iEEG Spectrum Visualization Application

Welcome to the iEEG Spectrum Visualization Application with Wavelet Denoising. This specialized tool offers an intuitive interface to visualize intracranial EEG data, both in the time and frequency domains. The application allows you to explore the raw signal and frequency spectrum of 16 channels, providing valuable insights into neural activities.

Additionally, the application provides a wavelet denoising feature to reduce noise in the EEG data, improving the clarity and accuracy of the visualizations.

Features

• File Scroll Bar: A horizontal scroll bar allowing users to traverse through the EEG data file seamlessly.
• Channel Selector: Select from 16 channels. Channels 0-7 correspond to the left hemisphere EEG, while channels 8-15 pertain to the right hemisphere EEG.
• Window Size Selector: Enables users to zoom in and out on the EEG data. The range varies from 1 to 60 seconds, with a default of 10 seconds.
• Detrend Toggle: Removes DC using a linear wavelet. This feature is crucial for ensuring clear, uninterrupted EEG data visualization and is turned on by default.
• Filter Order Selector: Choose from 1 to 4 to adjust the sharpness of the band filters. By default, the setting is at 2.
• Whole Spectrum Toggle: This overlays the non-filtered spectrum, highlighting frequencies below 0.5 Hz and above 30 Hz.
• Frequency Scale Selector: Adjust between Linear and Log scales for a more comprehensive view of the EEG data in the frequency domain.
• Wavelet Denoise: Apply wavelet denoising to the EEG data. Choose from various wavelet types and levels to refine the signal.

Output Visualizations

• Raw Signal Chart (Time Domain): Witness the EEG signals in their raw, unfiltered form.
• Spectrum Chart (Frequency Domain): Dive deep into the filtered EEG signals, categorized under Delta (0.5-4 Hz), Theta (4-8 Hz), Alpha (8-13 Hz), and Beta (13-30 Hz) bands. If the 'Whole Spectrum' feature is toggled on, additional insights into frequencies below 0.5 Hz and above 30 Hz will be provided.

Key Insights

While raw signals in the time domain might appear indistinguishable, processing and filtering them can bring out distinctive features. The differences become especially pronounced in the frequency domain, where variations in EEG segments become evident. Moreover, by utilizing a simple metric such as the band energy ratio, users can quickly determine the EEG state. Wavelet denoising can further enhance the signal quality and provide a clearer representation of the underlying neural activities.

Technical Details

The application employs Butterworth bandpass filters, set at a default 2nd order. For spectral analysis, the FFT (Fast Fourier Transform) is used in conjunction with a Hanning window. RMS averaging, exponential weighting, and a set number of averages (10) further enhance the analysis. Wavelet denoising is performed using various wavelet types and levels to refine the EEG signal and reduce noise.

Data Source

This application utilizes intracranial EEG data from NeuroVista, which is publicly available at www.ieeg.org.

Closing Note

By transforming and categorizing the raw EEG data, this application offers a profound understanding of the underlying neural activities. Whether you're a researcher, student, or just curious, the tool will provide invaluable insights into the intricacies of EEG recordings. The wavelet denoising feature further refines the data, enhancing the clarity and accuracy of the visualizations.