NEWS

NatMEG @ MEG UK

We would like to thank all the organizers of MEG UK at Aston University, Birmingham, for an superb and very pleasant conference with many new and old friends. The UK was clearly shown to be a leading place for MEG research, with nine (!) excellent MEG laboratories and top-level researchers in Nottingham, Cambridge, Glasgow, Derry, York, Oxford, Cardiff, UCL and Aston. NatMEG would especially like to congratulate and welcome to the MEG community the new MEG lab at Derry, University of Ulster, School of Computing and Intelligent Systems. They will soon be online and provide Northern Ireland with their first MEG lab.

A fabulous pre-conference FieldTrip & SPM workshop was hosted by Prof. Klaus Kessler and taught by Vladimir Litvak, Robert Oostenveld, Bernadette van Wijk, Gareth Barnes, Guillaume Flandin, Saskia Helbling, Ryszard Auksztulewicz, Johanna Zumer, Hongfang Wang and NatMEG’s Stephen Whitmarsh. Thanks to this workshop and excellent beamformer presentations, two improvements/issues to the beamformer analysis were identified that will soon be implemented in the FieldTrip beamformer by Robert Oostenveld (and tested by NatMEG):

  1. After using Maxfilter, the dimensionality of the data is reduced to 60-80 components, resulting in a rank-deficient cross-spectral density matrix (CSD, or covariance matrix when dealing with time-courses). This is not by itself a problem. However, when concatenating the data of different files/sessions – which is often the case in Elekta systems when recording for longer periods of time – the data is no-longer rank-deficient, but does contain components that are unique to each file/session. These should in most cases be considered unwanted. This means that after calculating the CSD or covariance matrix, the rank can (and should) be reduced to a maximum rank of the individual file/session, e.g. by using a PCA and keeping the first 60-80 components. This will be implemented in the Beamformer through a kappa parameter.
  2. Prof. Gary Green (York) proposed to normalize the beamformer through setting all non-diagonal values of the CSD/covariance matrix to zero when they are smaller than a set value, e.g. the smallest value of the diagonal. This will also be soon implemented in FieldTrip. You can find his paper here.

Finally, NatMEG presented it’s first poster. It was a great opportunity to get invaluable feedback, and we would very much like to thank those that visited and helped out in considering the future of the project, with special thanks to Prof. Uta Noppeney and Dr. Ali Mazaheri from the University of Birmingham, and Dr. Micah Allen from UCL. You can see the poster below.

Poster Birmingham Stephen Whitmarsh

NatMEG is proud of the evaluation of our one-week analysis workshop

What is your general impression about this workshop? 

