Example: Time-Frequency Analysis (MNE-Python)¶

This page explains the time_frequency_analysis_pipeline_mne.signalJourney.json example file, which documents a typical time-frequency analysis using MNE-Python.

Pipeline Overview¶

This MNE-Python pipeline demonstrates time-frequency analysis using Morlet wavelets: - Load cleaned data from ICA decomposition pipeline - Extract event-related epochs - Compute time-frequency decomposition using mne.time_frequency.tfr_morlet - Apply baseline correction and save results - Generate visualization plots

Pipeline Flowchart¶

flowchart TD
    A[Load Cleaned Data
mne.io.read_raw_fif] --> B[Extract Epochs
mne.Epochs]
    B --> C[Compute TFR
mne.time_frequency.tfr_morlet]
    C --> D[Apply Baseline Correction
tfr.apply_baseline]
    D --> E[Save TFR Results
tfr.save]

    %% Input file
    F["📁 sub-01_task-rest_desc-cleaned_eeg.fif
From: ICA Decomposition Pipeline"] --> A

    %% Inline data
    C --> V2["📊 Frequencies
[4, 8, 13, 30] Hz"]

    %% Final outputs
    E --> G["💾 sub-01_task-rest_desc-tfr_eeg.h5
Time-frequency results"]
    D --> H["💾 sub-01_task-rest_desc-tfr_plot.png
Visualization"]

    %% Quality metrics
    C --> Q1["📈 Frequency range: 4-30 Hz
Cycles: 2-15"]
    D --> Q2["📈 Baseline: [-0.2, 0] s
Mode: percent"]

    %% Styling
    classDef processStep fill:#e1f5fe,stroke:#01579b,stroke-width:2px
    classDef inputFile fill:#fff3e0,stroke:#e65100,stroke-width:2px
    classDef outputFile fill:#e8f5e8,stroke:#1b5e20,stroke-width:2px
    classDef inlineData fill:#f3e5f5,stroke:#4a148c,stroke-width:1px
    classDef qualityMetric fill:#f9f9f9,stroke:#666,stroke-width:1px

    class A,B,C,D,E processStep
    class F inputFile
    class G,H outputFile
    class V2 inlineData
    class Q1,Q2 qualityMetric

Key MNE-Python Features Demonstrated¶

Time-Frequency Functions¶

• mne.Epochs: Event-related epoch extraction
• mne.time_frequency.tfr_morlet: Morlet wavelet decomposition
• tfr.apply_baseline: Baseline correction for power changes
• tfr.save: Save time-frequency results in HDF5 format

Advanced Parameters¶

• Wavelet parameters: Adaptive cycles for different frequencies
• Baseline correction: Percent change from pre-stimulus period
• Frequency selection: Logarithmically spaced frequencies
• Output formats: Both HDF5 data and PNG visualization

Example JSON Structure¶

The time-frequency computation step demonstrates complex parameter documentation:

{
  "stepId": "3",
  "name": "Compute Time-Frequency Decomposition",
  "description": "Calculate time-frequency representation using Morlet wavelets.",
  "software": {
    "name": "MNE-Python",
    "version": "1.6.1",
    "functionCall": "mne.time_frequency.tfr_morlet(epochs, freqs=freqs, n_cycles=n_cycles, return_itc=False)"
  },
  "parameters": {
    "freqs": [4, 6, 8, 10, 13, 17, 22, 30],
    "n_cycles": [2, 3, 4, 5, 6.5, 8.5, 11, 15],
    "use_fft": true,
    "return_itc": false,
    "decim": 1,
    "n_jobs": 1
  }
}

Multi-Output Documentation¶

Steps can produce multiple related outputs:

"outputTargets": [
  {
    "targetType": "file",
    "location": "./derivatives/signaljourney/sub-01/eeg/sub-01_task-rest_desc-tfr_eeg.h5",
    "format": "HDF5",
    "description": "Time-frequency power data."
  },
  {
    "targetType": "file", 
    "location": "./derivatives/signaljourney/sub-01/eeg/sub-01_task-rest_desc-tfr_plot.png",
    "format": "PNG",
    "description": "Time-frequency plot visualization."
  }
]

Time-Frequency Analysis Features¶

Wavelet Parameter Selection¶

• Frequency-dependent cycles: Lower frequencies use fewer cycles for better temporal resolution
• Higher frequencies: More cycles for better frequency resolution
• Adaptive approach: Balances time-frequency trade-off across spectrum

Baseline Correction Methods¶

• Percent change: (power - baseline) / baseline * 100
• Ratio: power / baseline
• Decibel: 10 * log10(power / baseline)
• Z-score: (power - baseline_mean) / baseline_std

MNE-Python vs EEGLAB Comparison¶

Usage Notes¶

This example demonstrates: - Time-frequency decomposition with optimal parameters - Baseline correction for interpretable results - Multi-format outputs for analysis and visualization - Pipeline integration building on previous processing steps - Quality documentation for reproducible analysis

The pipeline showcases MNE-Python's comprehensive time-frequency analysis capabilities while maintaining full parameter transparency for reproducibility.