Press Release December 2009, Volume 33 Conens of his issue IN THE FOCUS New rtms simulaion device PowerMAG research 2 Produc Developmens Looking back a BrainVision Analyzer and RecView in 2009 3 Produc Developmens Brain-Compuer Inerface echnology: a challenging research field 4 exreme EEG EEG in a Rollercoaser - exreme EEG measured in he Europa- Park, Rus 6 Produc Developmens The acicap sysem has come of age 8 News in brief Downloads, Programs and Updaes 8 Brain Producs Disribuors News in Norh America Widespread changes for Brain Vision LLC 9 Suppor Tip BooCamp - BrainVision Analyzer on Apple Mac Sysems 9 Dear Cusomers, Dear Friends of Brain Producs, We wish you a happy and relaxing Chrismas holiday and a rewarding and successful 2010. May we also hank you for he rus you have placed in us and he successful business relaionship we have enjoyed. This year, we are again pleased o be able o share he fruis of our success: We shall be making a donaion of 4,500 which will allow he indoor paddling pool in he Salberghaus children s home in Puzbrunn near Munich o be renovaed nex year. The Salberghaus provides children from difficul backgrounds wih proecion, securiy, sabiliy and herapeuic assisance, hus opening up new perspecives in life for hem. The primary objecive is always o reurn he children o heir own families, which means ha he Salberghaus also works inensively wih he parens in order o foser he relaionship beween he parens and he child o esablish a new basis for rus. Our donaion from las year was used o finance an advenure playground and as you can see, i s really fun for he children! We also wish o offer ongoing suppor o he Salberghaus: We inend using par of he revenue from each full Analyzer 2 license sold over he coming year o allow wo children from he Salberghaus o spend a day on a farm. Our goal is o make i possible for all he children o spend a week on he farm. For raumaized children from difficul homes in paricular, conac wih animals helps hem o open up. Life in proximiy wih naure and responsibiliy for he animals on he farm provides hem wih a daily rouine of needs and purpose. We shall keep you up o dae wih he progress of our projec in our newsleers over he course of he year! For more informaion please visi /chariy.php Your Brain Producs Team Brain Producs Disribuors Brain Producs 6h Disribuor Meeing 11 Brain Producs Workshops Brain Producs EEG & fmri Workshop 12 News in brief Workshops, Conferences & Exhibiions 12 Brain Producs Inside Who is who - new Brain Producs members of saff 13 Conac & Imprin 13 Dr. Achim Hornecker Brain Producs CEO Alexander Svojanovsky Brain Producs CEO Pierlugi Casellone (Inernaional Sales Manager) Anja Egger (Markeing Manager) Gregor Enzian (Trainee Sofware Engineer) Dr. Lisa Glass (Scienific Suppor Coordinaor) Jens Gruner (CEO MES GmbH) Dr. Ingmar Guberle (Scienific Suppor Manager) Norber Hauser (Sofware Developer) Monika Heiss (Office Manager & Accouning) Dr. Michael Jachan (Sofware Developer) David Kadlec (Sales & Suppor MES GmbH) Markus Koelble (Sofware Developer) Helmu Luz (Logisics Manager) Dinh-Phuong Nguyen (Recorder Sofware Develop. Manager) Bernd Pich (Head of he R&D Deparmen) Dr. Mihai Prunescu (Sofware Developer) Dr. Davide Riccobon (Produc & Risk Manager) Sefanie Rudrich (Evens & Public Relaions) Maria Scha (Sofware Developer) Benjamin Spinnler (Logisics) Dirk Sellmacher (Sofware Developer) Dr. Rober Sörmer (Technical Suppor Manager) Chrisiane Srasser (Office Managemen) Johann Srasser (Producion Manager R&D) Erich Svojanovsky (CEO MES GmbH) Anne Vogel (Regulaory Affairs Manager) Ingo Wesermayer (IT Trainee) Kaja Wus (Technical Wrier) Illya Yaroshenko (IT Manager) page 1 of 13
IN THE FOCUS New rtms simulaion device PowerMAG research Advanced soluions for online EEG and rtms applicaions: A cooperaion beween MAG & More and Brain Producs by Alex Svojanovsky and Dr. Kersin Wendicke, MAG & More GmbH The abiliy o carry ou TMS and EEG a he same ime (online regisraion) offers a unique opporuniy o invesigae coricocorical neworks. This combinaion is of massive significance in he scienific communiy across he globe (for regisering funcional reorganizaion paerns following a sroke or deecing changes o inracerebral conneciviy wih degeneraive diseases). We have herefore se ourselves he ask of providing he bes possible echnical foundaion for hese kinds of scienific experimens. Using a combinaion of he BrainAmp DC or BrainAmp MR plus, he EasyCap TMS and he appropriae careful planning of he rial, he familiar TMS simulus arifac can be reduced o 5 ms [1]. Furher opimizaion (such as power-line noise and recharging arifacs) is, however, only possible in direc, close cooperaion wih a manufacurer of TMS equipmen. The desire o also be able o correc he simulus-induced arifacs in he EEG recording led o a fruiful collaboraion beween MAG & More and Brain Producs. One aspec of his unique full soluion was he echnical challenge of phase-synchronizing he TMS simulaor and he EEG sysem, along he same lines as he esablished clock synchronizaion beween he BrainAmp MR sysem and various MRI scanners. Along wih simulus iming ha is precisely mached o he sample rae and provision of a consan inerval beween he simuli by he simulaion hardware, a furher prerequisie is ha he TMS coil is securely aached and he es subjec resrained. For demonsraion purposes, we injeced TMS pulses ino a known signal (sine wave, 5 Hz, inernal BrainAmp es signal) in order o illusrae he effeciveness of he TMS simulus arifac correcion ha was now possible. The red curve shows he sinusoidal signal wih TMS arifacs and he black curve plos he adjused signals. Ch1 Ch1 200 µv R 128 (TMS Trigger) R 128 (TMS Trigger) Sync On I is already possible o correc arifacs using MR gradien correcion and ICA. However, we are of he opinion ha we have esablished a unique echnical foundaion ha will provide he impeus for he developmen of many new correcion echniques and scienific mehods. The PowerMAG research will be available in 30 Hz and 100 Hz varians. Designed for Brain Producs, his rtms range from MAG & More will be available exclusively from Brain Producs and heir disribuors as of January 2010. The PowerMAG research range offers special specificaions for online TMS and EEG applicaions: 200 µv R 128 (TMS Trigger) R 128 (TMS Trigger) Sync On Exremely reduced power-line noise on conac wih he coil No recharging arifacs (e.g. wih rtms) Trigger oupu opimized for EEG Analog inpu for exernal proocol conrol (iming, inensiy) Clock synchronizaion oupu for arifac correcion Hardware required: PowerMAG research (30 Hz or 100 Hz rtms), BrainAmp DC/MR plus, SyncBox, EasyCap TMS Reference 1. Clin Neurophysiol. 