  • As stated above, and also commented “by mouth” at the end of the work shop – very good planning and overall theme for the whole week, in combination with fantastic enthusiasm from all teachers, i.e. very very good!
  • Very good. Good program with, first a summary of the theoretical background and right after a hands-on session. Very interesting and enthusing keynote speakers especially for someone that has no clinical background. Also really brings forward the point that you can use MEG for practical/clinical purposes and not only research.
  • My overall impression is excellent, without a shadow of a doubt. All aspects deserve praise – from an idea of having a workshop “from EEG to MEG”, through the execution of theoretical and practical parts by an excellent team to interesting keynote lectures illustrating the use of MEG in clinical and research practice spiced up by methodological details. I also enjoyed multiple opportunities to interact with other participants who represented a wide range of interests in MEG as a tool in brain science. Teaching naturally was of particularly high quality (I am still wondering how it was possible to convey challenging issues/problems in such a digestible form). On this note, I would like to emphasize the golden balance between theoretical discussions, coverage of FieldTrip methods and hands-on data analysis practice. Since I had already been using FieldTrip and conducting signal analyses in my research work before the workshop, the pace suited me very well. I updated my knowledge and brushed up skills in effective use of the FieldTrip toolbox. At the end, I would like to thank the organizers for fulfilling their professional mission/commitment with passion and dedication, for didactical intuition, and sincere engagement and devotion to MEG community. I can only try to imagine how much work and hours were invested in this event, which makes me appreciate it even more.
  • My general impression was shaped by:
    • Meeting the right people at the right moment when I had questions concerning my current work in EEG and receiving the right answers from the hosts.
    • The density of internationally based guest lecturers in such a short time.
    • The opportunity of meeting other researchers in related fields. It is always exciting to meet people whose work I have read or cited. In numbers: 3 researchers whose work I have encountered and 4 potential collaborators.
    • The atmosphere of openness, of exchanging ideas and knowledge, was very different from my previous experiences at KI.
    • The facilities were overall impressive. Having dinner in the same building where O’Keefe will be holding his Nobel lecture was equally thrilling. Let me just say that not being able to open the windows for ventilation justified the coffee breaks, which were much appreciated.
    • The schedule was very well conceived.
    • The funding from Swedish Bioimaging…
    • The advanced skills I acquired in EEG/MEG data analysis
    • It was a fabulous top notch workshop, why? For several reasons:
      • Excellent teachers and excellent keynote speakers that connected the theory with research and applied research. Great recruiting!
      • We were given the right atmosphere to network and familiarize with our peers and teachers
      • Speed dating great idea, worked well
      • Having the welcome dinner and farewell lunch as well as the catering of coffee etc. in between really made it possible to network and address
      • contemporary and future study ideas/analysis etc.
      • Recording the Workshop was brilliant and was interestingly for me and three of my peers the way we did find this workshop in the first place.
      • Being quiet oblivious to MEG, besides having read the basic literature shared by Stephen, I really came to understand in what way MEG sets itself apart from EEG and
      • how it might therefore be the superior imaging tool in addressing certain research questions/ future projects.
      • It really got me thinking on the shortcomings of some of my EEG studies (of course relating to the methodological constraints) and therefore made me wonder throughout the week how some questions regarding my field of interest may be answered using MEG.
      • There is not much that can be improved upon, but one idea I would like to pitch. Regarding the hand on, it would be much easier for the teachers as well as the students if the teachers first of all run the entire script with all pupils watching what happens. When the teacher has gone Bandura on us (monkey see, monkey do), then the pupils might have more time to have a closer look at the code and do it themselves. Furthermore, little sticks and stones on the way would be easily picked up upon by the teachers and fixed instanteneously. Thus also reducing some time which would be spent running from one pupil to the next. But, the workshop was brilliant do not get me wrong, just an idea regarding the didactics.
  • I am really happy that I was a part of this workshop. I am very impressed by the whole structure of the workshop. First of all the teachers were very knowledgeable. Second, the mixture of the lectures on technical issues, hand-on sessions and lectures that gave a good grasp on application of the method was excellent . All together it was a great experience!!
  • Excellent. The program is broad and deep, the team dedicated and knowledgeable. It really made a difference, and I am confident that the full benefits will yet surface only with time and practice. I would really recommend the workshop to friends and co-workers. The time for the last tutorials (dipole fitting and beamforming) was a little short, however. One or two more hours in each would have been beneficial for me. However, it might also be too much, because fatigue was clearly kicking-in.
  • Top notch! Teaching of methods was very effective, made hard things easy and accessible. I love the wiki style updated tutorials, an example of how to accumulate knowledge in a practical way. My own experience of EEG is that methods need to be evaluated from what results they bring about. If I really try hard to find a critical angle on the workshop it could be that it was a bit method-heavy, just a short orientation about spectra dynamics in the brain could have been good. Perhaps a lecture titled: “What do we know about brain oscillations and their sources?”.
    • The workshop was perfect. Everything was very well organised. I was impressed by how each and every thing was planned and minor details were taken care of. The small participant group allowed a very good instructor to student ratio. The best part I liked about NatMEG is very warm welcoming and supporting attitude of you people. The welcome diner gave a good time to interact and to get to know each other. I would also like to appreciate the topic selection and the way they were presented. Starting from the very basic and gradually covering each and every step of the analysis along with the practical application highlighted by keynote speakers were amazing. Thank you so much for recording the lectures and for the hand-outs. Overall it was a great experience
    • General high academic level of lectures/workshops/keynote presentations, but with a great range from starting with importing data and then proceeding to more advanced operations, along with top-notch examples from the keynote speakers giving a hint of the further possibilities. Probably the best course I have attended.
    • Excellent, superb! Super condensed learning experience.  You are all very enthusiastic and pedagogic. This workshop was one of the best (research) weeks of my life (well, I am a nerd).
    • This was challenging! At the end, I would remember this as one of the most engaging workshop.
    •  Very well organized and an ambitious workshop!

