31]. the applied electrical field does not significantly  vary When it comes to DCS, which is regarded as the cur- between these spots, and therefore, minimum measured rent gold standard for mapping human cortex function, distances seem comparable. However, statistical analysis language performance has been repeatedly tested dur- of the four hotspots revealed a statistically significant dif - ing awake surgery [32–34]. Interestingly, Haglund et  al. ference of the mean distances between the hotspots and [33] showed that radical brain tumor resection without adjacent stimulation spots (Table  3), highlighting that postoperative permanent language deficits can only be the mean differences in distance measurements to the achieved reliably when the resection border is at least hotspot were different for the respective adjacent map - 10  mm away from the nearest language site determined ping points. This might show that spatial resolution of by DCS. Thus, 10 mm seems to be the spatial resolution rTMS for language mapping—at least within a certain of DCS-based language mapping during awake surgery. range—depends on the localization of cortical stimula- As our findings show, the spatial resolution of rTMS- tion, although the applied electrical field strength does based language mapping is slightly above 10 mm (Table 3; not significantly change during mapping. Yet, this effect Fig. 4), which is comparable to the results of DCS [33]. might also be due to the chosen distribution of our corti- cal stimulation spots, and we can state that the minimally Further considerations measured spatial resolution of rTMS was 11.1 ± 5.5 mm Regarding the distribution of no responses across the (Table 3). hemisphere, the opIFG and trIFG, which should roughly overlap with the classical Broca’s area, were not charac- Comparison to other modalities terized by a high rate of errors during rTMS. However, Overall, cortical language distribution has been under these regions showed error rates to a higher extent when extensive investigation during the last decades. In this it comes to other error categories, which were incorpo- context, one frequently used technique for the identi- rated into analysis of the same dataset in other publica- fication of cortical sites related to language function is tions focusing on different purposes [11, 12]. Hence, functional magnetic resonance imaging (fMRI). In its considerable error rates were achieved during mapping of most common form, this approach compares the blood the opIFG and trIFG, but belonging to other types than oxygenation level dependent (BOLD) signal during a no responses. As a possible explanation for the low rate Sollmann et al. BMC Neurosci (2016) 17:67 Page 8 of 10 of no responses in these regions when compared to pre- mapping parameters, like the stimulation frequency or vious investigations [6–8, 20], the stimulation frequency the angulation and shape of the magnetic coil, for exam- of 7 Hz (instead of 5 Hz) has to be considered, but the set ple [1, 11, 17, 36]. As a consequence, it cannot be the of objects, which was different when compared to these claim of the present study to define the definite spatial previous publications, might have also played a role. resolution of rTMS-based language mapping indepen- However, the distinct cause for the specific pattern of no dently from setup factors. Instead, this study at least pro- responses observed in the present study cannot be fully vides some practical orientation on the minimum spatial assessed within the scope of our approach. resolution for a commonly used mapping protocol, which Although the resolution measurements of the pre- has revealed to be reliable in terms of inducing transient sent study can be discussed in the light of  previous lit- errors during object naming while being well tolerable erature on DCS-based language mapping, there is no for the individual subject. opportunity to directly compare rTMS to DCS results Another potential limitation could be the fact that on an intra-individual basis in healthy subjects due to distance measurement was only performed for prede- the highly invasive character of DCS. As a consequence, fined cortical points. In this context, the rTMS resolu - the language-positive areas mapped in the present study tion measurement was limited by the distance between cannot be verified by the gold standard method, and a single stimulation points of the CPS, meaning that areas final decision about the possible essential character of a in-between the different stimulation spots were not sys - particular language-positive site cannot be made due to tematically examined. Thus, the results presented in the the comparatively high sensitivity and low specificity of current approach reflect a minimum resolution measure - rTMS [10]. However, this limitation can at least partly ment. Consequently, it