2006 Aug; 117(8):1699-707. Transcranial magneic simulaion and corical evoked poenials: A TMS/EEG co-regisraion sudy. Bonao C, Miniussi C, Rossini PM. Neurofisiologia IRCCS Cenro S. Giovanni di Dio Faebenefraelli, Via Pilasroni 4, 25125 Brescia, Ialy) Paern Recogniion, Wiley- Inerscience; 1 Ediion (March 27, 1992), Chaper 12 page 2 of 13
Produc Developmens Looking back a BrainVision Analyzer and RecView in 2009 by Dr. Achim Hornecker Following he release of Analyzer 2 in he Summer of 2008, BrainVision RecView was also updaed in April 2009 o mee he demands of modern sofware echnology. This aricle will briefly summarize he improvemens o hese wo producs ha have been released during he course of his year or ha have already been developed and are due for release soon. professional 2 professional RECVIEW We shall firs ake a look a Analyzer 2. Version 2.0 incorporaed many new feaures, and following he main release, i was firs necessary o correc some bugs which had no shown up during inensive bea esing. This resuled in he updaed Version 2.0.1 in Spring 2009. Bu even his updae was far more han merely a service pack of he kind we are familiar wih for Windows. Analyzer 2.0.1 also incorporaed a number of addiional feaures. To sar wih, he ICA Inverse module provides a powerful ool for manipulaing daa calculaed using ICA. Afer componens have been analyzed using ICA, he ineracive user inerface of ICA Inverse allows users o ry differen combinaions of back-ransformaion of he ICA ino he EEG. Arifac correcion and he exracion of relevan componens for source localizaion are jus wo of he poenial applicaions. The expor and impor of markers has also been improved, replacing he familiar soluion from Analyzer 1. Markers can now also be expored in XML forma, a daa forma ha has become very popular over he pas decade and which is suppored by many hird pary producs. This forma also makes i very easy o exrac he user properies of markers. The marker impor funcion is also able o read hese files o merge markers in new nodes. However, his module also incorporaes a new feaure ha makes i possible o impor daa from oher hisory nodes. Why is his useful? Imagine a siuaion where you have implemened some filers in he frequency domain and ransform he daa back o he ime domain using he FFT Inverse funcion of Analyzer 2. In his even, all your ime domain markers would have been los because hey made no sense in he frequency domain. The node feaure of he marker impor funcion allows you o capure he markers from he paren node before you perform he FFT. Bu new feaures were no he only changes in Analyzer 2.0.1. One fuure rend in compuer echnology will be he increasing number of processor cores incorporaed in a single PC. In Analyzer 2.0.1 we inroduced he kernel funcionaliy ha makes i possible o exploi such mulikernel archiecures o improve he performance of our ransforms. The filers module will be redesigned and released as he firs example of his group. For BrainVision RecView 1.3, we moved over o he new version of he Microsof.NET Framework, Version 3.5. This means ha anyone waning o wrie heir own filers for RecView is now able o use he Microsof Visual Sudio 2008 Express ediion, which can be downloaded free of charge from Microsof s websie. Oher improvemens in RecView 1.3 included he new segmenaion and averaging feaures which open up new possibiliies for online daa evaluaion, especially in combinaion wih online MR arifac correcion. The FFT was improved o allow i o also work wih segmens and overlapping ranges. This is of ineres if you are using RecView for biofeedback applicaions. A new online LORETA module makes i possible o apply biofeedback modules no only o channels or combinaions of channels bu also o he aciviy of defined regions in he brain. The virual channels produced by his module can be used in he same way as he channels from acquisiion. Many cusomers have also asked us o provide a special funcion, namely promoion of riggers on he parallel por depending on paricular evens exraced from he EEG. We will mee his demand by releasing a se of rigger modules, for insance a level rigger and a band power rigger and even an SDK for developing rigger applicaions no covered by he general-purpose funcionaliy. Anoher ineresing module for RecView is currenly in bea esing and is due for release soon. The bridge ool will deec gel bridges during acquisiion using a combinaion of correlaion mehods. This will provide new opions for conrolling and correcing he validiy of he daa during recording. So we see ha a large number of improvemens have already been released for hese wo producs and work is in hand on furher enhancemens. And we have no even menioned all he componens ha have been developed for research and for using BCI wih our equipmen and sofware. These aspecs are covered by anoher aricle in his issue. page 3 of 13