 

What did this workshop do to your enthusiasm, confidence and practical plans of doing MEG research? 

  • As I personally didn´t have any research plans I will answer on behalf of our plans of work up for clinical epilepsy patients instead – and then I will say: I´ve become very much more confident and enthusiastic over the issue!
  • It did wet my appetite for more.
  • Obviously, seeing the engagement and strong competence of the NatMEG team helps in growing enthusiasm even though I already was convinced of the importance of MEG for brain studies at the start of the workshop. The key lectures broadened my perception of the MEG applicability in neuroscience.  In addition, having been exposed to advanced signal analysis methods from both theoretical and practical perspective I undoubtedly became more confident and positive about actually using MEG data in my research. I realized that with a systematic and informed approach, maybe with some guidance by NatMEG people ;), I am in a position to harness MEG.
  • Some studies have shown that reflecting on enthusiasm might interfere with the actual enthusiasm, so I will simply continue on to the remaining points; I have become more confident in my EEG work from points of design, acquisition, and future steps for analysis. Concerning MEG I am now convinced that I would be able to conceive and realize a study in all its parts, under some guidance. Although my practical plans on starting a MEG-based project have been limited due to the availability of MEG scanners, in Germany and elsewhere, MEG research now seems much more realistic.
  • Well, although complex and more ‘difficult’ to interpret than EEG, the course gave me sufficient insight into the method such that I feel confident in getting to work with MEG. Of course there remains a great amount to learn but if there is one lab member that has such practical experience as for instance Stephen, and whose brains one may pick once in a while, then I could definitely kick off a project.
  • I think that it made me believe that the method is worth to use but also that it is accessible. So, definitely it made me enthusiastic about possible use of the method in my research.
  • It raised my enthusiasm, helped me build confidence in analysing my own data, and will certainly influence what I’ll be doing from now on, since I have MEG data ready to start! In this regard, I would recommend having one period, e.g. the last afternoon, to split the students in small subgroups to directly discuss what people are actually doing in their labs and how they could approach their data analysis. For example, in this course a reasonable division in subgroups could have been “epilepsy”, “ERP/ERFs” and “resting state/frequency analysis”. People working in each of these areas could thus spend around one or two hours talking with one of the teachers about specific questions and difficulties they have, and how to solve them. I think this would help people to go back with a very practical direction for their work in the short term
  • For me, MEG seems an accessible method I would like to use, more so than before the workshop. I have mostly stayed away from spectral and source analysis because I thought it was too much to learn, I don’t feel like that now. If I could I would start implementing a MEG-experiment right now. In practice it is quite tough for me to find time for such projects. I need to integrate my research with other tasks at my job, such as teaching or taking part in existing projects at the lab. I’ll try to work something out.
  • The workshop gave me a good picture of how to move forward with MEG analysis and gave a good boost to my enthusiasm. Thank you so much for organising it and giving us an opportunity to learn from the experts :)
  • I am a little more confident i setting up a processing pipeline. Good to also have seen and discussed some of the pitfalls and limitations in MEG – e.g. such as the different leakage problems. Definitely something to take into account in designing my experiment.
  • I now feel confident in reading articles. I am very excited about MEG research. At the same time, it is intimidating to see how complicated it can get. At the least, I do not feel intimidated anymore by MEG and will dare to think big (i.e., MEG is not only for elite researchers at OTHER universities).
  • I look forward to doing work in MEG. Specifically, exploiting MEG for answering my own questions.
  • I know more about the possibilities of MEG for source analysis and connectivity methods and hope to be able to use it in the near future.

