Frontal Alpha Asymmetry, a Potential Biomarker for the Effect of Neuromodulation on Brain’s Affective Circuitry - Preliminary Evidence from a Deep Brain Stimulation Study

ORIGINAL RESEARCH published: 04 Decembe 2017 doi: 10.3389/ nhum.2017.00584 F on al Alpha Asymme y, a Po en ial Bioma ke o he E ec o Neu omodula ion on B ain’s A ec i e Ci cui y—P elimina y E idence om a Deep B ain S imula ion S udy Lihua Sun1,2,Ja i Pe äkylä1,3 and Kaisa M. Ha ikainen1,3* 1Beha io al Neu ology Resea ch Uni , Tampe e Uni e si y Hospi al, Tampe e, Finland, 2Tu ku PET Cen e , Uni e si y o Tu ku, Tu ku, Finland, 3Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e, Tampe e, Finland Edi ed by: Juliana Yo dano a, Ins i u e o Neu obiology (BAS), Bulga ia Re iewed by: Gonzalo Ala con, King’s College London, Uni ed Kingdom Baha Gün ekin, School o In e na ional Medicine, Is anbul Medipol Uni e si y, Tu key *Co espondence: Kaisa M. Ha ikainen [email p o ec ed] Recei ed: 19 June 2017 Accep ed: 17 No embe 2017 Published: 04 Decembe 2017 Ci a ion: Sun L, Pe äkylä J and Ha ikainen KM (2017) F on al Alpha Asymme y, a Po en ial Bioma ke o he E ec o Neu omodula ion on B ain’s A ec i e Ci cui y—P elimina y E idence om a Deep B ain S imula ion S udy. F on . Hum. Neu osci. 11:584. doi: 10.3389/ nhum.2017.00584 Neu omodula ion echniques a ge ing limbic ci cui s can be used o ea e ac o y psychia ic o neu ological diso de s. Howe e , objec i e measu e o he impac o neu omodula ion on a ec i e b ain ci cui s is lacking. Deep b ain s imula ion a a key node o he limbic ci cui , he an e io halamic nuclei (ANT-DBS), is used o ea e ac o y epilepsy. While e ec i e in educing seizu es, pa ien s ha e epo ed subjec i e dep essi e symp oms as a side e ec . In line wi h po en ial ulne abili y o dep ession, we ha e p e iously shown ANT-DBS o inc ease a en ion alloca ion o h ea e idenced by beha io and b ain physiology. Righ wa d on al alpha asymme y wi h g ea e igh hemisphe ic ac i a ion is hough o e lec b ain physiology linked wi h dep ession and anxie y. To ha end, we in es iga ed whe he high- equency elec ic s imula ion a ANT in luences on al alpha asymme y. Fu he mo e, we explo ed he impac o DBS on emo ional modula ion o on al alpha asymme y and whe he i is linked wi h emo ional modula ion o esponse speed. Elec ical s imula ion a ANT led o an inc eased igh wa d on al alpha asymme y compa ed o si ua ions whe e s imula ion was o (F(1,12)= 14.09, p= 0.003) o he halamic con ol loca ion was s imula ed (F(1,12)= 10.19, p= 0.008), along wi h p olonged eac ion imes in he con ex o emo ional dis ac o s (F(1,7)= 16.66, p= 0.005). The change was speci ically d i en by inc eased ac i i y in he igh hemisphe e. Fu he mo e, we ound a co ela ion be ween he emo ional modula ion o on al alpha asymme y and emo ional in e e ence o esponse speed due o ANT s imula ion ( = 0.78, p= 0.02). In conclusion, DBS a ANT inc eased ela i e igh hemisphe ic ac i i y and his was linked wi h emo ional modula ion o beha io . P e ious s udies ha e linked on al alpha asymme y wi h emo ion ela ed symp oms and u he mo e, Vagus Ne e S imula ion (VNS) has been shown o modula e alpha asymme y. Thus, in he ligh o he p e ious li e a u e and he cu en indings, we sugges ha on al alpha asymme y along wi h emo ional in e e ence o esponse speed migh be a easible bioma ke o he e ec s o neu omodula ion on b ain’s a ec i e ci cui y in gene al. Keywo ds: deep b ain s imula ion, neu omodula ion, on al alpha asymme y, emo ion, bioma ke , epilepsy, an e io halamic nuclei, an e io nucleus o halamus F on ie s in Human Neu oscience | www. on ie sin.o g 1Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y INTRODUCTION Neu omodula ion echniques a e inc easingly used o ea ing se e e neu ological and psychia ic diso de s. Fo ins ance, deep b ain s imula ion (DBS) a he an e io halamic nuclei (ANT) is used o ea ing e ac o y epilepsy wi h la ge scale clinical s udies indica ing signi ican bene i s in educing seizu e equency (Fishe e al., 2010) and demons a ing sus ained bene i s wi hin a 5-yea ollow-up pe iod (Salano a e al., 2015). Along wi h ea men e ec s ANT-DBS has been linked wi h ad e se mood ela ed e ec s such as subjec i e dep ession- ela ed symp oms (Fishe e al., 2010). A ec i e ad e se e ec s a e no su p ising conside ing he ea men is a ge ing ANT, which is a key node in he limbic ci cui y. To ha end he e is a call o objec i e bioma ke s o he impac o neu omodula ion on a ec i e b ain unc ions. Bioma ke s which allow guiding neu omodula ion ea men s, o example pa ame e selec ion owa ds he op imal ea men e ec and minimal cogni i e and a ec i e side e ec s, ha e g ea clinical signi icance. The impo ance o such bioma ke s is emphasized especially when neu omodula ion is used o ea emo ional dys unc ions, o example dep ession. EEG p o ides a easible me hod o assessing online al e a ions in human b ain physiology and i is equen ly used in bo h clinical and esea ch se ings. Thus, i is no su p ising ha di e en EEG measu es ha e been sugges ed as po en ial bioma ke s o a ie y o b ain diso de s and o he impac o neu omodula ion on b ain unc ions (Coan and Allen, 2004; Thibodeau e al., 2006; G imshaw e al., 2014; Sun e al., 2017), as well as o guiding DBS elec ode a ge ing using co ical esponses o deep b ain s imuli p o iding in o ma ion abou op imal elec ode loca ions (Valen ín e al., 2013). Measu es quan i ying EEG equency in o ma ion such as he a conco dance ha e been sugges ed o ha e u ili y as a po en ial bioma ke o ea men esponse in DBS a subcallosal cingula e o ea men - esis an dep ession (B oadway e al., 2012). The p o ile o asymme ic EEG ac i i y o di e en equency bands has been sugges ed o e lec po en ial esponse o DBS a subgenual cingula e co ex (Qu aan e al., 2014). The sugges ed measu es ha e shown po en ial in dis inguishing esponde s om non- esponde s, which is impo an when selec ing pa ien s who will mos likely bene i om an in asi e neu omodula ion ea men . I is equally impo an o ha e bioma ke s o he immedia e impac o neu omodula ion on a ec i e unc ions. Such bioma ke s will allow pe sonally ailo ed neu omodula ion pa ame e adjus men s and p o ide immedia e eedback abou he impac o chosen pa ame e s on he a ec i e b ain ne wo ks and a ec i e unc ions enabling online op imiza ion o he s imula ion pa ame e s. In he cu en clinical p ac ice pa ame e s a e adjus ed o e he cou se o se e al mon hs o ollow up based on subjec i e epo s a he han objec i e bioma ke s e lec ing he impac o neu omodula ion on b ain’s a ec i e ci cui s. In sea ch o online bioma ke s o neu omodula ion echniques, we ha e p e iously shown ha Vagus Ne e S imula ion (VNS), a neu omodula ion echnique used o he ea men o e ac o y epilepsy and dep ession, has immedia e modula o y e ec on b ain’s on al alpha asymme y (Sun e al., 2017). We compa ed he on al asymme y sco e when cyclic VNS was u ned on s. when i was u ned o and ound ha VNS led o inc eased igh wa d on al alpha asymme y in he con ex o emo ional dis ac o s. In he con ex o neu al dis ac o s, howe e , VNS had no e ec on on al alpha asymme y. This inding e ealed an immedia e modula o y e ec o VNS on emo ion- ela ed b ain esponses and highligh ed he po en ial u ili y o on al alpha asymme y as an online bioma ke o he impac o neu omodula ion on b ain’s a ec i e unc ions. F on al EEG alpha asymme y indica ing inc eased ela i e igh hemisphe ic ac i i y has been equen ly linked wi h dep ession and anxie y (Coan and Allen, 2004; Thibodeau e al., 2006) e en hough he ecen me a-analysis sugges s qui e a iable and un eliable e ec s ( an de Vinne e al., 2017). Righ wa d on al EEG alpha band asymme y, i.e., ela i ely inc eased igh hemisphe e ac i a ion, has been epo ed in bo h cu en ly and p e iously dep essed subjec s compa ed o ne e dep essed subjec s (Hen iques and Da idson, 1990, 1991; Go lib, 1998; S ewa e al., 2011). The same pa e n o on al asymme ic ac i a ion is ound in in an s o mo he s wi h dep ession (Field and Diego, 2008; Wen e al., 2017). Also, igh wa d shi in on al asymme y is linked wi h cu en anxie y and i can po en ially p edic u u e anxie y (Toma ken and Da idson, 1994; Blackha e al., 2006; Ma he sul e al., 2008). Bo h es ing s a e and ask- ela ed on al alpha asymme y ha e been s udied. Res ing s a e on al alpha asymme y is eco ded when subjec s a e no engaged in beha io al asks. In con as , beha io al asks equi e a en ional engagemen and asks wi h emo ional elemen s may p omo e asymme ic ac i a ion o he b ain hemisphe es. The e o e, es ing s a e on al alpha asymme y indexes hemisphe ic b ain ac i i y a es , while ask- ela ed alpha asymme y indexes b ain ac i i y du ing a beha io al ask (Coan and Allen, 2004). I has been sugges ed ha ask- ela ed on al alpha asymme y may be a mo e powe ul indica o o p edisposi ion owa ds psychopa hology han es ing s a e alpha asymme y, possibly due o elimina ion o he uncon olled a iance o each subjec a es ing s a e (Coan e al., 2006; S ewa e al., 2014). Also, gi en he high demand o a en ion in beha io al asks, he ask- ela ed on al alpha asymme y may be e link b ain’s emo ional-p ocessing- ela ed asymme ic ac i a ion wi h cogni i e p ocesses, such as a en ion and execu i e con ol (G imshaw e al., 2014). P e iously, we ha e s udied he e ec s o ANT-DBS on human a en ion and execu i e unc ions wi h objec i e measu es using a compu e -based Execu i e-RT es (Ha ikainen e al., 2010, 2014). The expe imen al es has been de eloped based on ou p e ious neu ocogni i e s udies on heal hy subjec s and pa ien s wi h on al lesions while employing eac ion ime and e en - ela ed po en ial measu es (Ha ikainen e al., 2000, 2007; Ha ikainen and Knigh , 2003). The Execu i e-RT es equi es mul iple cogni i e p ocesses o be engaged simul aneously and has been shown o de ec sub le changes in a en ion and emo ion-a en ion in e ac ion due o di e en ypes o neu omodula ion F on ie s in Human Neu oscience | www. on ie sin.o g 2Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y echniques (Ha ikainen e al., 2014;Sun e al., 2015, 2016, 2017; Sun, 2016; Pe äkylä e al., 2017), ao ic al e eplacemen su ge y (Liima ainen e al., 2016), mild b ain auma ic inju y (Mäki-Ma unen e al., 2015) and o bi o on al lesions (Mäki-Ma unen e al., 2017). Using his es , we ound ha ANT-DBS led o inc eased emo ional in e e ence, e idenced by inc eased eac ion imes in he con ex o emo ional dis ac o s and inc eased a en ion alloca ion o h ea - ela ed s imuli, e idenced by inc eased cen o-pa ie al N2-P3 b ain po en ials (Ha ikainen e al., 2014; Sun e al., 2015). In he cu en s udy, we in es iga ed whe he ANT-DBS p omo es immedia e changes in asymme ic on al ac i a ion and whe he hese changes a e associa ed wi h p ocessing o nega i e emo ions. Gi en ou p e ious indings wi h ANT-DBS inc easing a en ional alloca ion o emo ionally nega i e s imuli (Ha ikainen e al., 2014; Sun e al., 2015), epo s whe e ANT-DBS has been linked wi h subjec i e dep essi e symp oms (Fishe e al., 2010) and me a-analysis linking on al igh wa d alpha asymme y ac i i y wi h dep ession (Coan and Allen, 2004; Thibodeau e al., 2006), we expec ed ANT-DBS o lead o inc eased igh wa d on al alpha