might be possible that the defi - be overcome by the results of recent studies that showed nite resolution of rTMS is finer-grained. Yet, all cortical a much higher specificity and positive predictive value points of stimulation were placed manually at the center between language maps generated by both mapping tech- of the gyri, meaning that finer-grained mapping might niques using the CPS, which again provides spatial reso- also have targeted sulci instead of cortical tissue, which lution data on a gyri level [1, 2]. To draw more definite is generally not regarded as adequate for the rTMS-based conclusions concerning resolution, a prospective study language mapping approach per se [6, 8]. We decided to including preoperative rTMS and intraoperative DCS follow the presented kind of approach because one major mapping among patients should be the next step. objective was to investigate the required cortical distance In the recent years, the differentiation of naming errors to distinguish between hotspots (stimulation points with into various predefined error categories has been dem - a high error rate) and adjacent mapping points (stimu- onstrated repeatedly within the scope of rTMS and DCS lation points with low error rates) in a practical setup. trials [4, 6, 8, 16, 20, 34]. Although other error categories Basically, this method is similar to the measurement pre- except no responses are based on stimulation-induced sented in Haglund et al., which represents a key reference disruption of important language subfunctions as well, for the accuracy of DCS-based language mapping during we decided to not take these error types into account awake surgery [33]. since no responses have proven to be among the most frequent error types [6, 20]. Incorporation of other error Conclusions types could have led to different hotspots, but poten - The present study examined the spatial resolution of tially on the basis of more unclear language disruption. rTMS-based language mapping via distance meas- Therefore, our current approach seems to be reasonable urement. In this context, rTMS is able to differenti - within the scope of one of the first studies regarding the ate between hotspots (stimulation points with a high exploration of rTMS-based language mapping resolution. error rate) and adjacent mapping points (stimulation However, we have to be aware of the fact that upcoming points with low error rates) at a minimum distance studies systematically analyzing other categories might of 11.1  ±  5.5  mm as measured for a parietal hotspot. further refine our minimum resolution measurements. According to these distance measurement results, the Additionally, the spatial resolution of rTMS should also spatial resolution of rTMS should principally allow for be evaluated for mapping of other cognitive functions the identification of a particular gyrus as it is the case for since such approaches are currently emerging. In this DCS-based language mapping during awake surgery as context, non-invasive assessment of calculation functions the current gold standard. by rTMS has been successfully performed recently, but spatial resolution for this purpose is vastly unknown [35]. Abbreviations Furthermore, it is already known that rTMS mapping 2‑D: two ‑ dimensional; 3‑D: three ‑ dimensional; aMTG: anterior middle results and electrical field shaping depend on various temporal gyrus; ANOVA: analysis of variance; ADM: abductor digiti minimi; Sollmann et al. BMC Neurosci (2016) 17:67 Page 9 of 10 APB: abductor pollicis brevis; BOLD: blood oxygenation level dependent; CI: 2. Krieg SM, Tarapore PE, Picht T, Tanigawa N, Houde J, Sollmann N, Meyer confidence interval; CPS: cortical parcellation system; DT: display time; DCS: B, Vajkoczy P, Berger MS, Ringel F, et al. Optimal timing of pulse onset direct cortical stimulation; EHI: Edinburgh Handedness Inventory; EMG: elec‑ for language mapping with navigated repetitive transcranial magnetic tromyography; fMRI: functional magnetic resonance imaging; IPI: inter‑picture stimulation. NeuroImage. 2014;100:219–36. interval; ITG: inferior temporal gyrus; MEG: magnetoencephalography; MEP: 3. Picht T, Schulz J, Vajkoczy P. The preoperative use of navigated motor evoked potential; MRI: magnetic resonance imaging; nTMS: navigated transcranial magnetic stimulation facilitates early resection of sus‑ transcranial magnetic stimulation; orIFG: orbital part of the inferior frontal pected low‑ grade gliomas in the motor cortex. Acta Neurochir ( Wien). gyrus; polIFG: polar inferior frontal gyrus; polMFG: polar middle frontal gyrus; 2013;155(10):1813–21. polMTG: polar middle temporal gyrus; polSFG: polar superior frontal gyrus; 4. Rosler J, Niraula B, Strack V, Zdunczyk A, Schilt S, Savolainen P, Lioumis polSTG: polar superior temporal gyrus; PTI: picture‑to ‑trigger interval; RMT: P, Makela J, Vajkoczy P, Frey D, et al. Language mapping in healthy resting motor threshold; rTMS: repetitive navigated transcranial magnetic volunteers and brain tumor patients with a novel navigated TMS stimulation; SD: standard deviation; TMS: transcranial magnetic stimulation; system: evidence of tumor‑induced plasticity. Clin Neurophysiol. VAS: visual analogue scale. 