 

The efforts of NatMEG as a national MEG research facility – are we going in the right direction? 

  • Yes, I think you are!
  • I think so. Educating the research community must be the key. When researchers know that it exists, and what the possibilities and limitations with MEG, then you are in the minds of the broader community and not just the few who are already enthusiastic about MEG.
  • It is hard to judge the national aspect but it is obvious to me that you have strong ambition, commitment and passion to popularize MEG use in research and clinical settings. Your willingness/dedication to involve scientists as well as practitioners from different fields and having a wide range of different competences is truly impressive. This in my eyes however goes well beyond the recent workshop.
  • I was very impressed by the many options to acquire data for different sensory modalities and in different domains of research. The options seem endless, but because of the early stages of the new facility it appears difficult to gauge the capacity limit of NatMEG. With applications in both clinical and cognitive neuroscience I think the centre is on the right way. I did not get to look too far behind the scenes and thus it is difficult to estimate, but in the future there might be need for permanent supporting staff, such as a research assistant or a technician with empirical experience.
  • I am pretty sure that given these workshop you will have a problem with too many people wanting to use the lab in the near future, no doubt.
  • With regard to hardware, there is little to nothing missing. Well, I of course would like to combine brain stimulation with MEG and investigate connectivity changes due to this. Therefore, having a TMS device or, since cheaper and quiet handy, tDSC for that matter would be great. But I am positive that our institute could borrow the device for common projects, as I am basically the only one working with TMS at our institute. What would be extremely interesting though is to investigate and compare the long/intermediate term effects of both brain stimulation methods using MEG (Stimulation of course offline), as the neurophysiology behind this remains unclear and is only understood in theoretical terms.
  • Definitely yes. It is really fantastic that NatMEG is so open for cooperation’s. Also, it is great that NatMEG wants to be involved and provide some help in planning an conducting study. This all makes researchers, that are beginners in the use of MEG more confidence to even consider this method in the further studies.
  • Seems totally in the right direction. A carefully equipped lab, with good computers, data storage carefully planned, analysis pipelines available, expertise in code programming for experimental designs, all available and ready to become a famous lab! It can certainly improve on the project and research of all people that come to use the facility, way beyond data acquisiton itself
  • I don’t know that well how you are doing, that is: If interesting projects are being proposed and realised in the right time. Perhaps we could start up some kind of EEG/MEG community to strengthen the role of NATMEG as a kind of hub for these methods? In the end I think what will matter is how many PhD students in Sweden will use MEG as a major method. I’d love to steer some of them towards MEG.
  • I hope so :)
  • Absolutely! These hands-on sessions combined with state of the art research talks just is the best. Suggestion: How about mini workshops (2-3 days) with a particular focus (e.g., oscillations). This could include discussions of data among the students.
  • Yes. And possibly towards an international level.
  • Absolutely! It is important to teach about the methodological possibilities, in order to make people think about the right questions to ask.

NatMEG is very happy to announce yet another unique one-week analysis workshop

MEGEEG1

 MEGEEG2

 

Our previous workshop was a great succes!

The link between lectures and hands-on [were most useful]. Demos were also nice breaks, but what really makes them useful is being able to go through them myself, and playing around

From knowing very little about MEG analysis, I now feel I have the tools to design a MEG study and also to analyse data from it. Not in a fluent way (like there are any such studies), but I feel equipped to start and to know how to formulate relevant questions when running into problem.

ki-sw-057

 

 

 

 

Analyzing Neural Time Series: Theory and Practice

NatMEG congratulates Mike X. Cohen with his new book! We are planning a seminar by Mike at NatMEG. Order the book now!

bookcover cohen