Pre-processing and the role of horses in the history of M/EEG

1. Pre-processingand the role of horses in the history ofM/EEG Vladimir Litvak Wellcome Trust Centre for Neuroimaging UCL Institute of Neurology

2. At the beginning there was a horse Berger originally had intended to study astronomy. While he was serving in the German army in the early 1890s, his horse slipped down an embankment, nearly seriously injuring Berger. His sister many miles away had a feeling he was in danger and got her father to telegram him. This astonished him so much that he switched to study psychology. (Blakemore, 1977) Hans Berger 1873-1941

3. What do we need? Events Possible source of artefact M/EEG signals Time axis (sampling frequency and onset) Sensor locations

4. ~300 sensors <128 electrodes

5. Evoked response vs. spontaneous activity active awake state pre-stim post-stim resting state falling asleep sleep deep sleep coma 50 uV Averaging 1 sec evoked response ongoing rythms

6. Conversion MEG CTF Neuromag BTi-4D Yokogawa EEG EEGLAB Biosemi Brainvision Neuroscan EGI ANT SPM5 … Biosig Arbitrary ASCII Matlab Fieldtrip raw timelock freq script GUI Convert button (fileio) spm_eeg_convert spm_eeg_ft2spm SPM8 dataset *.dat – binary data file *.mat – header file

7. Expert’s corner • The *.mat file contains a struct, named D, which is converted to an meeg object by spm_eeg_load. • The *.dat file is memory-mapped and linked to the object. • Special functions called ‘methods’ provide a simple interface for getting information from the object and updating it and ensure that the header data remain consistent.

10. Now lets take a step back

11. Sensor locations MEG: • Requires quite complex sensor representation including locations and orientations of the coils and the way MEG channels are derived from the sensors. • Sensor representation is read automatically from the original dataset at conversion. EEG: • Presently only requires electrode locations. In the future will also include a montage matrix to represent different referencing arrangements. • Usually electrode locations do not come with the EEG data. • SPM assigns default electrode locations for some common systems (extended 10-20, Biosemi, EGI – with user’s input). • Individually measured locations can be loaded; requires co-registration.

12. Understanding coordinate systems Coordinate systems can differ in their origin, units and orientation. • MNI coordinates are defined using landmarks inside the brain. • Advantage: locations can be related to anatomy • Disadvantage: co-registration to a structural scan is required • Head coordinates are defined based on the fiducials. Commonly used for MEG, but the definition differs between different MEG systems. • Advantage: once the location of the head is expressed in head coordinates, it can be combined with sensor locations even if the subject moves. • Disadvantage: requires fiducials; if the fiducials are moved, the coordinate system changes. • Device coordinates are defined relative to some point external to the subject and fixed with respect to the measuring device. • Advantage: head locations can be compared between different experiments and subjects. • Disadvantage: head location needs to be tracked.

13. Understanding coordinate systems In SPM8 • Before co-registration • MEG sensors are represented in head coordinates in mm. • EEG sensors can be represented in any Cartesian coordinate system. Units are transformed to mm. • After co-registration • MEG sensor representation does not change. The head model is transformed to head coordinates. • EEG sensors are transformed to MNI coordinates.

14. Epoching Definition: Cutting segments around events. Need to know: • What happens (event type, event value) • When it happens (time of the events) Need to define: • Segment borders • Trial type (can be different triggers => single trial type) Note: • SPM8 only supports fixed length trials (but there are ways to circumvent this). • The epoching function also performs baseline correction (using negative times as the baseline).

15. Filtering • High-pass – remove the DC offset and slow trends in the data. • Low-pass – remove high-frequency noise. Similar to smoothing. • Notch (band-stop) – remove artefacts limited in frequency, most commonly line noise and its harmonics. • Band-pass – focus on the frequency of interest and remove the rest. More suitable for relatively narrow frequency ranges.

16. Filtering - examples Unfiltered 5Hz high-pass 10Hz high-pass 45Hz low-pass 10Hz low-pass 20Hz low-pass

17. EEG – re-referencing Average reference

18. EEG – re-referencing • Re-referencing can be used to sensitize sensor level analysis to particular sources (at the expense of other sources). • For other purposes (source reconstruction and DCM) it is presently necessary to use average reference. This will be relaxed in the future. • Re-referencing in SPM8 is done by the Montage function that can apply any linear weighting to the channels and has a wider range of applications.

19. Artefacts Eye blink EEG MEG Planar

20. Artefacts • SPM8 has an extendable artefact detection function where plug-ins implementing different detection methods ca be applied to subsets of channels. • Presently, amplitude thresholding, jump detection and flat segment detection are implemented. • Plug-in contributions are welcome. • In addition, topography-based artefact correction method is available (in MEEGtools toolbox).

21. Averaging - horses, once again. In the 1870s Sir Francis Galton (1822-1911), became the first scientific sportsman. He derived the common facial features of winning horses by photographically superimposing heads of race-winning thoroughbreds. In addition to its innovation to sports, this is a first in the extraction of common features and blurring of noise. Encouraged by his horse racing success Galton went on to identify the common physical features in the faces of violent criminals and murderers of the 1880s. He hoped to be able to detect potential violent offenders before they committed their crimes. He superimposed photographs to obtain composites that, in this case, turned out to be nothing out of the ordinary.

22. Galton, 1880s

23. Robust averaging Kilner, unpublished Wager et al. Neuroimage, 2005

24. Robust averaging • Robust averaging is an iterative procedure that computes the mean, down-weights outliers, re-computes the mean etc. until convergence. • It relies on the assumption that for any channel and time point most trials are clean. • The number of trials should be sufficient to get a distribution (at least a few tens). • Robust averaging can be used either in combination with or as an alternative to trial rejection.

25. So that’s how we got to this point

26. Generating time x scalp images

27. Quiz Time What is the crucial difference between M/EEG and fMRI from the point of view of data analysis ?

28. How can we characterize this? ?

29. Fourier analysis • Joseph Fourier (1768-1830) • Any complex time series can be broken down into a series of superimposed sinusoids with different frequencies

30. Fourier analysis Original Different amplitude Different phase

31. Methods of spectral estimation – example 1 Hilbert transform Morlet 3 cycles Morlet 5 cycles Multitaper Morlet 7 cycles

32. Methods of spectral estimation – example 2 Morlet w. fixed window Morlet 5 cycles Optimized multitaper Hilbert transform Morlet 7 cycles Multitaper Morlet 9 cycles

33. Robust averaging for TF Unweighted averaging Robust averaging

34. TF rescaling Raw Log Rel [-7 -4]s Diff [-7 -4]s LogR [-7 -4]s Rel [-0.5 0.5]s

35. Thanks to: The people who contributed material to this presentation (knowingly or not): • Stefan Kiebel • Jean Daunizeau • Gareth Barnes • James Kilner • Robert Oostenveld • Hillel Pratt • Arnaud Delorme • Laurence Hunt and all the members of the methods group past and present

36. Wavelets Frequency