Purpose: Decompose the signal into intrinsic mode functions.
Method: EMD (Empirical Mode Decomposition) with a maximum of 5 modes.
How it works:
- The signal is decomposed into intrinsic modes using the EMD algorithm.
- The drop-down list contains 6 values:
- IMF: no decomposition is applied; the signal remains unchanged
- 1: mode 1
- 2: mode 2
- 3: mode 3
- 4: mode 4
- 5: mode 5
- If a specific mode is selected, only that mode is used:
signal = imf(:, selected_mode)
Use cases:
- Isolating specific frequency components
- Removing baseline drift
- Analyzing nonlinear and non-stationary processes
Physiological explanation of EMD:
Empirical Mode Decomposition is a method for decomposing a signal into “intrinsic modes” without using predefined basis functions (unlike the Fourier transform).
What modes are:
- Modes are oscillatory signal components that have physical meaning.
- Each mode has its own frequency and amplitude.
- Modes adapt to the data (they are not fixed frequencies as in Fourier analysis).
Physiological meaning:
- Mode 1 (high-frequency): the fastest components, possible noise
- Modes 2–3: main rhythms (alpha, beta)
- Mode 4: slower components, signal details
- Mode 5 (low-frequency): slow changes, baseline drift, and possibly artifacts
Why use a specific mode:
- Drift removal: Mode 5 often contains baseline drift (due to sweating or electrode movement).
- Rhythm isolation: A specific mode may contain the rhythm of interest (for example, alpha).
- Artifact removal: Some modes may contain artifacts that can be excluded.
Advantages of EMD:
- Adaptivity: modes fit the signal
- Nonlinearity: nonlinear processes can be analyzed
- Non-stationarity: works with signals whose properties change over time