logo
logo

PyLeida

Leading Eigenvector Dynamics Analysis (LEiDA) in Python logo

Contents

pyleida is a Python toolbox to apply the Leading Eigenvector Dynamics Analysis (LEiDA) framework to functional Magnetic Resonance Imaging (MRI) data. It contains all the necessary tools to apply the framework from the beggining to the end of the pipeline or workflow, save results, generate reports, and figures.

About LEiDA

LEiDA is a powerful tool to characterize the temporal evolution of dynamic functional connectivity (dFC) based on fMRI phase-coherence connectivity with reduced dimensionality, which becomes useful in the analysis of multi-dimensional dFC data. Importantly, by focusing solely on its dominant connectivity pattern instead of the whole upper triangular part of the phase-coherence matrices, LEiDA is more robust to high-frequency noise, overcoming a limitation affecting all quasi-instantaneous measures of functional connectivity. Indeed, it allows detecting the precise epochs when the variance of the dFC becomes dominated by a different pattern, even if the dFC evolves more smoothly. Moreover, beyond significantly reducing the dimensionality of the data and allowing for improved temporal resolution, LEiDA offers the advantage that recurrences of the same pattern are more clearly detected, hence improving the signal-to-noise ratio in the functional connectivity dynamics (DFC) analysis.

The LEiDA framework has demonstrated the ability to detect quantifiable features (serving as potential biomarkers) of clinically-relevant brain functional subsystems (see Publications).

Reference: Cabral et al. (2017). Cognitive performance in healthy older adults relates to spontaneous switching between states of functional connectivity during rest. Scientific reports, 7(1), 5135. https://doi.org/10.1038/s41598-017-05425-7

Notes

logoSource code can be found online at the site of the repository: https://github.com/PSYCHOMARK/leida-python

This package was created by PSYCHOMARK, an open-source software project for neuroimaging analysis to serve in Psychopathology and Psychophysiology research, with applications for diagnosis, drug development and theranostics.

See PSYCHOMARK contributors here.

Main classes

pyleida.Leida

Class to execute the LEiDA pipeline, explore the results, generate reports, and figures.

pyleida.DataLoader

Class to retrieve and explore the LEiDA results.

Modules

pyleida.clustering

The module 'pyleida.clustering' provides a class and functions to identify and explore the BOLD phase-locking states or patterns using K-Means clustering

pyleida.data_utils

The module 'pyleida.data_utils' provides generic functions to manipulate/handle data, and load the neccesary files to run the 'Leida' class

pyleida.dynamics_metrics

The module 'pyleida.dynamics_metrics' provides functions to compute the metrics from dynamical systems theory

pyleida.plotting

The module 'pyleida.plotting' provides functions to generate plots and visual representations.

pyleida.signal_tools

The module 'pyleida.signal_tools' provides functions to compute relevant information from BOLD time series.

pyleida.stats

The module 'pyleida.stats' provides functions to execute statistical analyses on the dynamical system theory metrics.