Produc Developmens Brain-Compuer Inerface echnology: a challenging research field by Maria Scha A Brain-Compuer Inerface is a sysem for conrolling a compuer or oher device by human voliion wihou using muscle or peripheral nerve aciviy. A significan amoun of BCI research relaes o he ask of developing BCI applicaions based on non-invasive brain signal recordings such as EEG. Unlike invasive or parially-invasive mehods which require elecrodes o be implaned inside he brain or he corical surface, non-invasive EEG recordings pose no risk for he person being examined. A lo of effor has herefore been devoed o his field despie he problems resuling from he poor signal-o-noise raio found in scalp EEG recordings. Oher challenges have o be faced in addiion o he poor signal qualiy. Even hough some characerisic brain aciviy paerns are well known o be correlaed wih cerain asks (for example, modulaions of he µ-rhyhm in response o real or imagined movemens), he way hese paerns occur in differen individuals can differ considerably. In addiion o his phenomenon of iner-subjec-variabiliy, BCI developers and users also have o cope wih iner-sessionvariabiliy which is caused by he non-saionariy of brain paerns over ime. Markers via LPT Por Feedback Info (TCP/IP) One way of dealing wih hese problems is o provide inensive raining o subjecs: A subjec who is o use a BCI sysem ries o learn o produce a cerain paern of brain aciviy reliably and hus o adap o a given BCI sysem. Anoher and more promising sraegy is o auomaically adap he BCI sysem o he individual who is using i in each new session by applying machine learning echniques. To make his adapaion possible, a BCI experimen usually consiss of hree phases: Firs a supervised raining session is conduced and recorded. The recorded daa is hen analyzed offline. This involves he recogniion of he relevan subjec- and ask-specific feaures and he raining of a classifier on he exraced feaures. Feaure exracion ofen consiss in finding opimal spaial and emporal filers, for example by means of blind source separaion echniques such as Independen Componen Analysis or Common Spaial Paerns algorihms. Classifiers can be derived using mehods such as linear discriminan analysis, Suppor Vecor Machines or Closed feedback loop professional RECVIEW neural neworks, o name bu a few. During he feedback phase, he relevan feaures are exraced and he classifier oupu assigns a conrol signal o he exraced feaures. The above-menioned problems are only hose resuling from neurophysiology and signal processing. However, anyone developing a BCI sysem from scrach faces a large number of echnical challenges in erms of informaics and real-ime processing. The ime and effor required o perform such a ask ofen exceed he resources available in a BCI research projec dedicaed o a precisely defined research field. Markers & Daa via USB professional RECORDER Markers and Daa (TCP/IP) Recen years have seen he emergence of a number of open-source BCI frameworks which can relieve BCIresearchers of he majoriy of his developmen effor and allow hem o concenrae on finding answers o heir individual research quesions. Unforunaely, no all frameworks inerface wih all ypes of recording hardware. We are currenly rying o improve he inegraion of Brain Producs hardware in order o enrich he possibiliies available o cusomers ineresed in he BCI field. One widely used and well-documened BCI framework is BCI2000 (www.bci2000.org), a generalpurpose sysem for brain-compuer inerface (BCI) research. I can also be used for daa acquisiion, simulus presenaion, and brain monioring applicaions. Is user conribuion area feaures an RDA clien which inerfaces wih he BrainVision Recorder as well as a source module ha makes i possible o access he Brain Producs V-Amp direcly. Apar from he possibiliy of recording in he Analyzer-compaible formas EDF and GDF, he inegraed BCI2000 expor ool allows users o expor daa and saes (which correspond o BrainVision markers) in BrainVision forma. As a resul, Analyzer 2 users can apply ICA ransformaions during offline analyses. To make i possible o benefi from hese capabiliies in feedback scenarios, we provide a corresponding macro which is soon o be followed by an expor componen which expors ICA marices in a BCI2000 and MATLAB compaible forma. Expored ICA marices can hus be used as spaial filers in BCI2000. page 4 of 13
Anoher plaform for developing and esing BCI and real-ime neuroscience applicaions is he C++ based OpenViBE (hp:// openvibe.inria.fr). This plaform is remarkable for is high level of modulariy and mees he needs of programmers and nonprogrammers alike hanks o he provision of a user-friendly graphical language which allows non-programmers o design a BCI wihou wriing a single line of code. OpenViBE provides parial suppor for access o he BrainAmp via RDA clien using he Brain- Vision Recorder. This means ha he implemenaion is no robus and here is currenly no associaed documenaion. Forunaely, work is currenly underway on he direc inegraion of he Brain Producs V-Amp and his is expeced o be complee by he end of 2009. The relaively new BCI++ (www.sensibilab.campuspoin.polimi.i) is dedicaed o he developmen and fas prooyping of BCI sysems, PC-driven proocols for a variey of bio-signal acquisiion paradigms and BCI-based applicaions. Apar from a module for signal acquisiion, sorage and display, real-ime execuion and managemen of cusom algorihms, i feaures a Graphic User Inerface module based on a high-level 2D/3D Graphic Engine (Irrlich). Work on direc V-Amp inegraion is currenly in progress. Alhough an RDA clien exiss, we were no able o operae i successfully in our ess. The MATLAB-based BCI2VR (www.engineering.vcu.edu/eegbci/ bci2vr_index.hml) is a sofware sysem for neural signal processing and classificaion. Unforunaely, Brain Producs hardware is currenly no suppored. Cusomers who wan o use his framework will need o procure he BrainAmp/V-Amp SDK and inerface wih he MATLAB Daa acquisiion oolbox. Naurally, we have also considered he quesion of he exen o which we can use he Brain Producs sofware componens o build a BCI applicaion. The simples way o answer his quesion was o ry. We decided o implemen a P300 speller applicaion, one of he firs BCI spelling devices developed by Farewell and Donchin [1]. The P300 Speller is based on he eliciaion and recogniion of an even-relaed poenial referred o as P300. This posiive deflecion in he EEG is characerisic of a