asymme y, i.e., ela i ely inc eased igh hemisphe e ac i a ion. MATERIALS AND METHODS Subjec s Thi een pa ien s wi h ANT-DBS ea men o e ac o y epilepsy pa icipa ed in his s udy. Twel e subjec s included in ou p e ious publica ions (Ha ikainen e al., 2014) we e complemen ed wi h one new subjec . Subjec s had on al (n= 4), empo al (n= 1), occipi al (n= 2) and mul i ocal (n= 6) epilepsies, and he main e iologies we e co ical dysplasia (n= 5) and encephali is (n= 5). Th ee subjec s had unknown e iology. Six subjec s had pa hological indings in Magne ic Resonance Imaging (MRI) and se en had no mal MRIs. All subjec s we e using one o ou an i-epilep ic medica ions. The wide he e ogenei y o he pa ien s wi h a ie y o ac o s po en ially con ibu ing o beha io and EEG was accoun ed o by a wi hin-subjec s udy design. The s udy was app o ed by he Regional Re iew Boa d a Tampe e Uni e si y Hospi al, Tampe e, Finland. All subjec s p o ided hei w i en consen and olun a ily pa icipa ed in he s udy acco ding o he guidelines se o h in he Decla a ion o Helsinki go e ning he ea men o human subjec s. Execu i e RT Tes Pa ien s pe o med he Execu i e-RT es (Ha ikainen e al., 2010), a isual a en ion ask wi h emo ional dis ac o s, while hei scalp EEG was eco ded. Du ing he es , pa ien s sa one me e away om a 2100 compu e sc een. The beha io al pa adigm was p esen ed using P esen a ion so wa e (Neu obeha io al Sys em Inc., Be keley, CA, USA). The whole es consis ed o wo sessions, each session consis ing o 16 blocks and each block consis ing o 64 wo-second ials. The onse o each ial was a iangle poin ing ei he up o down and las ing 150 ms. A e he iangle, he e was a lapse o 150 ms be o e a Go/NoGo signal. Go/NoGo signal was a a ic ligh in ei he g een o ed colo p esen ed o 150 ms. The colo o he a ic ligh indica ed whe he subjec should espond o no , i.e., a Go o a NoGo ial. In a Go ial, subjec s we e supposed o p ess one bu on using hei index (down bu on) o middle inge (up bu on) o indica e he o ien a ion o he p e iously p esen ed iangle. In a NoGo case, pa ien s should no p ess any bu ons. The meaning o he a ic ligh colo changed om block o block so ha in hal o he blocks g een indica ed Go and ed NoGo signal and in ano he hal he meaning was in e ed so ha ed indica ed Go and g een indica ed NoGo. In he middle ci cle o he a ic ligh he e was a schema ic line d awing dis ac o , which was ei he a nega i ely alenced shape o a spide o emo ionally neu al con ol dis ac o composed o he exac ly same g aphical componen s and esembling a lowe . Beha io al ou comes o he es include eac ion imes and di e en ypes o e o s. He e we epo only emo ion ela ed beha io al indings. Analysis o e o s did no gi e emo ion- ela ed signi icance, see Pe äkylä e al. (2017). S imula ion P o ocol All pa ien s ecei ed ANT-DBS ea men o e ac o y epilepsy wi h DBS leads bila e ally in he an e io halami. ‘‘ANT loca ion’’ e e s he e o elec ode loca ions inside ANT o immedia ely adjacen o ANT wi h wo subjec s ha ing ‘‘ANT loca ion’’ sligh ly ou side o ANT (see also Sun e al., 2015). One subjec used le hand o pe o ming he beha io al es . Du ing he expe imen , bipola s imula ion was used o allow be e a ge ing o ANT. The DBS lead has ou elec odes ou o which we used wo adjacen elec odes closes o ANT when ANT was s imula ed, wi h one as nega i e con ac and he o he as posi i e con ac . Con inuous elec ical s imula ion wi h a cons an 5 mA cu en , 140 Hz equency and 90 µs pulse wid h was gi en when s imula ion was ON. When s imula ion was OFF, no cu en was deli e ed. When he con ol loca ion ou side ANT was s imula ed, he wo a hes con ac s om ANT we e chosen as ac i e con ac s. The con ol loca ion was e e ed o as ‘‘ou side ANT loca ion (OA)’’ in ou p e ious s udy (Ha ikainen e al., 2014). The con ol loca ion con ac s we e ypically in adjoining in e io s uc u e o ANT, e.g., mediodo sal nucleus. The expe imen consis ed o wo sessions (see Figu e 1). In one session s imula ion was al e na i ely ON o OFF a ANT, and in he o he session s imula ion was al e na i ely ON o OFF a con ol loca ion. The s a ing o de o he sessions was coun e balanced be ween subjec s. Each s a us, e.g., ANT ON, las ed o 5–6 min o allow o comple ing wo blocks o es ing. The e o e, eigh blocks o es s we e done wi h ANT ON s a us, eigh blocks o es s we e wi h con ol loca ion ON s a us, and 16 blocks we e done wi h OFF s a us (i.e., ANT OFF and con ol loca ion OFF). F on al Alpha Asymme y EEG signal was eco ded using he 64-channel Ag/AgCl elec odes and he QuickAmp EEG ampli ie (Gilching, Ge many). EEG signal was p ocessed o line using he B ain Analyze 2 so wa e (B ain P oduc s GmbH, Ge many). Fi s , signal was down sampled o 250 Hz. Ocula mo emen s we e co ec ed using he ICA ocula co ec ion unc ion o F on ie s in Human Neu oscience | www. on ie sin.o g 3Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y FIGURE 1 | Schema ic illus a ion o he s imula ion p o ocol. (A) The DBS lead and he s imula ed si es. Each lead has ou po en ially ac i e con ac s. In bipola s imula ion, wo con ac s we e ac i e wi h one ca hode and he o he anode. Typically, supe io con ac pai was in he an e io halamic nuclei (ANT) and in e io con ac pai was conside ed o be ou side ANT assigned as he halamic con ol loca ion. (B) The whole expe imen o each subjec included 32 blocks o es ing. In he i s 16 blocks, ei he he ANT o he con ol loca ion was s imula ed, o s imula ion was u ned o . In he second 16 blocks, he o he loca ion was s imula ed, o s imula ion was u ned o . The o de o s imula ed si es was coun e balanced. DBS, deep b ain s