The evaluation of local and global approaches in assessment of baroreflex sensitivity using sequence technique

This paper evaluates the assessment of baroreflex sensitivity (BRS) using the most common time domain method - sequence technique. The BRS estimation is based on a linear regression between systolic blood pressure (SBP) and pulse intervals (PI) values in baroreflex sequences (BS). Since BS beats represent a very small percentage of total beat number (about 10% in rats), there exist doubts about the accuracy of the estimated slopes, especially since BSs are usually three bits long and there are number of outliers. For these reasons, a global estimator with outlier rejection rule is evaluated, with both baroreflex sequences and baroreflex events (BE). The BRS estimation techniques were done in datasets obtained from radiotelemetred rats treated with: saline, beta₁-adrenergic antagonist metoprolol, alpha₁-adrenergic antagonist prazosin and muscarinic receptor antagonist atropine. The results show that, even the number of beats in BE increases up to the 50% of the total, their use slightly improves the distinguishing among the experimental groups. The global estimator, both on BSs and on BEs, shows the best performance in terms of variability.