subjec s reacion o an unusual arge simulus. One paradigm ha is widely used o provoke he P300 response in experimenal conexs is he socalled oddball paradigm: Wihin a series of simuli, an unexpeced, devian simulus is presened in a random locaion and wih a small probabiliy of occurrence. The P300 speller paradigm is a varian of he oddball paradigm which is used in a BCI applicaion for he spelling of characers. The paradigm makes use of a selecion of characers (leers, numbers or icons) arranged in a marix. The characer marix usually consiss of six rows and six columns. The subjec focuses on he desired characer while he rows and columns flash in a random sequence. The arge simulus occurs when he characer which he subjec is focusing on is highlighed. The raio of he arge o he neural simuli is 1:5. By deecing he row-and column flashes ha, on average, elici he sronges P300 responses, he sysem ries o recognize wha characer he user has been focusing on. While using BrainVision Recorder as recording sofware, BrainVision Analyzer for offline analysis and RecView for online processing, i was sill necessary o implemen he speller paradigm iself. The poin of ineres here lies in synchronizing he paradigm aciviy wih he recording, i.e. marking he poin a which he differen simuli occur in he recorded daa. To do his, we used he amplifier rigger por. The speller paradigm sends a TTL signal which encodes he flash even o he parallel por and his hen appears as simulus marker in he recorded daa. We addiionally designed an Analyzer expor componen which implemened a sepwise feaure selecion algorihm based on linear discriminan analysis [2]. This expor componen idenifies relevan feaures and classifier weighs and expors he informaion o an xml file. This xml file is subsequenly read ino a RecView P300 feedback filer. This filer performs he online evaluaion of he recorded daa and reurns he feedback informaion o he speller paradigm via TCP/IP. In his way, a full feedback loop is esablished. Alhough we have no ye scienifically sudied he informaion ransfer rae in an experimenal seing, we have been encouraged by he spelling resuls achieved by various BP colleagues. This prooype illusraes he possibiliies ha Brain Vision RecView offers as a feedback ool and provides us wih an incenive o coninue exending he capabiliies of our sofware beyond he habiual areas of applicaion. References [1] Farwell LA, Donchin E. Talking off he op of your head: oward a menal proshesis uilizing even-relaed brain-poenials. Elecroenceph Clin Neurophysiol 1988; 70:510 23. [2] Geoffrey J. McLachlan: Discriminan Analysis and Saisical Paern Recogniion, Wiley-Inerscience; 1 Ediion (March 27, 1992), Chaper 12 page 5 of 13
exreme EEG EEG in a Rollercoaser - exreme EEG measured in he Europa-Park, Rus by Dr. Michael Jachan, Maria Scha, Parick Briz and Dr. Achim Hornecker In earlier exreme EEG sudies, we showed ha i is possible o ake EEG measuremens from humans while in moion, for example when playing able ennis. Hungarian researchers have even aken our exreme EEG Pack o he hosile environmen of he Himalayas, where hey recorded EEGs while mounaineering. We in he Freiburg Sofware Developmen Team had he idea of aking EEG measuremens under condiions of unnaurally high acceleraion, such as occur during a rollercoaser ride. This shor pilo sudy could pave he way oward a new branch of brain research ino FEAR, ANXIETY and DANGER! The Europa-Park in Rus, Baden-Würemberg (www.europapark.de), is an amusemen park for he whole family. Older children and eenagers enjoy he exciemen of riding high-power, high-ech rollercoasers, such as he Silver Sar (www.clipfish.de/video/800851/ silversar-europa-park-rus) or he Blue Fire (www.youube.com/ wach?v=ak72rtl3bm). We made a wo-day appoinmen wih he very welcoming eam a he Europa-Park who allowed us exclusive use of he Silver Sar on Sepember 14 and he Blue Fire he day afer. Of course, o benefi from an exclusive session on a rollercoaser you have o ge up early in he morning, and we were herefore probably he firs people ever o follow up a quick breakfas wih an 8 am rollercoaser ride. Silver Sar Ride We spen he firs session raining ourselves on he Silver Sar. I can sill remember ha my whole body was shaking even during our preparaions. We mouned he exreme EEG on Maria who had never ridden he Silver Sar. I, Michael, had already experienced his ride wo years previously. We prepared he acicap and conneced he hree devices: The lapop, he FirsAmp, and he aci- CAP conrol box. We sowed hese in he rucksack, seaed Maria in one of he fron seas and mouned a wo-channel acceleraion sensor on he car. The sensor s x and y channels corresponded o acceleraions in he lef/righ and up/down direcions. We were no able o measure he hird componen, forward-backward. We could no find any way o moun he camera safely on he car, so we decided o go wihou i ime was running ou I ook he sea nex o Maria, leaving Achim as groundsaff o push he safey buon, while he official from he crew, siing in he conrol cabin, sared o winch us up. The Silver Sar s design uses a passive powering principle: The car is slowly pulled o a heigh of 73 meers and is hen released ino almos verical free fall. OK, i is no compleely verical bu when you are siing in he Silver Sar, i feels like i! The whole ride of 1.6 km is powered only by he kineic energy gained from lifing he car. Maria and I screamed more han once, especially during he firs, and seepes, downward secion. Afer a ride lasing approximaely 65 seconds, he car reurned o he plaform and we checked he equipmen. We saw ha boh channels of he acceleraion sensor had gone off he connecors o he FirsAmp had become loose. For he second ride, we fixed hem o he FirsAmp using leucoplas. 3... 2... 