imula ion. This igu e is adap ed om Pe äkylä e al. (2017). he Analyze 2 so wa e, whe e signals a e decomposed in o independen componen s wi h he ex ended In omax algo i hm and one o wo componen s co esponding eye mo emen s we e ejec ed. Then he signal was e- e e enced o he cen al EEG elec ode Cz and il a ed wi h bandpass il e a 3–30 Hz. EEG signal was segmen ed in o 2-s segmen s, wi h each s a ing om he onse o he ial, i.e., he iangle. EEG segmen s wi h ampli udes exceeding ±80 µV we e emo ed. The powe spec um (µV2/Hz) o all segmen s was calcula ed using he Fas Fou ie ans o ma ion and a e aged o each condi ion. Spec al powe (µV2) a he alpha ange (8–13 Hz) was expo ed. Finally, alpha powe o EEG elec odes F3 and F4 was log- ans o med and he asymme y sco e o on al alpha was calcula ed by sub ac ing he alue a F3 om he alue a F4 (log- ans o med spec al powe a F4—log- ans o med spec al powe a F3). S a is ics F on al alpha asymme y and hemisphe ic alpha ac i i y we e analyzed o he whole g oup (n= 13) and o a subg oup o subjec s who comple ed he ask wi h good pe o mance (n= 8). A subjec was conside ed a good pe o me i subjec ’s o al e o a e was below 15% (Ha ikainen e al., 2014). F on al alpha asymme y sco es and eac ion imes we e analyzed using epea ed measu es analysis o a iance using Emo ion (emo ional, neu al) and S imula ion (ANT ON, OFF, con ol loca ion ON) as ac o s. Hemisphe ic alpha ac i i y was analyzed using epea ed measu es analysis o a iance using Emo ion, S imula ion and Hemisphe e (le , igh ) as ac o s. In subg oup analysis, he ac o S imula ion has wo le els (ANT ON and OFF) in o de o con ol o mul iple compa ison. Be o e analysis, dis ibu ion o he da a was es ed o no mali y and ans o med by sub ac ing subjec mean om subjec ’s alues i necessa y. Signi ican in e ac ion e ec s we e ollowed by pos hoc analysis. Bon e oni-adjus ed signi icance c i e ia p= 0.017 was used o co ec o mul iple compa ison. All s a is ical indings we e con i med wi h non-pa ame ic s a is ical app oach using Wilcoxon signed ank es as de ailed in ou p e ious s udy (Sun e al., 2015). Co ela ion Analysis We s udied how ANT-DBS modula es emo ional p ocessing by analyzing changes ANT-DBS causes in emo ion- ela ed alpha asymme y and emo ion- ela ed eac ion imes. Sub ac ing alpha asymme y sco es in he con ex o neu al s imuli om alpha asymme y sco es in he con ex o emo ion- ela ed s imuli gi es an index o he impac o emo ional s imuli on alpha asymme y, wi h o he ac o s con ibu ing o he alpha asymme y sco es con olled o by sub ac ion. Simila ly, sub ac ing eac ion imes in con ex o neu al s imuli om RTs in con ex o emo ion- ela ed s imuli gi es an index indica ing he impac o emo ional s imuli on esponse speed. A e de i ing hese measu es o each s imula ion s a us (i.e., ANT ON and OFF) sepa a ely, we sub ac ed ob ained alues when s imula ion was OFF om he alues when ANT s imula ion was ON, esul ing in ne impac indices o ANT-DBS and emo ion- ela ed s imuli on on al alpha asymme y sco es and eac ion imes. A e de i ing he ne impac indices o on al alpha asymme y and eac ion imes, we conduc ed a co ela ion analysis whe e we co ela ed changes caused by ANT s imula ion on eac ion imes and alpha asymme y linked o emo ion. This analysis was done only o good pe o me s. Because eac ion ime and alpha asymme y sco es a ies hugely be ween subjec s, hey we e ans o med o z-sco e be o e he analysis and all calcula ions we e done using z-sco es. All s a is ical analysis was done in R s a is ics ( e sion 3.2.3). Repea ed measu e analysis o a iance was done using he ez package (Law ence, 2013). RESULTS F on al Alpha Asymme y and Hemisphe ic Ac i i y, Whole G oup Analysis o on al alpha asymme y sco es esul ed in a main e ec o S imula ion (F(2,24)= 10.01, p= 0.0007, η2 G= 0.074). To compa e he di e ence be ween s imula ion s a us, a pai wise analysis was done. S imula ion o ANT led o inc eased igh wa d F on ie s in Human Neu oscience | www. on ie sin.o g 4Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y FIGURE 2 | ANT-DBS led o inc eased igh wa d on al alpha asymme y. (A) When ANT was s imula ed, enhanced igh wa d on al asymme y (smalle asymme y sco e) was ound in compa ison o si ua ions when s imula ion was OFF and when halamic con ol loca ion (Con ol) was s imula ed. ∗S a is ically signi ican ; n.s., no signi icance. E o ba s ep esen Fishe ’s leas signi ican di e ence. (B) Indi idual changes in he on al alpha asymme y when ANT was s imula ed and when i was no . on al alpha asymme y (i.e., ela i ely less alpha powe on he igh ) compa ed o OFF (F(1,12)= 14.09, p= 0.003, η2 G= 0.085) and ON a con ol loca ion (F(1,12)= 10.19, p= 0.008, η2 G= 0.104), Figu e 2A. Unde he Bon e oni-adjus ed signi icance c i e ia, he e was an opposi e end when con ol loca ion was s imula ed wi h dec eased igh wa d on al alpha asymme y compa ed o OFF si ua ions (F(1,12)= 6.61, p= 0.025, η2 G= 0.021). Subjec speci ic e ec s o ANT-DBS on on al alpha asymme y a e illus a ed in Figu e 2B. Analysis o he hemisphe ic alpha ac i i y esul ed in an in e ac ion e ec be ween S imula ion and Hemisphe e (F(2,24)= 14.09, p= 0.0007, η2 G= 0.041). Unde he Bon e oni- adjus ed signi icance c i e ia, he e was a end owa ds inc eased igh hemisphe ic ac i i y when ANT was s imula ed (F(1,12)= 5.89, p= 0.03, η2 G= 0.14), wi h le 1.32 ±1.11 µV and igh 1.16 ±1.07 µV. This endency did no exis when con ol loca ion was s imula ed (F(1,12)= 0.01, p= 0.75, η2 G= 0.004) o when s imula ion was OFF (F(1,12)= 0.53, p= 0.48, η2 G= 0.039). F on al Alpha Asymme y, Hemisphe ic Ac i i y and Emo ional