Exploring Cortical Responses to Blood Flow Restriction through Deep Learning.

Area of Science:

• Neuroscience
• Exercise Physiology
• Biomedical Engineering

Background:

• Blood flow restriction (BFR) training enhances muscle hypertrophy and strength.
• The neural mechanisms underlying BFR's effects are not well understood.
• Brain-computer interfaces (BCIs) offer a method to investigate cortical activity.

Purpose of the Study:

• To explore the impact of BFR on cortical activity.
• To investigate the feasibility of using deep learning (DL)-based BCIs for analyzing BFR-induced neural changes.
• To assess the generalizability of BFR-related cortical responses across subjects and time.

Main Methods:

• Magnetoencephalography (MEG) was used to record cortical responses in six subjects under three conditions: before, during, and after BFR.
• Data preprocessing included standardization and Euclidean-space alignment.
• A deep learning model (BaseNet) was employed to classify MEG data, with testing across within-subject, cross-subject, and cross-time splits.

Main Results:

• Within-subject classification accuracy exceeded 90%, indicating detectable individual cortical responses to BFR.
• Cross-subject models performed at chance level (33%), revealing significant inter-individual variability.
• Cross-time models achieved accuracy above 50%, suggesting some temporal consistency in responses.

Conclusions:

• Cortical activity patterns associated with BFR are detectable at an individual level using DL-based BCIs.
• Significant inter-individual variability in BFR responses poses challenges for generalizing findings.
• Further research is needed to understand and potentially overcome the individual specificity of BFR-induced neural changes.