1... second ride. Unforunaely, his ime all he EEG channels wen off afer 15 seconds. However, he acceleraion sensor yielded a good signal on boh channels. The res of he working day on Monday was spen examining he daa using Analyzer 2. Three of he 16 EEG channels were compleely pollued by high-ampliude arifacs. A comparison of hese hree channels wih he acceleraion signals from he second ride showed ha hey seemed o be highly correlaed and indicaed each curve on he rollercoaser s pah. This migh indicae effecs due o defecive conacs. I has o be admied ha, in general, he daa qualiy was very low because physiological arifacs, mos of hem muscle arifacs, were superimposed on he EEG. We had repeaedly screamed ou loud and our bodies mus surely have acivaed every possible muscle o counerac he exernal forces. However, no exernal arifacs, such as 50 Hz mains signal, were presen. Blue Fire Ride For he second day s research, Maria and Achim had he idea of mouning he camera on a biker s helme using copious amouns of Gaffer Tape. During he Blue Fire experimen, Maria wore he helme while I was responsible for he EEG cap. The Blue Fire is a more modern ride and uses a sophisicaed principle o power he car: I is a caapul launch coaser. The car sars ou on a horizonal rack, where i is acceleraed by a 2500 horsepower linear moor which carries i from zero o 100 km/h in 2.5 secoooonds. Achim agreed wih he saff ha we could have wo rides in a row and hen ake a break o check he equipmen before finally aking wo more rides. The linear acceleraion phase a he sar is he mos exciing par. Then comes a seep horseshoe-shaped urn wih a oal heigh of approximaely 40 m followed by a full loop and a oal of hree screw urns which provide an enjoyable, exciing aerial experience wih a maximum acceleraion of 3.8G. Afer a ride of 45 seconds, i is over. Wow! Have you ever ridden a rollercoaser four imes in a row wihou even having o ge ou and back ino he car? Weren we lucky! During he second day, we did no experience any echnical problems excep for he fac ha he acceleraion sensor s x-channel came loose again. Even hough he daa qualiy was excellen in erms of exernal influences, here were sill a lo of muscle arifacs, in fac oo many for he daa o be useable for convenional EEG processing. We observed relaively srong exernal arifacs during he acceleraion phase. These were probably caused by he highpower linear moors. Daa Analysis Firs, we analyzed he inerference caused by he 2500 horsepower linear moors during he acceleraion phase. We exraced he 2.5 seconds afer he sar, a phase which was refleced by a huge ari- page 6 of 13
fac in all four rides. We compued he FFT based on 500-millisecond inervals averaged over all channels and all four rides. Fig. 1 illusraes his averaged FFT (logarihmic scale). The arifac can be described as a low-pass noise wih decreasing specral densiy in he band 0-100 Hz. We hen decided o evaluae he global specral EEG properies as a funcion of acceleraion. Firs, he EEG was filered in he band 0.5 32 Hz and he acceleraion signals were filered in he band 0.5-4 Hz using 48dB/Oc filers. Figure 2 depics he ime-domain signaure of he acceleraion sensor. All four rides yielded very similar paerns. The EEG s Pz channel was excluded because i conained arifacs which migh have been due o a defecive conac. For he remaining EEG channels, we compued an ICA o perform arifac rejecion. The Cz channel conained approximaely 14% arifacs and hese were discarded. We auomaically se markers on he maxima and minima of he acceleraion channels. These were hen used o generae 38 low-acceleraion segmens (group A) and 51 high-acceleraion segmens (group B), each of lengh 500 ms. An FFT wih a Hanning window was applied o boh groups and he group averages were compued. The resuling specra for hree example channels (O1, O2, Oz) are shown in Figure 1. I can be seen ha he values for high acceleraions (green lines) are noiceably higher han for low acceleraions. However, we do no claim o have found a significan difference in his simple experimen. We are also unable o say wheher his specral difference is due o he EEG solely or due o sources such as he EMG or due o exernal elecromagneic inerferences. Conclusion In his pilo sudy involving an EEG recorded during a rollercoaser ride, we clearly reached he echnical limis of EEG acquisiion. We are proud o say ha our hardware was able o wihsand exreme acceleraion forces of up o 3.8G. In he fuure, i will be an easy ask o solve minor problems such as loose connecors by improving he elecronics housing. The acicap s abiliy o measure he EEG when acive elecrodes and, moreover, acive shielding were employed mean ha exernal arifacs remained a a very low level. However, when dealing wih exernal power sources of several megawas, we were able o idenify he emporal locaion of he arifacs in he daa. A video, a par of he recorded daa and a repor on his exreme EEG Projec can also be found in our exreme EEG Daabase a /exreme_eeg.php?ab=2 Figure 1 Figure 2 Figure 3 Legend figure 1: BLACK: Toal average FFT specrum of he 2.5 second acceleraion phase. The arifac is 0-100 Hz broad-band noise wih decreasing ampliude. Legend figure 2: BLACK: The acceleraion sensor signal of he firs Blue Fire ride. The signals for all four rides are similar. Legend figure 3: RED: Power specral densiy for group A (low acceleraion). The hin line indicaes mean value plus sandard deviaion. GREEN: Power specral densiy for group B (high acceleraion). The hin line indicaes mean value plus sandard deviaion. page 7 of 13
