Comparison of heart rate and heart rate variability metrics derived from laboratory ECG signals and BOLD fMRI

Motivation: Ensuring that HR and HRV metrics obtained from different signals and testing environments are consistent is crucial for combining cardiovascular data from multiple sources and enhancing the robustness of studies. Goal(s): We evaluated the consistency of HR and HRV (SDNN and RNSSD) measurements obtained in the laboratory and the MRI scanner. Approach: Mean-HR, SDNN, and RMSSD calculated from inter-beat intervals derived from ECG R-R interval recorded in the laboratory, and from the peaks of cardiac cycles directly derived from rs-fMRI data itself, without using scanner's built-in PPG signal were compared. Results: Correlations were high for mean-HR, moderate for SDNN and low for RMSSD. Impact: Using MRI data alone to assess HR enables more comprehensive, integrated research on cardiac and brain interactions. It may contribute to enhancing multimodal data analysis and developing robust, non-invasive protocols for studying cardiac and brain interactions in various applications.