L5 - Alternative Methods Flashcards
1
Q
How many APs can occur in one second?
A
- AP takes 3 ms
- Max APs = 1000/3 = 333
- Neurons are limited to 333 spikes per second
2
Q
Explain the 10-20 System of Electrode Placement.
A
- Reference electrodes placed on the ear / mastoid behind the ears
- Measure from nation to inion: first and last 10% are where first and last electrodes go – remaining 3 separated by 20%
- Same process from ear to ear and around the head
- Z represents sagittal midline: broken down into frontal, central, and parietal
- RH even numbers, LH odd numbers
3
Q
What are the two common EEG montages?
A
- Montage = position that electrodes are glued onto the scalp
- Bipolar montage: Two electrodes per amplifier channel (each electrode has its own reference)
- Referential montage: One common reference electrode for all electrodes
4
Q
How does the differential operational amplifier work?
A
- Two signals go into one amplifier – one positive and one negative
- If same voltage enters both, one cancels out the other. This cancels any shared voltage between the two electrodes (common mode)
- Common mode is the “noise” while the differences are averaged over many trials
5
Q
What is an Event Related Potential?
A
- Brief change in slow-wave EEG signal in response to a sensory stimulus
- EEG response in a given channel is digitized by sampling the voltage across the electrode / reference pair at a sampling frequency of at least 2f (f is highest frequency of interest for the signal)
- If you take many recordings at each point in time and add them together, creates signal averaging
6
Q
Explain how signal averaging is used in EEG
A
- Each stimulus presentation produces a small signal response embedded in noise
- At any instant in time, the voltage measured across the electrode of interest and its reference electrode consists of a small signal and a large amount of noise
- Signal is neural response while noise is random neural activity
- If you add the same tiny signal each time, the signal will gradually get bigger. Because the noise will be equally often above and below zero, it should cancel itself out
- The signal to noise ratio grows as the number of sweeps increases
- The signal waveform is always the same across sweeps, so at each time point, the signal component of the voltage slowly increases relative to the noise component
- Noise component is random: At each point in time, over a large number of sweeps, noise will cancel itself out
7
Q
Explain the sentence, “The signal voltage across the electrode pair at each instant in time is filtered, amplified, digitized, and stored in computer memory.”
A
- Filtering: Selectively passing certain frequencies (low-pass, high-pass, bandpass)
- Amplifying: Increasing the original signal voltage by a factor to render it useful for recording, displaying, or computing
- Digitizing: Converting the voltage measured on a continuous scale to discrete symbols such as numerical digits
- 8 bit scale: 1 part in 256
- 10 bit scale: 1 part in 1024
- 12 bit scale: 1 part in 4096
- A larger analog scale allows for larger fluctuations that can still be digitized
8
Q
What is the Nyquist Criterion?
A
- The frequency of digital sampling must be greater than twice the highest frequency of interest
- E.g. If you want to analyze EEG components at 20 Hz, you must sample above 40 Hz
9
Q
List 4 functions of ERP recordings.
A
- Normal functions of brain pathways
- Cognitive processing during learning about the stimulus
- Hemispheric differences
- Planning and execution of movement
10
Q
Compare single-cell recordings, EEG/ERP, CT scan, MRI / fMRI, and PET in terms of spatial and temporal resolution.
A
- Single cell recordings: Highest resolution, but lowest generalizability; invasive.
- EEG / ERP: High temporal resolution but low spatial resolution
- CT scan: Quick snapshot, but no functional information
MRI / fMRI: High spatial resolution, low temporal resolution - PET: Shows metabolism, but low spatial and temporal resolution
