Pypreclin

Pypreclin preprocesses macaque functional MRI (fMRI) data to correct experimental and technical artifacts and register images to a macaque brain template for downstream analyses.

Key Features:

• Implementation: Python-based module for automated preprocessing of macaque fMRI datasets.
• Automated pipeline: Fully automated pipeline that processes raw fMRI images using state-of-the-art algorithms.
• Artifact handling: Addresses artifacts arising from body movements and intracranial leads.
• Field strength compatibility: Demonstrated compatibility with data acquired at 3T and at ultra-high fields.
• Subject state compatibility: Handles data from both awake and anesthetized macaques.
• Contrast agent support: Accommodates data acquired with and without iron oxide contrast agents.
• Coil support: Supports data acquired with single loop coils and multichannel phased-array coils.
• Implanted hardware: Accommodates scenarios involving intracranial implanted electrodes.
• Registration: Registers fMRI data to a macaque brain template across tested conditions.
• Validation dataset use: Validated on datasets from the Primate Neuroimaging Data-Exchange across institutions and protocols.
• Task and resting analyses: Evaluated on resting-state and auditory evoked fMRI studies, including statistical localization of activations.

Scientific Applications:

• Resting-state macaque fMRI preprocessing: Preprocessing for resting-state connectivity analyses across different field strengths, coils, and subject states.
• Auditory evoked fMRI analysis: Preprocessing and statistical localization of auditory evoked activations relative to gray matter structures at corrected statistical levels.
• Cross-site and multi-protocol studies: Harmonization and preprocessing of macaque fMRI data from multiple institutions and protocols via validation on the Primate Neuroimaging Data-Exchange.
• Studies with implanted hardware or contrast agents: Preprocessing for datasets acquired with intracranial electrodes or iron oxide contrast agents.

Methodology:

Processes raw fMRI images via a fully automated pipeline using state-of-the-art algorithms, addresses artifacts from body movements and intracranial leads, registers data to a macaque brain template, and performs statistical localization of activations at corrected significance levels.