Physiological Confounds in BOLD-fMRI and Their Correction.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Imaging

Background:

• Functional magnetic resonance imaging (fMRI) is crucial for understanding brain function.
• Physiological confounds significantly impact fMRI data quality, particularly at higher magnetic fields.
• Key confounds include respiratory and cardiac fluctuations, thoracic movements, and cardiac pulsatility.

Purpose of the Study:

• To review the mechanics, performance, and limitations of physiological confound correction techniques in fMRI.
• To discuss the impact of these correction methods on fMRI study outcomes.
• To highlight challenges and future directions in addressing physiological noise in fMRI.

Main Methods:

• Review of existing literature on physiological confound correction in fMRI.
• Analysis of four major physiological confounds: breathing, heart rate, thoracic movement, and cardiac pulsatility.
• Evaluation of correction techniques including model-based approaches, independent component analysis, and machine learning.

Main Results:

• Correction methods have improved the detection of task-activated voxels and reduced false positives/negatives in functional connectivity.
• Model-based methods require external physiological data, which is often unavailable.
• Independent component analysis needs prior component number knowledge, and machine learning methods require further validation.

Conclusions:

• Physiological confound correction is vital for robust fMRI research, especially with advanced imaging techniques.
• Current correction methods have limitations, necessitating ongoing development and validation.
• Future research should focus on refining existing techniques and exploring novel approaches for comprehensive physiological noise removal in fMRI.