Modula ion o Beha io , Good Pe o me s To inc ease s a is ical powe and con ol o mul iple compa ison, subg oup analysis was done only o ANT ON and OFF si ua ions. Speci ically, ins ead o using a h ee-le el ac o S imula ion (ANT ON, OFF and con ol loca ion ON), we used a wo-le el ac o S imula ion (ANT ON and OFF) in he ollowing s a is ical analysis. Analysis o he on al alpha asymme y sco es o good pe o me s esul ed in compa able esul s o whole g oup analysis, i.e., a main e ec o S imula ion (F(1,7)= 9.75, p= 0.017, η2 G= 0.38). Addi ionally, in e ac ion e ec be ween S imula ion and Emo ion was ma ginally signi ican (F(1,7)= 4.94, p= 0.06, η2 G= 0.003). Analysis o he hemisphe ic alpha ac i i y esul ed in an in e ac ion e ec be ween S imula ion and Hemisphe e (F(1,7)= 9.75, p= 0.017, η2 G= 0.05). Pos hoc analysis esul ed in a main e ec o S imula ion on he igh hemisphe e (F(1,7)= 6.58, FIGURE 3 | ANT s imula ion led o signi ican ly inc eased igh hemisphe ic ac i i y, as indica ed by educed alpha powe . E o ba s ep esen Fishe ’s leas signi ican di e ence. ∗S a is ically signi ican . p= 0.037, η2 G= 0.3) bu no he le hemisphe e (F(1,7)= 0.42, p= 0.54, η2 G= 0.03; Figu e 3). Reac ion ime analysis esul ed in an in e ac ion e ec be ween Emo ion and S imula ion (F(1,7)= 20.75, p= 0.003, η2 G= 0.03). Pos hoc analysis indica ed a main e ec o Emo ion when ANT was s imula ed (F(1,7)= 16.66, p= 0.005, η2 G= 0.026), wi h slowe eac ion imes in con ex o Emo ional (512 ±144 ms) han Neu al s imuli (495 ±147 ms). The e was also a main e ec o Emo ion when s imula ion was OFF (F(1,7)= 19.41, p= 0.003, η2 G= 0.22) wi h as e eac ion imes in con ex o Emo ional (486 ±118 ms) in con as o Neu al s imuli (492 ±119 ms). Co ela ion be ween F on al Alpha Asymme y Sco es and Emo ion-Rela ed Reac ion Times To in es iga e an associa ion be ween he e ec o ANT-DBS on emo ional modula ion o on al alpha asymme y and emo ional modula ion o RTs, we did a co ela ion analysis using Pea son’s co ela ion. The modula ion o ANT-DBS on esponse speed and F on ie s in Human Neu oscience | www. on ie sin.o g 5Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y FIGURE 4 | Deduc ed modula o y e ec s o emo ion and ANT s imula ion show a posi i e co ela ion be ween alpha asymme y sco es and he eac ion imes. alpha asymme y sco es (see ‘‘Ma e ials and Me hods’’ sec ion o he measu es) linked o emo ion we e co ela ed ( = 0.69, p= 0.03; Figu e 4). DISCUSSION We ound in his s udy ha du ing a cogni i e ask wi h emo ional dis ac o s high equency elec ical s imula ion a ANT inc eased igh wa d on al alpha asymme y compa ed o s imula ion a he halamic con ol loca ion o o no s imula ion. Inc eased igh wa d on al alpha asymme y sugges s ela i ely inc eased igh on al ac i i y, which has been p e iously linked wi h anxie y diso de s and dep ession (Coan and Allen, 2004; Thibodeau e al., 2006). This inding is in line wi h ou p e ious inding o ANT-DBS inc easing a en ion alloca ion o h ea (Ha ikainen e al., 2014; Sun e al., 2015), as well as wi h subjec i e epo s o dep ession- ela ed symp oms due o ANT-DBS (Fishe e al., 2010). Fu he mo e, we unco e ed he e ec o ANT-DBS on beha io al and physiological b ain esponses ela ed o emo ional s imuli wi h o he con ibu ing ac o s sub ac ed. A e de i ing hese ne ANT-DBS impac indices we ound a co ela ion be ween emo ion-linked alpha asymme y sco es and eac ion imes. We ha e p e iously shown ha emo ional modula ion o on al alpha asymme y is also al e ed by ano he neu omodula ion me hod impac ing he a ec i e ci cui s i.e., VNS (Sun e al., 2017). To ha end, we sugges ha ask- ela ed on al alpha asymme y along wi h measu e o emo ional in e e ence o beha io could be used as a po en ial bioma ke o objec i ely asses and gea he a ec i e e ec s o neu omodula ion ea men s. Hemisphe ic asymme y in he p e on al ac i a ion, as measu ed by he EEG powe in he alpha band, has been p oposed o p edic b ain’s esponse o a ec i ely alenced s imuli (Da idson e al., 1990). Based on he Da idson’s app oach-wi hd awal heo y, he le on al b ain egion may be mo e ac i e owa d posi i e emo ional s imuli linked wi h app oach- ela ed beha io s and he igh on al egion mo e ac i e owa d nega i e emo ional s imuli linked wi h wi hd awal- ela ed beha io s. Gene ally, alpha powe dec eases wi h inc easing cogni i e ac i i y (P u schelle e al., 1996; Klimesch, 1999). The on al asymme y is mos commonly compu ed by sub ac ing he na u al log o le hemisphe e alpha powe (EEG elec ode F3) om he na u al log o igh hemisphe e alpha powe (F4) while using he cen al Cz as e e ence elec ode (Coan and Allen, 2004). This app oach gene a es a unidimensional scale indica ing he ela i e ac i i y o each b ain hemisphe e, wi h high asymme y sco e indica ing ela i ely g ea e le on al ac i a ion and lowe sco es ep esen ing ela i ely g ea e igh on al ac i a ion. This adi ional app oach o measu ing on al alpha asymme y based on asymme y sco es bea s se e al ad an ages. Fi s ly, i simpli ies he da a analysis p ocess and makes i s in e p e a ion somewha s aigh o wa d. Fu he , compa ed o sepa a e hemisphe e analysis, i enables p ope con ol o indi idual di e ences, including he subjec speci ic beha io al ac i a ion and indi idual di e ences in skull hickness leading o non-neu ogenic di e ences in absolu e powe alues (Ha mon-Jones and Allen, 1997; Su on and Da idson, 1997; Coan and Allen, 2003). Also, hese asymme y sco es a e epo ed o show highe in e nal consis ency (Toma ken e al., 1992; Jones e al., 1997). F on al alpha asymme y may be a mo e obus ma ke han ERPs in de ec ing he a ec