FAST functional connectivity implicates P300 connectivity in working memory deficits in Alzheimer's disease.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Transient functional connectivity in electroencephalogram (EEG) research is challenging due to noise and spurious correlations.
• High-temporal resolution EEG data offers valuable insights but is often confounded by signal quality.
• Existing methods struggle to accurately discriminate between conditions in noisy EEG signals.

Purpose of the Study:

• To introduce a novel methodology, Filter Average Short-term (FAST) functional connectivity, to overcome limitations in measuring transient EEG functional connectivity.
• To enhance the analysis of high-temporal resolution brain activity by filtering noise and spurious correlations.
• To apply the FAST method to identify differences in visual short-term memory (VSTM) binding deficits in Alzheimer's disease (AD) cohorts.

Main Methods:

• Averaging long-term functional connectivity across a cohort to create a stable connectivity matrix for a given task.
• Utilizing the average connectivity matrix as a filter to analyze individual subjects' transient functional connectivity.
• Simulating noisy event-related potentials (ERPs) to validate the FAST method's discriminative power compared to standard approaches.

Main Results:

• Simulations demonstrated that FAST accurately discriminates differences in noisy ERPs where other methods failed.
• Application to Alzheimer's disease (AD)-related mild cognitive impairment (MCI) cohorts revealed significant differences in VSTM binding tasks.
• No significant differences were observed in the shape task, but reproducible differences were found in the P300 ERP range for the binding task.

Conclusions:

• The FAST functional connectivity method provides sensitive measurements of transient brain activity.
• This technique can effectively filter noise and spurious correlations in EEG data.
• FAST holds potential for obtaining clinically significant results, particularly in neurodegenerative disease research.