2014;125(3):526–36. 5. Picht T, Mularski S, Kuehn B, Vajkoczy P, Kombos T, Suess O. Navigated Authors’ contributions transcranial magnetic stimulation for preoperative functional diagnostics NS: manuscript preparation, data acquisition, data handling, data analysis, in brain tumor surgery. Neurosurgery. 2009;65(6 Suppl):93–8 (discussion statistics, literature review. TH, LT, SI, SM, TBB: data acquisition, data handling, 98–9). data analysis. FR, BM, SK: data acquisition, data handling, data analysis, study 6. Sollmann N, Tanigawa N, Ringel F, Zimmer C, Meyer B, Krieg SM. supervision. All authors read and approved the final manuscript. Language and its right‑hemispheric distribution in healthy brains: an investigation by repetitive transcranial magnetic stimulation. NeuroIm‑ Authors’ information age. 2014;102P2:776–88. Nico Sollmann is a Ph.D. student at the Department of Neurosurgery. 7. Hernandez‑Pavon JC, Makela N, Lehtinen H, Lioumis P, Makela JP. Eec ff ts Theresa Hauck is a medical doctor and lab intern. Lorena Tussis is a master of navigated TMS on object and action naming. Front Hum Neurosci. student and lab intern. Sebastian Ille and Stefanie Maurer are residents 2014;8:660. at the Department of Neurosurgery. Tobias Boeckh‑Behrens is an attend‑ 8. Lioumis P, Zhdanov A, Makela N, Lehtinen H, Wilenius J, Neuvonen T, Han‑ ing neuroradiologist, and Sandro M. Krieg is an attending neurosurgeon. nula H, Deletis V, Picht T, Makela JP. A novel approach for documenting Florian Ringel is vice chief, and Bernhard Meyer is chief of the Department of naming errors induced by navigated transcranial magnetic stimulation. J Neurosurgery. Neurosci Methods. 2012;204(2):349–54. 9. Pascual‑Leone A, Gates JR, Dhuna A. Induction of speech arrest and Author details counting errors with rapid‑rate transcranial magnetic stimulation. Neurol‑ Department of Neurosurgery, Klinikum rechts der Isar, Technische Universität ogy. 1991;41(5):697–702. München, Ismaninger Str. 22, 81675 Munich, Germany. TUM‑Neuroimaging 10. Picht T, Krieg SM, Sollmann N, Rosler J, Niraula B, Neuvonen T, Savol‑ Center, Klinikum rechts der Isar, Technische Universität München, Ismaninger ainen P, Lioumis P, Makela JP, Deletis V, et al. A comparison of language Str. 22, 81675 Munich, Germany. Section of Neuroradiology, Department mapping by preoperative navigated transcranial magnetic stimulation of Radiology, Klinikum rechts der Isar, Technische Universität München, Isma‑ and direct cortical stimulation during awake surgery. Neurosurgery. ninger Str. 22, 81675 Munich, Germany. 2013;72(5):808–19. 11. Hauck T, Tanigawa N, Probst M, Wohlschlaeger A, Ille S, Sollmann N, Mau‑ Acknowledgements rer S, Zimmer C, Ringel F, Meyer B, et al. Stimulation frequency determines Nico Sollmann gratefully acknowledges the support of the graduate school of the distribution of language positive cortical regions during navigated our university. transcranial magnetic brain stimulation. BMC Neurosci. 2015;16(1):5. 12. Hauck T, Tanigawa N, Probst M, Wohlschlaeger A, Ille S, Sollmann N, Competing interests Maurer S, Zimmer C, Ringel F, Meyer B, et al. Task type affects location of FR and SK are consultants for BrainLAB AG (Feldkirchen, Germany). SK is con‑ language‑positive cortical regions by repetitive navigated transcranial sultant for Nexstim Oy (Helsinki, Finland). The authors declare that they have magnetic stimulation mapping. PLoS ONE. 2015;10(4):e0125298. no competing interests. 13. Snodgrass JG, Vanderwart M. A standardized set of 260 pictures: norms for name agreement, image agreement, familiarity, and visual complexity. Availability of data and materials J Exp Psychol Hum Learn. 1980;6(2):174–215. All data used for analysis are presented in the manuscript. The discussion 14. Niskanen E, Julkunen P, Saisanen L, Vanninen R, Karjalainen P, Kononen M. and conclusions only rely on the data presented. Raw mapping data can be Group‑level variations in motor representation areas of thenar and ante ‑ provided upon request. rior tibial muscles: Navigated Transcranial Magnetic Stimulation Study. Hum Brain Mapp. 2010;31(8):1272–80. Ethics approval and consent to participate 15. Rossini PM, Barker AT, Berardelli A, Caramia MD, Caruso G, Cracco RQ, The local ethical committee of the Technische Universität München approved Dimitrijevic MR, Hallett M, Katayama Y, Lucking CH, et al. Non‑invasive the experimental procedures (Registration Number: 2793/10) in accordance electrical and magnetic stimulation of the brain, spinal cord and with the Declaration of Helsinki. Written informed consent was obtained from roots: basic principles and procedures for routine clinical application. all volunteers. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol. 1994;91(2):79–92. Funding 16. Corina DP, Loudermilk BC, Detwiler L, Martin RF, Brinkley JF, Ojemann The study was financed by institutional grants from the Department of Neuro ‑ G. Analysis of naming errors during cortical stimulation mapping: surgery and the Section of Neuroradiology. implications for models of language representation. Brain Lang. 2010;115(2):101–12. Received: 17 June 2016 Accepted: 19 October 2016 17. Epstein CM, Lah JJ, Meador K, Weissman JD, Gaitan LE, Dihenia B. Opti‑ mum stimulus parameters for lateralized suppression of speech with magnetic brain stimulation. Neurology. 1996;47(6):1590–3. 18. Ruohonen J, Karhu J. Navigated transcranial magnetic stimulation. Neuro‑ physiol Clin. 2010;40(1):7–17. 19. Picht T, Schmidt S, Brandt S, Frey D, Hannula H, Neuvonen T, Karhu References J, Vajkoczy P, Suess O. Preoperative functional mapping for rolandic 1. Tarapore PE, Findlay AM, Honma SM, Mizuiri D, Houde JF, Berger MS, brain tumor surgery: comparison of navigated transcranial magnetic Nagarajan SS. Language mapping with navigated repetitive TMS: proof of stimulation to direct cortical stimulation. Neurosurgery. 2011;69(3):581–8 technique and validation. NeuroImage. 2013;82:260–72. (discussion 588). Sollmann et al. BMC Neurosci (2016) 17:67 Page 10 of 10 20. Krieg SM, Sollmann N, Tanigawa N, Foerschler A, Meyer B, Ringel F: Corti‑ dominance for language assessed by magnetoencephalographic imag‑ cal distribution of speech and language errors investigated by visual ing. Ann Neurol. 2012;71(5):668–86. object naming and navigated transcranial magnetic stimulation. Brain 29. Hirata M, Goto T, Barnes G, Umekawa Y, Yanagisawa T, Kato A, Oshino Struct Funct. 2016;221(4):2259–86. S, Kishima H, Hashimoto N, Saitoh Y, et al. Language dominance and 21. Opitz A, Windhoff M, Heidemann RM, Turner R, Thielscher A. How the mapping based on neuromagnetic oscillatory changes: comparison with brain tissue shapes the electric field induced by transcranial magnetic invasive procedures. J Neurosurg. 2010;112(3):528–38. stimulation. NeuroImage. 2011;58(3):849–59. 30. Frye RE, Rezaie R, Papanicolaou AC. Functional neuroimaging of language 22. Thielscher A, Opitz A, Windhoff M. Impact of the gyral geometry on the using magnetoencephalography. Phys Life Rev. 2009;6(1):1–10. electric field induced by transcranial magnetic stimulation. NeuroImage. 31. Lev MH, Grant PE. MEG versus BOLD MR imaging: functional imaging, the 2011;54(1):234–43. next generation? AJNR Am J Neuroradiol. 2000;21(8):1369–70. 23. Bijsterbosch JD, Barker AT, Lee KH, Woodruff PW. Where does transcranial 32. Tate MC, Herbet G, Moritz‑ Gasser S, Tate JE, Duffau H. Probabilistic map magnetic stimulation ( TMS) stimulate? Modelling of induced field maps of critical functional regions of the human cerebral cortex: Broca’s area for some common cortical and cerebellar targets. Med Biol Eng Comput. revisited. Brain. 2014;137(Pt 10):2773–82. 2012;50(7):671–81. 33. Haglund MM, Berger MS, Shamseldin M, Lettich E, Ojemann GA. Cortical 24. Walsh V, Cowey A. Transcranial magnetic stimulation and cognitive neu‑ localization of temporal lobe language sites in patients with gliomas. roscience. Nat Rev Neurosci. 2000;1(1):73–9. Neurosurgery. 1994;34(4):567–76 (discussion 576). 25. Binder JR, Frost JA, Hammeke TA, Cox RW, Rao SM, Prieto T. Human brain 34. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after language map‑ language areas identified by functional magnetic resonance imaging. J ping for glioma resection. N Engl J Med. 2008;358(1):18–27. Neurosci. 1997;17(1):353–62. 35. Maurer S, Tanigawa N, Sollmann N, Hauck T, Ille S, Boeckh‑Behrens T, 26. Binder JR. Functional MRI is a valid noninvasive alternative to Wada test‑ Meyer B, Krieg SM: Non‑invasive mapping of calculation function by ing. Epilepsy Behav. 2011;20(2):214–22. repetitive navigated transcranial magnetic stimulation. Brain Struct Funct. 27. Forster MT, Hattingen E, Senft C, Gasser T, Seifert V, Szelenyi A. Navigated 2016;221(8):3927–3947. transcranial magnetic stimulation and functional magnetic resonance 36. Bolognini N, Ro T. Transcranial magnetic stimulation: disrupt‑ imaging: advanced adjuncts in preoperative planning for central region ing neural activity to alter and assess brain function. J Neurosci. tumors. Neurosurgery. 2011;68(5):1317–24 (discussion 1324–5). 2010;30(29):9647–50. 28. Findlay AM, Ambrose JB, Cahn‑ Weiner DA, Houde JF, Honma S, Hinkley LB, Berger MS, Nagarajan SS, Kirsch HE. Dynamics of hemispheric Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Journal

BMC Neuroscience – Springer Journals

Published: Oct 24, 2016

There are no references for this article.