neural-signal-analysis 0.2.8

A comprehensive Python library for human brain/ cortical organoid/spheroid eeg/ecog/mea data analysis including FFT, Higuchi Fractal Dimension, Transfer Entropy, and more.

Neural Signal Analysis Library

The neural_signal_analysis library is a comprehensive Python package for EEG data analysis. It encompasses a wide range of functionalities including FFT (Fast Fourier Transform), Higuchi Fractal Dimension, Transfer Entropy, Welch's Power Spectral Density, and more, providing a toolkit for researchers and developers working with neural signal data.

Installation

To install neural_signal_analysis, simply run:

pip install neural_signal_analysis

Ensure you have Python 3.7 or later installed.

Modules:

• FFT (Fast Fourier Transform)
• frequency_maximum_power
• higuchi_fractal_dimension
• MFDFA_neural (Multifractal Detrended Fluctuation Analysis)
• phase_space_2d
• phase_space_3d
• spectral_centroids
• spectral_edge_density
• spectral_entropy_signals
• STFTsignal (Short-Time Fourier Transform)
• transfer_entropy_all_signals
• transfer_entropy_Hemispheric
• transfer_entropy_regional
• welchsPSD (Power Spectral Density)

Example usage:

import welchsPSD (no need for "from neural_signal_analysis import welchsPSD")

Contributing

Contributions to neural_signal_analysis are welcome! If you have suggestions for improvements or encounter any issues, please feel free to open an issue or submit a pull request on our github repo for this library (https://github.com/Metaverse-Crowdsource/EEG-Chaos-Kuramoto-Neural-Net)

Collaborations

If you would like to collaborate with us, feel free to join our Discord channel (https://discord.gg/545bMkgf), or check out our projects at (https://github.com/Synthetic-Intelligence-Labs)

License

This project is licensed under the CC BY-SA 4.0 License - see the LICENSE file for details.

---

Fast Fourier Transform (FFT) Module:

Example Usage:

import numpy as np from FFT import compute_fft, plot_psd, process_eeg_data

Example for single-channel data

single_channel_data = np.random.rand(1000) # Example data fs = 250 # Example sampling frequency frequencies, psd = compute_fft(single_channel_data, fs) plot_psd(frequencies, psd, channel_name='Channel 1')

Example for multi-channel data

multi_channel_data = np.random.rand(5, 1000) # 5 channels, example data process_eeg_data(multi_channel_data, fs)

---

Frequency Maximum Power Module:

Example Usage:

import numpy as np from frequency_maximum_power import compute_peak_frequency, plot_frequency_spectrum, process_eeg_data

Example for single-channel data

single_channel_data = np.random.rand(1000) # Example data fs = 250 # Example sampling frequency peak_frequency = compute_peak_frequency(single_channel_data, fs) plot_frequency_spectrum(single_channel_data, fs, channel_name='Channel 1')

Example for multi-channel data

multi_channel_data = np.random.rand(5, 1000) # 5 channels, example data process_eeg_data(multi_channel_data, fs)

---

Higuchi Fractal Dimension Module:

Example Usage:

import numpy as np from higuchi_fractal import higuchi_fd, higuchi_fd_multichannel

For single-channel data

single_channel_data = np.random.rand(1000) hfd_single = higuchi_fd(single_channel_data, k_max=10)

For multi-channel data

multi_channel_data = np.random.rand(1000, 3) hfd_multi = higuchi_fd_multichannel(multi_channel_data, k_max=10)

---

Multifractal Detrended Fluctuation Analysis (MFDFA) Module:

Example Usage:

import numpy as np from mfdfa_neural import calculate_mfdfa, plot_mfdfa_results

Sample EEG data (5 channels x 1000 data points)

eeg_data = np.random.rand(5, 1000) lags = np.linspace(1, 100, 100) qs = np.linspace(-5, 5, 100) channel_names = ['Frontal', 'Parietal', 'Temporal', 'Occipital', 'Central']

---

Phase Space Analysis Module:

Example Usage:

import numpy as np from phase_space import process_phase_space_analysis

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) emb_dim = 3 max_delay = 20

Process phase space analysis

phase_space_results = process_phase_space_analysis(eeg_data, emb_dim, max_delay)