Produc Developmens The acicap sysem has come of age by Bernd Pich I is a problem ha all of us know: New producs someimes spawn new problems. As is known, we were no immune from such problems wih acicap, bu finally hey have now been resolved for some ime. Admiedly, i was no possible o implemen he changes overnigh, bu a lo has happened around acicap over he pas year. So why are we only producing his summary now, you may be asking yourself. Well, we are of he opinion ha argumens, developmens and figures should no be jus a maer of heory. They should be suppored by pracical evidence. Of course, ha is only possible afer sufficien daa has been colleced, and ha akes ime. Wha changes have we already made o he aci- CAP? Firs and foremos, we have been using a new casing compound and housing maerial for some ime. This measure alone made i possible o reduce he failure rae of he acicap sysem o below 1 % wihin a year. (invesigaed before delivery). If a cusomer neverheless idenifies a faul in heir acicap sysem, he new producion and repair deparmen ha we have esablished a our Gilching headquarers allows us o recify he faul wihin a very shor ime (approx. 2 working days) or o supply he cusomer wih an equivalen replacemen. In addiion, numerous small changes and improvemens ha have also been made make he acicap sysem simple and fas o work wih. Take for example he newly developed cable clips ha preven he individual elecrode cables from wising ogeher, or he reaining clip ha is now fied o boh he acicap conrol box and he splier box. The clips permi he boxes o be simply clipped o he es subjec s clohing during he experimen. The complee reworking of he acicap elecrode housing shape makes i far easier o clean he acicap elecrodes. Wha changes are planned for he acicap in fuure? Even if a lo has already been done, and he acicap sysem is now sable and reliable in operaion, we shall coninue sriving o expand and opimize he sysem ino he fuure. We are planning o launch for special applicaion cases, such as acicap modules especially for BCI. Addiional equipmen, such as special children s caps are also under consideraion, and he sysemaic inegraion of he acicap sysem in our sofware will coninue o pogress (e.g. impedance measuremen direc from our BrainVision Recorder sofware). One impressive piece of evidence ha we are moving in he righ direcion is our exreme EEG projec, where he acicap sysem was able o demonsrae is qualiies in many differen ways. Even under exreme condiions (EEG acquisiion in he Himalayas or on a roller coaser), our cusomers and we were always able o rely on he sabiliy of he sysem and on he acquisiion of high-qualiy daa. Bu despie all his, we shall coninue o srive for he perfec acicap sysem... News in brief: Downloads, Programs and Updaes Oc 22nd, 2009 / New Analyzer 2 Updae available! Analyzer 2 now suppors Windows 7. Moreover his updae fixes a bug in he Progress Bar of Analyzer 2. Please visi our Analyzer 2 Download Secion (/downloads.php?kid=9&ab=2) o download he updae. Nov 6h, 2009 / Updaed versions of RDA Cliens available Updaed versions of he RDA cliens for MATLAB and Pyhon can be downloaded in our Recorder Download Area (/ downloads.php?kid=2&ab=5). Boh updaes include he consideraion of daa resoluion in example calculaion. Nov 23rd, 2009 / Updaed V-Amp/FirsAmp Conrol Componen available An updaed version of he V-Amp/FirsAmp Conrol Componen can be found in our Recorder Download Secion (/ downloads.php?kid=2&ab=4). Conained bugfixes are: >> impedance measuremen for Ground and Reference elecrodes >> ener of decimals ino he workspace-field [Gradien] for AUX-Channel All Updaes and New Modules can be downloaded on our websie a /downloads.php. If you d like us o keep you posed on any new Updae for BrainVision Analyzer 2, please regiser for our Analyzer 2 Newsflash a /a2_newsflash.php page 8 of 13
Brain Producs Disribuors News in Norh America Widespread changes for Brain Vision LLC by Parick Briz, Brain Vision LLC Since 2004,Brain Vision LLC has been he sole Norh American disribuor of Brain Producs echnology and equipmen. Recenly, here have been many new and exciing developmens. My name is Parick Briz, formerly responsible for suppor, workshops and sales a Brain Producs German headquarer in Gilching. Some of you may remember my name from aricles I have conribued o previous newsleers. On July 1s 2009, I ook over responsibiliy for he day-o-day business operaions of Brain Vision LLC from Ingmar Guberle, he name many of you have come o associae wih Brain Vision LLC in recen years. Ingmar has no lef us, bu will now be focusing on his asks as he Cusomer and Scienific Suppor Manager for Brain Producs. I, on he oher hand, am looking forward o coninuing he excellen service Brain Vision LLC has provided o he Norh American marke in he pas. Furher developmens have been made necessary by he consisen rapid rise in demand for Brain Producs hard- and sofware soluions in he USA. To ensure a high and reliable level of service, he eam has been joined by wo new moivaed members. My assisan, Grechen Koch, joined Brain Vision LLC afer graduaing in biochemisry from Penn Sae Universiy. Naalie Dion is Brain Vision s new office manager and brings us her exensive and valuable experience in he efficien conduc of modern business managemen. Finally, he recruimen of so many new personnel mean ha we definiely needed a bigger office. As he ime had come for us o move from our exising premises, we decided o relocae he company o he Eas Coas. Therefore, I should like o ask you o noe Brain Vision s new address: Brain Vision LLC, Meridian Parkway, Suie 300, Durham, NC 27713, Unied Saes Please le us know how we can be of service o you - we are always delighed o help! Suppor Tip BooCamp - BrainVision Analyzer on Apple Mac Sysems by David Kadlec, MES Medizinechnik GmbH One frequenly asked quesion is wheher our Brain Producs sofware and Analyzer 2.0 in paricular works wih Apple Mac Sysems. I has been possible o run Windows on Apple Mac Sysems as of 2006, since when users have been able o insall, connec and work wih he enire Brain Producs range on Apple hardware. The reason for his is he change in Apple s chip supplier. Apple has now swiched o Inel echnology and, as a resul, i is now possible o insall he Windows XP, Windows Visa and Windows 7 operaing sysems wihou problems. Broadly speaking, here are wo ways o use our sofware and hardware producs on a Mac. The firs is o insall a Windows version wih Boo Camp. When using Boo Camp, you mus firs sar he sysem in naive mode because his sofware is no an emulaion produc like Parallels Deskop, VMware Fusion, Virual Box. If Win- Pic 1: Using Analyzer 2.0 in VMware and Safari-Browser in