eddymotion

eddymotion implements an open-source framework for volume-to-volume artifact estimation in neuroimaging that generalizes beyond eddy current distortions of diffusion MRI.

Key Features:

• Gaussian Process Regressor Model: The framework implements a Gaussian Process regressor model for alignment estimation of neuroimaging data.
• Bayesian Optimization Layer: A Bayesian optimization layer is integrated on an existing registration framework to enable alignment estimation.
• Alternative Model Implementation: The tool allows for easy alternative model implementation.
• Multi-modal Applicability: The framework can be used for non-dMRI imaging modalities.

Scientific Applications:

• Generalized Artifact Identification and Correction: The volume-to-volume modeling framework enables generalized artifact identification and correction in neuroimaging beyond diffusion MRI.
• Neuroimaging Data Alignment: The framework estimates the alignment of neuroimaging data enabled by volume-to-volume modeling.

Methodology:

The framework implements a Gaussian Process regressor model and a Bayesian optimization layer on an existing registration framework to estimate the alignment of neuroimaging data.