ERP components are defined by following approaches
- Source localization
- Principal component analysis (PCA)
- Independent component analysis (ICA)
- Time frequency analysis
In source localization techniques, component is equivalent to source localization. It find the set of single component scalp distribution that can be summated to give best fit to observed scalp distribution over time.
PCA and ICA make no biophysical assumptions, but instead use statistical properties of data to derive the scalp distribution of components. PCA is designed to find an un-mixing matrix in which each component has it scalp distribution. ICA designed to find un-mixing matrix that maximizes independence of each component so that every individual component represents the largest possible amount of information.
Fig1: ERP waveform components
ERP are often defined in terms of polarity, scalp distribution and latency. For example visual sensory components are more dependent on latency. Polarity and timing information cannot identify a component, but scalp distribution can help. Each subcomponent is actually a total component as a whole, which reflects the functionally and anatomically defined process. Therefore it is important to intelligently combine all the factors that might lead to ERP component and make a converging evidence of component source origin.
Fig2: Observed waveform vs underlying components
Fig3: Original wave form and new waveform with C2 amplitude increased (latency of C2).Increased amplitude has decreased the amplitude of peak 2(c). Original wave form and new form with C2 amplitude decreased. Decreased amplitude had increased the amplitude of peak 2(d).
ERP has been used in scientific community in the following domain
- Determining which cognitive or neural process differ across conditions across age groups
- Determining brain has completed set of processes
- Uncovering new mental process and subdividing know processes.
- Monitoring of processing in situations in which typical behavioral dis-orders cannot be studied.
Methods for isolating ERP component
- To focus on specific component with known amplitude, that is usually larger than the rest of components and can be easily identified.
- For example P300 is greater than 10 micro volts.
- To focus on a task, that allows only one or two ERP components to vary across different conditions.
- We focused on less number of components; it is easy to avoid overlap of components, and making measurements easier.
- Subtracting overlying ERP components by creating difference waves between conditions.
- LRP (lateral readiness potential) is difference wave created by subtracting waves of response with non-responsive wave.
Note: Fig2&3 are taken from The Oxford Handbook of Event-Related Potential Components. I dont own copyright of pictures.