i e e ec s o ANT-DBS in epilepsy pa ien s. ERP analysis equi es high quali y EEG wi h ew a i ac s and iden i iable ERP componen s di icul o ob ain in some pa ien s wi h e ac o y epilepsy (Ha ikainen e al., 2014; Sun e al., 2015). EEG signal om pa ien s wi h epilepsy is o en con amina ed by epilep ic discha ges and o he pa hological EEG oscilla ions esul ing in indis inguishable and widely a iable ERP componen s making he use o ERP de i ed bioma ke s less eliable. Mo eo e , ob aining eliable ERP bioma ke s equi es se e al epe i ions and long es ing pe iods incon enien in a clinical se ing. Thus, obus and quickly de i able bioma ke s allowing o immedia e, online assessmen o he impac o neu omodula ion on a ec i e ci cui s a e needed. Task-elici ed on al alpha asymme y du ing emo ional challenge has been shown o be a mo e obus indica o o ulne abili y o dep ession han es ing s a e alpha asymme y (S ewa e al., 2014). In he cu en s udy, he cogni i e ask engaging on al ne wo ks was combined wi h emo ional challenge i.e., emo ional dis ac o . The compu e based es o execu i e unc ions used in he cu en s udy mimics e e yday challenges wi h demand o a ie y o cogni i e p ocesses, like a en ion and execu i e con ol simul aneously. To ha end, he ask- ela ed on al alpha asymme y may be e e lec b ain’s a ec i e unc ions in he con ex o e e yday cogni i e challenges han es ing s a e alpha asymme y. Fu he mo e, combining EEG asymme y measu e wi h a beha io al ask including measu e o he ex en o emo ional F on ie s in Human Neu oscience | www. on ie sin.o g 6Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y in e e ence allows o elec ophysiological bioma ke o be combined wi h a beha io al bioma ke linking he obse ed e ec o neu omodula ion on bo h b ain physiology and beha io . Ou indings along wi h p e ious li e a u e sugges ha ANT-DBS may modula e a ec i e unc ions owa ds po en ial ulne abili y o dep ession, as i causes inc eased igh hemisphe ic ac i i y and an inc eased a en ion alloca ion o h ea , a hallma k o dep essi e symp oms. Howe e , cau ion is wa an ed when ex apola ing he cu en indings o clinical se ing, pa ially due o ou expe imen al se ing o s imula ion pa ame e s di e ing om hose used clinically. In ou s udy, we used con inuous s imula ion wi h bipola mode. In he clinical ea men , monopola and in e mi en s imula ion is ypically used (e.g., 1-min ON and 5 min OFF). Fu he mo e, we did no e alua e dep ession- ela ed symp oms in his s udy, hus no di ec conclusions can be d awn on whe he he obse ed igh wa d alpha asymme y and inc eased emo ional in e e ence o beha io due ANT-DBS e lec ulne abili y o dep ession. Taken oge he , b idging he immedia e e ec s ound in he cu en s udy o long- e m ou comes as well as di ec ly linking hese immedia e b ain and beha io al esponses o ANT-DBS o subjec i e mood ela ed symp oms equi e u he s udies. The cu en indings highligh he ole o ANT in human emo ional p ocessing. Al hough ANT has long been hough c i ical o he p ocessing o emo ion, and known as a key node o he Papez ci cui (Papez, 1937), di ec e idence suppo ing ANT’s ole in human a ec i e b ain unc ions is sca ce (Ha ikainen e al., 2014; Sun e al., 2015). High- equency elec ic s imula ion a ANT, hough o mimic a e e sible lesion, p o ides a unique oppo uni y o s udy i s ole in emo ion and a en ion. Fu he mo e, u ning his key node o he limbic ci cui y on and o by means o DBS when subjec s a e ac i ely engaged in a cogni i e ask and in e mi en ly dis ac ed by biologically ele an h ea ela ed emo ional s imuli, while EEG and pe o mance a e measu ed, p o ides in aluable insigh in o a ec i e b ain ci cui s and how hei unc ion is e lec ed in beha io al and b ain esponses in humans. An app oach allowing online manipula ion o a ec i e ci cui s and measu emen o causally linked al e a ions in a ec i e esponses is also in aluable in iden i ying and assessing po en ial bioma ke s o he e ec o neu omodula ion on b ain’s a ec i e ci cui y in gene al. Lack o objec i e signs o neu omodula ion on a ec i e ci cui s calls o on-line bioma ke s o moni o ing and op imizing neu omodula ion ea men . Unlike when ea ing mo emen diso de s wi h DBS whe e one can objec i ely obse e an immedia e educ ion o mo emen ela ed symp oms, when ea ing psychia ic diso de s, such as dep ession and obsessi e- compulsi e diso de wi h DBS, he e a e no ob ious signs e lec ing he desi ed e ec s on limbic ci cui s which could be used o guide pa ame e selec ion owa ds he op imal ea men e ec . In he cu en s udy, we demons a ed he modula o y e ec o ANT-DBS on b ain’s a ec i e ci cui y and highligh ed a po en ial new online bioma ke o he immedia e e ec s o neu omodula ion on b ain dynamics, i.e., combining on al alpha asymme y and emo ional in e e ence o esponse speed. Howe e , we did no link hese measu es wi h mood ela ed symp oms. Thus, u u e s udies a e needed o assess he u ili y o on al alpha asymme y du ing a ask engaging on al ne wo ks as a bioma ke , which possibly could be used o minimizing he a ec i e side e ec s o ANT-DBS in he ea men o e ac o y epilepsy and o guiding neu omodula ion ea men s o psychia ic diso de s linked wi h on al dys unc ion and asymme y, such as dep ession and anxie y, owa ds he op imal ea men e ec . In conclusion, we showed immedia e e ec s on on al alpha asymme y due o neu omodula ion a ge ed a a key node o he limbic ci cui . Fu he , we p o ided causal e idence ha neu omodula ion which impac s a ec i e ci cui s, ANT- DBS, in luences alpha asymme y and he modula ion o alpha asymme y is linked wi h emo ional modula ion o beha