Mac-OS on a 13 MacBook Pic. 2: Analyzer 2.0 in VMware mode in full screen image dows is insalled correcly, ogeher wih all he drivers for he USB pors, hen we can confirm you can use our sysems wihou problems. The greaes inconvenience when working wih Windows as naive sofware is ha no one-click swichover o he MAC OS sofware is possible. You firs have o shu down Windows and hen resar in Mac mode. This operaion always akes abou 5 minues. If you prefer o work wih emulaion sofware or a virual machine, hen we sugges ha you insall he emulaion sofware Parallels Deskop or VMware Fusion. Wha is a virual machine? I is he sofware equivalen of a real physical PC on which you can insall and run operaing sysems and applicaions. In he case of VMware and Parallels, a virual machine corresponds o a PC. You can hink of a virual machine as an ordinary PC ha is insalled inside anoher PC and imiaes he acions of he various hardware componens page 9 of 13
presen in he hos PC such as he processor, RAM and hard disk drive. The Mac on which he emulaion is running is referred o as he hos and he virual machine running in i is known as he gues. The argumens in favor of hese soluions are he fas swiching beween MAC OS and Windows, i.e. you can observe boh machines in real ime if necessary, i is no necessary o resar your MAC, and you can inegrae your insalled Windows Boo Camp configuraion. Runimes/operaing imes are no a concern. Even hough Windows is emulaed, he performance of daa analyses ec. suffer from no complicaions or appreciable delays. So far, here is no way of launching he insalled sofware wih Xover which does no need o boo Windows XP. Using his mehod, you sar programs wih MAC OS direcly from he hard disk. However, our sofware canno be run in his way. One problem experienced by mos new users of Apple hardware is working wih he Apple keyboard. However, our sofware is designed for he usual Windows keyboards. For example, users searching for he delee (Forward Delee) key are someimes confused by he fac ha here is no such key on he Apple MacBook and MacBook Pro builin keyboards. The mos convenien soluion is o work wih he righ/ lef mouse buons wih an exernal Windows keyboard conneced. Pic. 3: Analyzer 2.0 in he background in VMware 3.0 mode. VMware 3.0 is running he Boo Camp pariion. Because PC and Mac keyboards differ, you have o press cerain key combinaions o access various PC commands on MacBook, MacBook Pro or exernal Apple keyboards. You should noe ha o access he funcion keys, you mus press he Fn key plus he required funcion key. Please refer o he able below. PC-Keyboard Apple, exernal Keyboard MacBook and MacBook Pro inergraed Keyboard Prin F14 Scroll Lock Break/Suspend F15 F16 Backspace key Delee Delee Delee (forward deleion) Delee x> Fn + Delee Pase Cmd+V Cmd+V Num Lock Num Fun+num lock F6 Command (Windows-Logo-Key) Apple Logo Key Apple Logo Key Al Al-Key Al-Key page 10 of 13
Brain Producs Disribuors Brain Producs 6h Disribuor Meeing by Pierluigi Casellone and Sefanie Rudrich I has been a long radiion a Brain Producs o regularly ge ogeher wih our disribuors from all over he world. During hese 2-day meeings ypically he recen pas is reviewed, new produc developmens are presened and poenial problems as well as heir soluions are discussed. Brain Producs 6h Disribuor Meeing was held jus a couple of weeks ago (November 16h/17h, 2009) a he Holiday Inn Hoel in Munich / Germany. 25 aendees represening 21 counries were joining us and eager o see wha Brain Producs has sofware- and hardware-wise in he pipeline for 2010. Bu, le s see more in deail wha he given alks were abou! facs. The consrucion of he PowerMAG minimizes he impac of he coil recharge arifac as well as mains sysem noise on EEG daa. Moreover he Sofaxic (neuronavigaion sysem) does no only help users o aim for he righ arge area bu also records coninuously he posiion of he TMS coil wih respec o he head posiion during he whole measuremen. This informaion can be impored in our BrainVision Analyzer 2 sofware and used hroughou he EEG daa analysis. For example during a TMS evoked poenial analysis his funcion is very useful o discard all he epochs in which he posiion of he TMS coil deviaed oo much from he arge area. If we add o he above he fac ha when using our BrainAmp sysems scieniss have found physiological aciviy jus 5 msec afer a TMS simulaion, i is clear ha he combinaion of hese 3 producs is ruly somehing unique in he world of EEG/TMS measuremens. Day 2 In his second day of our meeing he given alks focused on opics like wha Brain Producs offers for BCI research (now and in he fuure), MRI safey aspecs ha need o be aken ino consideraion when combining EEG & fmri as well as which soluions he company EasyCap (EEG cap maker) offers for differen research applicaions. Day 1 The schedule of he firs meeing day focused on he presenaion of upcoming hard- and sofware developmens as well as our new soluions for EEG recording during TMS simulaion. The laer is an exremely ineresing applicaion for which Brain Producs and is new business parners MAG & More (TMS Simulaor maker) and EMS (Neuronavigaion Sysem manufacurer) have invesed a lo of resources. The idea is o offer our cusomers a produc bundle including allha-you-need for advanced EEG and TMS co-regisraion. As described more in deail in anoher aricle in his Press Release, phase synchronizing he maser clock of he PowerMAG (TMS simulaor) wih he BrainAmp s inernal clock is one of he unique feaures offered by his produc combinaion. Recording drif-free TMS pulses and markers allows very accurae evoked poenial analysis as well as opens up new scenarios in he correcion of he TMS-induced ari- A lo of aenion was raised by he presenaion of a P300 speller (see relaed aricle in his issue of our Press Release), a sofware applicaion ha allows spelling words by using he brain elecrical aciviy measured a he scalp level. Our colleague Michael Jachan, who wore a 16-channel acicap sysem conneced o a V-Amp 16 amplifier, was able o spell words suggesed direcly by he enhusiasic audience. Alhough he meeing program was prey packed on boh days we waned o make sure - as always - ha i s no all abou work and business only. Therefore we organized a bowling ournamen in he evening of he firs day and he diversion was very well acceped by everybody. We ruly hope everybody aending enjoyed he meeing as much as we did and would like o hank all aendees again for heir valuable inpus. We are looking forward o keep fruifully collaboraing wih all our disribuors in he fuure and hope o see many of hem again a our nex disribuor meeing. page 11 of 13
