1 Definition 2 Functional Requirements 3 Syntax 4 Semantics

1 Definition

The EEG Processing Type defines the allowable operations, methods, and processing descriptors used for Electroencephalography (EEG) data by reusing Common Definitions for: Header, Algorithm, Algorithms, FeatureClass, Features.

2 Functional Requirements

EEG Processing Type

The EEG Processing Type shall:

  • Fix Domain = EEG.
  • Validate Operation against EEG‑specific enumerations.
  • Validate Method against EEG domain signal‑processing and ML techniques.
  • Allow Algorithm to be either a string identifier or an AlgorithmObject.
  • Allow Algorithms to be an array of Algorithm items.
  • Support FeatureClass to specify the category of extracted features.
  • Require Features to be a non‑empty array of unique strings.

Operations

The EEG Processing Type shall support operations:

  • BandPower
  • SpectralAnalysis
  • EventRelatedPotential
  • ArtifactRemoval
  • ICADecomposition
  • SourceLocalization
  • SleepStageClassification
  • Epoching
  • ConnectivityAnalysis
  • TimeFrequencyAnalysis

These operations cover spectral, temporal, Event‑Related Potential (ERP)‑related, connectivity, artefact processing, and state classification tasks in EEG workflows.

Method Validation

The EEG Processing Type shall validate Method against:

  • FFT
  • WelchPSD
  • MorletWavelet
  • ICA
  • CSP
  • sLORETA
  • Beamforming
  • NotchFilter
  • BandpassFilter
  • DCNN
  • LSTM

These methods include classical spectral tools, decomposition techniques, source modelling, filtering, and deep‑learning‑based approaches.

3 Syntax

https://schemas.mpai.community/AIH1/V1.0/data/EEGProcessingType.json

4 Semantics

Label Description
Header EEG Processing Type Header, Standard “AIH-EET-Vx.y”
Domain Constant value "EEG". Identifies that this Processing Type applies exclusively to EEG data. No other value is permitted.
Operation Specifies the EEG‑specific processing step to be performed. Enumerated list includes: BandPower, SpectralAnalysis, EventRelatedPotential, ArtifactRemoval, ICADecomposition, SourceLocalization, SleepStageClassification, Epoching, ConnectivityAnalysis, TimeFrequencyAnalysis.
BandPower Operation measuring energy in canonical or custom EEG frequency bands.
SpectralAnalysis Operation performing decomposition of EEG into frequency spectrum components (power spectral density, etc.).
EventRelatedPotential Operation computing event‑locked temporal responses (ERP components).
ArtifactRemoval Operation removing blinking, muscle artefacts, line noise, movement artefacts, etc.
ICADecomposition Operation applying ICA to decompose EEG signals into independent components.
SourceLocalization Operation estimating brain source activity generating the scalp EEG signals.
SleepStageClassification Operation classifying sleep stages (N1, N2, N3, REM, Wake).
Epoching Operation segmenting continuous EEG into time intervals aligned to events or time windows.
ConnectivityAnalysis Operation computing functional connectivity (coherence, PLV, PLI).
TimeFrequencyAnalysis Operation analyzing how frequency content evolves over time (wavelets, multitaper).
Method Processing technique used to implement the operation. Must be one of: FFT, WelchPSD, MorletWavelet, ICA, CSP, sLORETA, Beamforming, NotchFilter, BandpassFilter, DCNN, LSTM.
Algorithm Either a string identifier or an AlgorithmObject from CommonDefinitions. Represents the algorithm used within the method/operation.
AlgorithmObject.Name Required name of the algorithm (e.g., “FFT‑Welch”, “DeepConvNet”).
AlgorithmObject.Version Optional version identifier for the algorithm.
AlgorithmObject.Params Free‑form object containing configuration parameters for the algorithm.
Algorithms An array of Algorithm entries, each either a string ID or an AlgorithmObject.
FeatureClass Category describing the type of EEG features (e.g., spectral bands, ERP components, wavelets, connectivity measures). Defined in CommonDefinitions.
Features A non‑empty array of unique strings specifying individual EEG feature names (e.g., alpha_power, N100_amplitude, PLV_theta). Defined in CommonDefinitions.
Trace Provenance information and Time.
DescrMetadata Descriptive Metadata