io . Thus, we conclude ha an index de i ed om al e a ion o on al alpha asymme y along wi h emo ional in e e ence o pe o mance e lec he unc ioning o limbic ci cui s and he impac o neu omodula ion on hem. While clinical ele ance o such an index equi es u he esea ch, i p o ides an objec i e measu e o wide ange o scien i ic expe imen s ha aim ei he a be e unde s anding dynamics o a ec i e ci cui s, hei dys unc ion o he impac o neu omodula ion on hem. Fu he mo e, s imula ing he wo closely loca ed halamic egions lead o dis inc la e aliza ion e ec s sugges ing on al alpha asymme y migh be used o p o ide addi ional in o ma ion on he DBS elec ode loca ion. AUTHOR CONTRIBUTIONS LS con ibu ed o da a analysis and w i ing he a icle. JP was in ol ed in s a is ical analysis and w i ing he a icle. KMH con ibu ed o expe imen al design, supe ision o da a analysis and w i ing he a icle. ACKNOWLEDGMENTS This s udy was suppo ed by he Academy o Finland and Pi kanmaa Hospi al Dis ic Compe i i e Resea ch Fund. REFERENCES Blackha , G. C., Minnix, J. A., and Kline, J. P. (2006). Can EEG asymme y pa e ns p edic u u e de elopmen o anxie y and dep ession? A p elimina y s udy. Biol. Psychol. 72, 46–50. doi: 10.1016/j.biopsycho.2005.06.010 B oadway, J. M., Hol zheime , P. E., Hilimi e, M. R., Pa ks, N. A., DeVylde , J. E., Maybe g, H. S., e al. (2012). F on al he a co dance p edic s 6-mon h an idep essan esponse o subcallosal cingula e deep b ain s imula ion o ea men - esis an dep ession: a pilo s udy. Neu opsychopha macology 37, 1764–1772. doi: 10.1038/npp.2012.23 F on ie s in Human Neu oscience | www. on ie sin.o g 7Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y Coan, J. A., and Allen, J. J. B. (2003). F on al EEG asymme y and he beha io al ac i a ion and inhibi ion sys ems. Psychophysiology 40, 106–114. doi: 10.1111/1469-8986.00011 Coan, J. A., and Allen, J. J. B. (2004). F on al EEG asymme y as a mode a o and media o o emo ion. Biol. Psychol. 67, 7–49. doi: 10.1016/j.biopsycho.2004. 03.002 Coan, J. A., Allen, J. J. B., and McKnigh , P. E. (2006). A capabili y model o indi idual di e ences in on al EEG asymme y. Biol. Psychol. 72, 198–207. doi: 10.1016/j.biopsycho.2005.10.003 Da idson, R. J., Ekman, P., Sa on, C. D., Senulis, J. A., and F iesen, W. V. (1990). App oach-wi hd awal and ce eb al asymme y: emo ional exp ession and b ain physiology: I. J. Pe s. Soc. Psychol. 58, 330–341. doi: 10.1037/0022- 3514.58.2.330 Field, T., and Diego, M. (2008). Ma e nal dep ession e ec s on in an on al EEG asymme y. In . J. Neu osci. 118, 1081–1108. doi: 10.1080/00207450701 769067 Fishe , R., Salano a, V., Wi , T., Wo h, R., Hen y, T., G oss, R., e al. (2010). Elec ical s imula ion o he an e io nucleus o halamus o ea men o e ac o y epilepsy. Epilepsia 51, 899–908. doi: 10.1111/j.1528- 1167.2010.02536.x Go lib, I. H. (1998). EEG Alpha asymme y, dep ession, and cogni i e unc ioning. Cogn. Emo . 12, 449–478. doi: 10.1080/026999398379673 G imshaw, G. M., Fos e , J. J., and Co ballis, P. M. (2014). F on al and pa ie al EEG asymme ies in e ac o p edic a en ional bias o h ea . B ain Cogn. 90, 76–86. doi: 10.1016/j.bandc.2014.06.008 Ha mon-Jones, E., and Allen, J. J. B. (1997). Beha io al ac i a ion sensi i i y and es ing on al EEG asymme y: co a ia ion o pu a i e indica o s ela ed o isk o mood diso de s. J. Abno m. Psychol. 106, 159–163. doi: 10.1037/0021-843x. 106.1.159 Ha ikainen, K. M., and Knigh , R. (2003). ‘‘La e al and o bi al p e on al co ex con ibu ions o a en ion,’’ in De ec ion o Change: E en -Rela ed Po en ial and MRI Findings, ed. J. Polich (Ne he lands: Kluwe Academic P ess), 99–116. Ha ikainen, K. M., Ogawa, K. H., and Knigh , R. T. (2000). T ansien in e e ence o igh hemisphe ic unc ion due o au oma ic emo ional p ocessing. Neu opsychologia 38, 1576–1580. doi: 10.1016/s0028-3932(00) 00072-5 Ha ikainen, K. M., Ogawa, K. H., Sol ani, M., and Knigh , R. T. (2007). Emo ionally a ousing s imuli compe e o a en ion wi h le hemispace. Neu o epo 18, 1929–1933. doi: 10.1097/wn .0b013e3282 1ca18 Ha ikainen, K. M., Sun, L., Pol i aa a, M., B ause, M., Leh imaki, K., Haapasalo, J., e al. (2014). Immedia e e ec s o deep b ain s imula ion o an e io halamic nuclei on execu i e unc ions and emo ion- a en ion in e ac ion in humans. J. Clin. Exp. Neu opsychol. 36, 540–550. doi: 10.1080/13803395.2014.913554 Ha ikainen, K. M., Waljas, M., Iso ii a, T., Das ida , P., Liima ainen, S., Solbakk, A. K., e al. (2010). Pe sis en symp oms in mild o mode a e auma ic b ain inju y associa ed wi h execu i e dys unc ion. J. Clin. Exp. Neu opsychol. 32, 767–774. doi: 10.1080/13803390903521000 Hen iques, J. B., and Da idson, R. J. (1990). Regional b ain elec ical asymme ies disc imina e be ween p e iously dep essed and heal hy con ol subjec s. J. Abno m. Psychol. 99, 22–31. doi: 10.1037/0021-843x.99.1.22 Hen iques, J. B., and Da idson, R. J. (1991). Le on al hypoac i a ion in dep ession. J. Abno m. Psychol. 100, 535–545. doi: 10.1037/0021-843x.100. 4.535 Jones, N. A., Field, T., Da alos, M., and Pickens, J. (1997). EEG s abili y in in an s/child en o dep essed mo he s. Child Psychia y Hum. De . 28, 59–70. doi: 10.1023/A:1025197101496 Klimesch, W. (1999). EEG alpha and he a oscilla ions e lec cogni i e and memo y pe o mance: a e iew and analysis. B ain Res. Re . 