Impact of physiological noise in characterizing the functional MRI default-mode network in Alzheimer's disease

Area of Science:

• Neuroimaging
• Physiological Monitoring
• Alzheimer's Disease Research

Background:

• Functional connectivity of the default-mode network (DMN) is often reduced in Alzheimer's disease (AD) patients compared to healthy controls.
• Physiological processes like breathing and heart beating can introduce noise into functional magnetic resonance imaging (fMRI) signals.
• This physiological noise may differentially affect fMRI data between AD patients and healthy participants, potentially confounding DMN connectivity findings.

Purpose of the Study:

• To investigate the impact of cardiac and respiratory physiological noise on DMN functional connectivity in Alzheimer's disease (AD).
• To assess whether correcting for physiological noise can resolve observed differences in DMN connectivity between AD patients and healthy controls.

Main Methods:

• Collected resting-state fMRI data concurrently with cardiac and respiratory recordings from AD patients and age-matched healthy participants.
• Estimated physiological responses phase-locked and non-phase-locked to cardiac and respiratory cycles.
• Corrected resting-state fMRI data for physiological noise using population-specific physiological response functions.

Main Results:

• Physiological responses in AD patients were significantly delayed compared to healthy participants (cardiac: 3.00 ± 0.51 s; respiratory: 3.96 ± 0.52 s; both p < 0.0001).
• After physiological noise correction, the default-mode network (DMN) estimated by seed-correlation became more localized.
• The previously observed significant differences in DMN connectivity between AD patients and healthy controls became insignificant post-correction.

Conclusions:

• Physiological noise significantly influences resting-state fMRI measurements of DMN connectivity in Alzheimer's disease.
• Controlling for physiological noise is crucial for accurate characterization of DMN alterations in AD.
• This highlights the importance of incorporating physiological noise correction in fMRI analyses for clinically relevant findings in neurodegenerative diseases.