Assuming phase_space_results contains the embedded data for 3 channels

titles = ['Channel 1', 'Channel 2', 'Channel 3'] output_dir = '/path/to/save/plots' create_phase_space_plot(phase_space_results, titles, output_dir)

---

3D Phase Space Analysis Module:

Example Usage:

import numpy as np from phase_space_3d import delay_embedding, create_3d_phase_space_plots

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) emb_dim = 3 delay = 20

---

Spectral Centroids Analysis Module:

Example Usage:

import numpy as np from spectral_centroids import process_eeg_data

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) fs = 250 # Sampling frequency in Hz

Process EEG data and plot spectral centroids

process_eeg_data(eeg_data, fs)

---

Spectral Edge Density Analysis Module:

Example Usage:

import numpy as np from spectral_edge_density import process_eeg_data

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) fs = 250 # Sampling frequency in Hz percentage = 90 # Percentage threshold for spectral edge

Process EEG data and plot spectral edge frequencies

process_eeg_data(eeg_data, fs, percentage)

---

Spectral Entropy Analysis Module:

Example Usage:

import numpy as np from spectral_entropy_signals import process_eeg_data

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) fs = 250 # Sampling frequency in Hz nperseg = 128 # Segment length for Welch's method

Process EEG data and plot spectral entropy

process_eeg_data(eeg_data, fs, nperseg)

---

Short-Time Fourier Transform (STFT) Module:

Example Usage:

import numpy as np from STFT_signal import process_eeg_data

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) fs = 250 # Sampling frequency in Hz window_size = 128 # Segment length for STFT in samples

Process EEG data and plot STFT

process_eeg_data(eeg_data, fs, window_size)

---

Transfer Entropy Regional Analysis Module:

Example Usage:

import numpy as np from transfer_entropy_regional import process_granular_eeg_data

Sample EEG data (multiple channels)

eeg_data = np.random.rand(5, 1000) # Example with 5 channels eeg_channels = ['Ch1', 'Ch2', 'Ch3', 'Ch4', 'Ch5'] channel_groups = {"Group1": ['Ch1', 'Ch2'], "Group2": ['Ch3', 'Ch4']}

Compute transfer entropy

te_results = process_granular_eeg_data(eeg_data, fs=1000, num_bins=10, k=1, l=1, eeg_channels=eeg_channels, channel_groups=channel_groups) print(te_results)

---

Transfer Entropy Hemispheric Analysis Module:

Example Usage:

import numpy as np from transfer_entropy_hemispheric import process_eeg_data

Sample EEG data (multiple channels)

eeg_data = np.random.rand(10, 1000) # Example with 10 channels eeg_channels = ['Ch1', 'Ch2', ..., 'Ch10'] left_channels = ['Ch1', 'Ch2', 'Ch3', 'Ch4', 'Ch5'] right_channels = ['Ch6', 'Ch7', 'Ch8', 'Ch9', 'Ch10']

Compute transfer entropy between hemispheres

te_left_to_right, te_right_to_left = process_eeg_data(eeg_data, fs=1000, num_bins=10, k=1, l=1, eeg_channels=eeg_channels, left_channels=left_channels, right_channels=right_channels) print(f"Transfer Entropy from Left to Right: {te_left_to_right}") print(f"Transfer Entropy from Right to Left: {te_right_to_left}")

---

Transfer Entropy Regional Analysis Module:

Example Usage:

import numpy as np from transfer_entropy_regional import process_eeg_data

Example EEG data with multiple channels

eeg_data = np.random.rand(10, 1000) # Assume 10 channels eeg_channels = ['Channel 1', 'Channel 2', ..., 'Channel 10'] regions = { "Frontal": ['Channel 1', 'Channel 2'], "Temporal": ['Channel 3', 'Channel 4'] # Add other regions as needed }

Compute transfer entropy between regions

te_results = process_eeg_data(eeg_data, fs=1000, num_bins=10, k=1, l=1, eeg_channels=eeg_channels, regions=regions) print(te_results)

---

Welch's Power Spectral Density Analysis Module:

Example Usage:

import numpy as np from welchsPSD import process_eeg_data

Sample EEG data (3 channels x 1000 data points)

eeg_data = np.random.rand(3, 1000) fs = 250 # Sampling frequency in Hz

Process EEG data and plot Power Spectral Density

process_eeg_data(eeg_data, fs)