Brain Producs Workshops Brain Producs EEG & fmri Workshop by Dr. Lisa Glass and Sefanie Rudrich As in previous years Brain Producs exhibied a he annual meeing of he Sociey for Psychophysiological Research (SPR). This year s conference, which was held in Berlin / Germany from Ocober 21s 24h, araced almos 1000 scieniss from all over he world way more han in pas years. Many aendees no only ook ime o aend he various sessions bu also frequenly visied he echnical exhibiion ha accompanied he meeing. As we assumed ha many of our cusomers would join he 49h Annual Meeing of he SPR, we also held an EEG & fmri workshop a he CHARITÉ Berlin righ afer he conference (Ocober 25h). The program of he workshop included various opics ranging from Pain & Gain of EEG & fmri o he presenaion of new approaches for fmri imaging-arifac correcion in ulrahigh-frequency EEG. In a live demonsraion direcly in he MR-Scanner, he proper preparaion of an EEG & fmri recording was shown, crucial experimenal issues discussed, and hen - during he recording - he impressive online arifac correcion was demonsraed. These daa recordings were used aferwards during an ineracive session o show how o successfully correc he daa offline for various ypes of MR-relaed arifacs. Among he speakers were Brain Producs Dr. Ingmar Guberle as well as Dr. Pera Rier, Rober Becker (M.Sc.) and Frank Freyer (M.Sc.) from he EEG-fMRI Group of he Berlin Dr. Pera Rier Neuroimaging Cener, whom we would like o hank very much for his excellen workshop. Once he workshop organized in cooperaion wih he CHARITÉ s Deparmen of Neurology was announced, he run on he few seas available was jus amazing. Wihin a few hours of announcing he workshop, all of he places were booked ou. This has convinced us ha we should hold such special opic 1-day workshops (in addiion o our regular 2-day workshops) more ofen in he fuure. If you are ineresed in any of our workshops please remember o regiser your ineres hrough our websie a / wsreques.php or email us a workshop@brainpoducs.com News in brief: Workshops Analyzer 2 User Workshop (deailed opics TBA soon) Analyzer 2 Workshop on EEG & fmri Denver (USA), February 23rd o 25h Edinburgh (Scoland), March 4h & 5h For more deailed informaion on upcoming workshops please visi our websie a /workshops.php To announce your ineres in aending any workshop we offer, please regiser a /wsreques.php News in brief: Conferences & Exhibiions EEG-Tage Munich (Germany), February 24h o 26h in cooperaion wih MES Medizinelekronik GmbH DGKN Halle (Germany), March 18h o 20h in cooperaion wih MES Medizinelekronik GmbH TeaP Saarbrücken (Germany), March 22nd 24h in cooperaion wih MES Medizinelekronik GmbH CNS Monreal (Canada), April 17h o 20h in cooperaion wih Brain Vision LLC ISMRM Sockholm (Sweden), May 1s o 7h Human Brain Mapping Barcelona (Spain), June 6h o 10h For more informaion on he conferences we are abou o aend, please visi our websie a /evens.php page 12 of 13
Brain Producs Inside Who is who: Ingo Wesermayer, IT Specialis Trainee Having jus finished school Ingo Wesermayer decided o become an IT Specialis and applied a Brain Producs IT deparmen. Being an IT enhusias for a long ime, we are very confiden ha he will be a grea aid supporing our IT Manager, Illya Yaroshenko, and will succeed in his appreniceship. Welcome o our eam, Ingo! Who is who: Benjamin Spinnler, Logisics Deparmen We are happy o welcome Benjamin Spinnler o our eam! Having worked in he shipping and logisics deparmen of an aerospace company in previous years Benjamin will now suppor our Logisics Manager, Helmu Luz. Benjamin is replacing Andreas Schmid whom we wish good luck for his personal and professional fuure. Who is who: Chrisiane Srasser, Office Managemen Chrisiane Srasser used o work as an adminisraive assisan in he pharmaceuical indusry wihin he pas four years. From now on she will suppor Monika Heiß in order processing and office managemen and hus replace Juliane Riedel. Juliane lef he company o pursue a career in he markeing field. Welcome on board, Chrisiane! Who is who: Norber Hauser, Sofware Developmen Our sofware developmen eam is growing furher! Saring in Ocober, Norber Hauser will suppor our Recorder sofware developmen manager, Dinh- Phuong Nguyen, in advancing our sofware and drivers. Having developed inerfaces and hardware conrollers in he field of neurological diagnosics for abou 20 years now, Norber is an exremely experienced sofware programmer and we are very happy o welcome him o our eam! This Press Release is published by Brain Producs GmbH, Zeppelinsrasse 7, 82205 Gilching, Germany. Phone +49 (0) 8105 733 84 0, Noice of Righs All righs reserved in he even of he gran of a paen, uiliy model or design. For informaion on geing permission for reprins and excerps, conac markeing@brainproducs.com. Unauhorized reproducion of hese works is illegal, and may be subjec o prosecuion. Noice of Liabiliy The informaion in his press release is disribued on as As Is basis, wihou warrany. While every precauion has been aken in he preparaion of his press release, neiher he auhors nor Brain Producs GmbH, shall have any liabiliy o any person or eniy wih respec o any loss or damage caused or alleged o be caused direcly or indirecly by he insrucions conained in his book or by he compuer sofware and hardware producs decribed here. Copyrigh 2009 by Brain Producs GmbH page 13 of 13