TRACC-PHYSIO: Time-Domain Resolution-Aligned Cross-Correlation to Estimate PHYSIOlogical Coupling and Time Delays in...

To validate a method to assess cardiac and respiratory brain pulsations in slowly sampled dynamic MR scans. Cardiac and respiratory pulsations are key drivers of neurofluid circulation. However, resolving their temporal dynamics requires fast imaging, which is not achievable in many dynamic MR acquisitions.

We systematically validated TRACC-PHYSIO, a Time-domain Resolution-Aligned Cross-Correlation framework for estimating physiological coupling and time delays in dynamic MRI. This method uses simultaneously recorded cardiac and respiratory waveforms as external references and applies time-shifted cross-correlation to estimate physiological coupling strength and relative pulse delay in dynamic MRI data. Two metrics are derived: Peak Coupling Coefficient (Peak CorrCoeff), which quantifies coupling strength, and TimeDelay, which estimates the relative arrival time of the physiological impulse in the brain. TRACC-PHYSIO was first evaluated with in vivo fMRI (TR = 363 ms) by comparing the Peak CorrCoeff with physiological bandpower derived from spectral analysis. Performance was further assessed using simulations of realistically modeled dynamic MR signals across repetition time (TR = 50-3000 ms) and acquisition durations (60-360 s).

In fMRI, TRACC-derived cardiac and respiratory Peak CorrCoeff were strongly associated with their respective spectrum-derived physiological bandpower (all tests Pearson r > 0.90). In simulations, both Peak CorrCoeff and TimeDelay were estimated with no mean bias and low temporal errors across TRs and acquisition durations.

TRACC-PHYSIO is a validated time-domain framework for quantifying cardiac and respiratory coupling strength and estimating millisecond-scale relative pulse delays in standard dynamic MR acquisitions.