29, 169–195. doi: 10.1016/s0165-0173(98)00056-3 Law ence, M. A. (2013). ez: Easy analysis and isualiza ion o ac o ial expe imen s. R Package Ve sion 4.2–2. A ailable online a : h p://c an. - p ojec .o g/package=ez Liima ainen, J., Pe äkylä, J., Jä elä, K., Sis o, T., Yli-Hankala, A., and Ha ikainen, K. M. (2016). Imp o ed cogni i e lexibili y a e ao ic al e eplacemen . In e ac . Ca dio asc. Tho ac. Su g. 23, 630–636. doi: 10.1093/ic s/i w170 Mäki-Ma unen, V., Kuusinen, V., B ause, M., Pe äkylä, J., Pol i aa a, M., dos San os Ribei o, R., e al. (2015). Enhanced a en ion cap u e by emo ional s imuli in mild auma ic b ain inju y. J. Neu o auma 32, 272–279. doi: 10.1089/neu.2014.3557 Mäki-Ma unen, V., Kuusinen, V., Pe äkylä, J., Ogawa, K. H., B ause, M., B ande , A., e al. (2017). G ea e a en ion o ask- ele an h ea due o o bi o on al lesion. J. Neu o auma 34, 400–413. doi: 10.1089/neu.20 15.4390 Ma he sul, D., Williams, L. M., Hopkinson, P. J., and Kemp, A. H. (2008). In es iga ing models o a ec : ela ionships among EEG alpha asymme y, dep ession, and anxie y. Emo ion 8, 560–572. doi: 10.1037/a00 12811 Papez, J. W. (1937). A p oposed mechanism o emo ion. 1937. J. Neu opsychia y Clin. Neu osci. 7, 103–112. doi: 10.1176/jnp.7.1.103 Pe äkylä, J., Sun, L., Leh imäki, K., Pel ola, J., Öhman, J., Mö önen, T., e al. (2017). Causal e idence om humans o he ole o mediodo sal nucleus o he halamus in wo king memo y. J. Cogn. Neu osci. 29, 2090–2102. doi: 10.1162/jocn_a_01176 P u schelle , G., S ancák, A., and Neupe , C. (1996). E en - ela ed synch oniza ion (ERS) in he alpha band—an elec ophysiological co ela e o co ical idling: a e iew. In . J. Psychophysiol. 24, 39–46. doi: 10.1016/s0167- 8760(96)00066-9 Qu aan, M. A., P o zne , A. B., Daskalakis, Z. J., Giacobbe, P., Tang, C. W., Kennedy, S. H., e al. (2014). EEG powe asymme y and unc ional connec i i y as a ma ke o ea men e ec i eness in DBS su ge y o dep ession. Neu opsychopha macology 39, 1270–1281. doi: 10.1038/npp. 2013.330 Salano a, V., Wi , T., Wo h, R., Hen y, T. R., G oss, R. E., Nazza o, J. M., e al. (2015). Long- e m e icacy and sa e y o halamic s imula ion o d ug- esis an pa ial epilepsy. Neu ology 84, 1017–1025. doi: 10.1212/WNL. 0000000000001334 S ewa , J. L., Coan, J. A., Towe s, D. N., and Allen, J. J. B. (2011). F on al EEG asymme y du ing emo ional challenge di e en ia es indi iduals wi h and wi hou li e ime majo dep essi e diso de . J. A ec . Diso d. 129, 167–174. doi: 10.1016/j.jad.2010.08.029 S ewa , J. L., Coan, J. A., Towe s, D. N., and Allen, J. J. (2014). Res ing and ask-elici ed p e on al EEG alpha asymme y in dep ession: suppo o he capabili y model. Psychophysiology 51, 446–455. doi: 10.1111/psyp. 12191 Sun, L. (2016). Impac o Neu omodula ion on Cogni i e and A ec i e B ain Func ions in Humans. Tampe e: Tampe e Uni e si y P ess. Sun, L., Pe äkylä, J., Holm, K., Haapasalo, J., Leh imäki, K., Ogawa, K. H., e al. (2017). Vagus ne e s imula ion imp o es wo king memo y pe o mance. J. Clin. Exp. Neu opsychol. 39, 954–964. doi: 10.1080/13803395.2017.12 85869 Sun, L., Pe äkylä, J., Ko alainen, A., Ogawa, K. H., Ka hunen, P. J., and Ha ikainen, K. M. (2016). Human b ain eac s o ansc anial ex aocula ligh . PLoS One 11:e0149525. doi: 10.1371/jou nal.pone.01 49525 Sun, L., Pe äkylä, J., Pol i aa a, M., Öhman, J., Pel ola, J., Leh imäki, K., e al. (2015). Human an e io halamic nuclei a e in ol ed in emo ion-a en ion in e ac ion. Neu opsychologia 78, 88–94. doi: 10.1016/j.neu opsychologia.2015. 10.001 Su on, S. K., and Da idson, R. J. (1997). P e on al b ain asymme y: a biological subs a e o he beha io al app oach and inhibi ion sys ems. Psychol. Sci. 8, 204–210. doi: 10.1111/j.1467-9280.1997. b00413.x Thibodeau, R., Jo gensen, R. S., and Kim, S. (2006). Dep ession, anxie y, and es ing on al EEG asymme y: a me a-analy ic e iew. J. Abno m. Psychol. 115, 715–729. doi: 10.1037/0021-843x.115.4.715 Toma ken, A. J., and Da idson, R. J. (1994). F on al b ain ac i a ion in ep esso s and non ep esso s. J. Abno m. Psychol. 103, 339–349. doi: 10.1037/0021-843x. 103.2.339 Toma ken, A. J., Da idson, R. J., Wheele , R. E., and Kinney, L. (1992). Psychome ic p ope ies o es ing an e io eeg asymme y: empo al s abili y and in e nal consis ency. Psychophysiology 29, 576–592. doi: 10.1111/j.1469- 8986.1992. b02034.x Valen ín, A., Ga cía Na a e e, E., Chel a ajah, R., To es, C., Na as, M., Vico, L., e al. (2013). Deep b ain s imula ion o he cen omedian halamic nucleus o F on ie s in Human Neu oscience | www. on ie sin.o g 8Decembe 2017 | Volume 11 | A icle 584 Sun e al. Neu omodula ion Induced F on al Alpha Asymme y he ea men o gene alized and on al epilepsies. Epilepsia 54, 1823–1833. doi: 10.1111/epi.12352 an de Vinne, N., Volleb eg , M. A., an Pu en, M. J. A. M., and A ns, M. (2017). F on al alpha asymme y as a diagnos ic ma ke in dep ession: ac o ic ion? A me a-analysis. Neu oimage Clin. 16, 79–87. doi: 10.1016/j.nicl.2017. 07.006 Wen, D. J., Soe, N. N., Sim, L. W., Sanmugam, S., Kwek, K., Chong, Y.-S., e al. (2017). In an on al EEG asymme y in ela ion wi h pos na al ma e nal dep ession and pa en ing beha io . T ansl. Psychia y 7:e1057. doi: 10.1038/ p. 2017.28 Con lic o In e es S a emen : The au ho s decla e ha he esea ch was conduc ed in he absence o any comme cial o inancial ela ionships ha could be cons ued as a po en ial con lic o in e es . Copy igh © 2017 Sun, Pe äkylä and Ha ikainen. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (CC BY). The use, dis ibu ion o ep oduc ion in o he o ums is pe mi ed, p o ided he o iginal au ho (s) o licenso a e c edi ed and ha he o iginal publica ion in his jou nal is ci ed, in acco dance wi h accep ed academic p ac ice. No use, dis ibu ion o ep oduc ion is pe mi ed which does no comply wi h hese e ms. F on ie s in Human Neu oscience | www. on ie sin.o g 9Decembe 2017 | Volume 11 | A icle 584