1. EEG The electroencephalogram (EEG) measures the activity of large numbers (populations) of neurons. First recorded by Hans Berger in 1929. EEG recordings are noninvasive, painless, do not interfere much with a human subject’s ability to move or perceive stimuli, are relatively low-cost. Electrodes measure voltage-differences at the scalp in the microvolt (μV) range. Voltage-traces are recorded with millisecond resolution – great advantage over brain imaging (fMRI or PET).

2. EEG Standard placements of electrodes on the human scalp: A, auricle; C, central; F, frontal; Fp, frontal pole; O, occipital; P, parietal; T, temporal.

3. EEG

5. Many neurons need to sum their activity in order to be detected by EEG electrodes. The timing of their activity is crucial. Synchronized neural activity produces larger signals.

6. The Electroencephalogram A simple circuit to generate rhythmic activity

7. The Electroencephalogram Two ways of generating synchronicity: a) pacemaker; b) mutual coordination 1600 oscillators (excitatory cells) un-coordinatedcoordinated

8. EEG EEG potentials are good indicators of global brain state. They often display rhythmic patterns at characteristic frequencies

9. EEG EEG suffers from poor current source localization and the “inverse problem”

10. EEG EEG rhythms correlate with patterns of behavior (level of attentiveness, sleeping, waking, seizures, coma). Rhythms occur in distinct frequency ranges: Gamma: 20-60 Hz (“cognitive” frequency band) Beta: 14-20 Hz (activated cortex) Alpha: 8-13 Hz (quiet waking) Theta: 4-7 Hz (sleep stages) Delta: less than 4 Hz (sleep stages, especially “deep sleep”) Higher frequencies: active processing, relatively de-synchronized activity (alert wakefulness, dream sleep). Lower frequencies: strongly synchronized activity (nondreaming sleep, coma).

11. EEG Power spectrum:

12. EEG - ERP ERP’s are obtained after averaging EEG signals obtained over multiple trials (trials are aligned by stimulus onset).

13. MEG The MEG laboratory Images courtesy of CTF Systems Inc.

14. MEG Measures changes in magnetic fields that accompany electrical activity.

15. MEG An example (auditory task):

19. MEG Three task conditions: MEG results 1 -- Listening to tones that were delivered with a delay of about 5s. A random time was added to prevent stimulus prediction. The signal is an average over about 80 stimulus presentations. 2 -- Reacting to acoustic stimuli. The same stimulus presentation as in (1) but now the subject was told to press on an air cushion as soon as possible after the tone was heard. 3 -- Synchronizing with a rhythm. Here the tones were presented regularly with a frequency of 1 Hz. The subject was told to press the air cushion in synchrony with the stimulus. 123 Viktor Jirsa (FAU): http://www.ccs.fau.edu/~jirsa/Imaging.html

20. Sensing Techniques-Cat Scans

22. Cat Scans use x-rays to show structures Really precise maps Hard to determine functions

23. PET Positron Emission Tomography Requires the injection of a positron-emitting radioactive isotope (tracer) Examples: C-11 Glucose analogs (metabolism) O-15 water (blood flow or volume) C-11 or O-15 carbon monoxide PET tracers must have short half-life, e.g. C-11 (20 min.), O-15 (2 min.). Cyclotron! Positron + electron  2 gamma ray beams. Gamma radiation is detected by ring of detectors, source is plotted in 2-D producing an image slice.

24. Sensing Techniques-PET Radioactive element decays, gives off positron Positron moves a short distance and gives off two gamma rays in opposite directions

25. Sensing Techniques-PET

26. PET Scans Eyes ClosedWhite LightComplex Scene

27. PET Scans Episodic Task Semantic Task Difference

28. PET Scans

29. PET

30. PET - Examples In cognitive studies, a subtraction paradigm is often used.

31. PET - Examples Another example of control and task states, and of averaging over subjects: Marc Raichle

32. PET - Examples M. Raichle (a) Passive viewing of nouns; (b) Hearing of nouns; (c ) Spoken nouns minus viewed or heard nouns; (d) Generating verbs.

33. PET - Examples M. Raichle

34. PET - Examples M. Raichle PET images taken at different times, e.g. during learning, can be compared.

35. PET - Examples PET images are pretty to look at... … and can be combined with other imaging modalities, here MRI.

36. Functional Magnetic Resonance Imaging Typical MRI Scanner

37. MRI - fMRI Subjects are placed in a strong external magnetic field. Spin axes of nuclei orient within the field. External RF pulse is applied. Spin axes reorient, then relax. During relaxation time, nuclei send out pulses, which differ depending on the microenvironment (e.g. water/fat ratio). The Physics (sort of)... fMRI – functional MRI Allows fast acquisition of a complete image slice in as little as 20 ms. Several slices are acquired in rapid succession and the data is examined for statistical differences. Hemoglobin is “brighter” than deoxyhemoglobin. Oxygenated blood is “brighter” - active areas are “brighter”. BOLD-fMRI

38. PET - MRI in Comparison

39. Sensing Techniques-fMRI Functional Magnetic Nuclear Resonance Imaging Similar to Pet, uses radio frequency information given off by water Gives better time (6 seconds) and spatial (2 mm) resolution than Pet or Cat scans Technology still under revision Slight danger to subjects Expensive ($300 an hour)

40. fMRI Method

41. MRI Scans

43. fMRI - Examples

44. Stimulus: “checkerboard pattern” V1 responses Jezzard/Friston 1994

46. fMRI High-Resolution Mapping Kim et al., 2000 BOLD fMRI in cat cortex (level of area 18)

47. fMRI High-Resolution Mapping Kim et al., 2000

48. PET and fMRI - Similarities and Differences - Different biological signal. Yet, both pick up a signal related to bulk metabolism (not electricity). - fMRI has better temporal (<100 ms) and spatial resolution (1 mm and less) - fMRI does not involve radioactive tracers and subjects can be measured repeatedly, over many trials. - PET images generally represent “idealized averages”. fMRI images are often registered with structural scans to show individual anatomy. - For both, images can be aligned for multiple subjects. - fMRI is widely available, PET is not. - fMRI does not allow localization of neurotransmitters or receptors etc. - For both, it can be tricky to get stimuli to the subject.

49. Data Analysis Issues Neuroimaging (PET/fMRI): Activation values, spatial resolution, averaging, image alignment and registration. EEG/MEG: Current source localization (inverse problem), time domain data sets, frequency power spectrum, correlation and coherency.

50. Summary Appropriate technology depends on question ERP has good temporal resolution CAT, Pet, MRI have good to fair spatial resolution, only PET has any functional capture fMRI has reasonably good spatial, reasonably good temporal, but is expensive