Novel Application of Multiscale Cross-Approximate Entropy for Assessing Early Changes in the Complexity between Systolic Blood Pressure and ECG R-R Intervals in Diabetic Rats

Abstract

Cardiac autonomic neuropathy (CAN) is a common complication of diabetes mellitus, and can be assessed using heart rate variability (HRV) and the correlations between systolic blood pressure (SBP) and ECG R-R intervals (RRIs), namely baroreflex sensitivity (BRS). In this study, we propose a novel parameter for the nonlinear association between SBP and RRIs based on multiscale cross-approximate entropy (MS-CXApEn). Sixteen male adult Wistar Kyoto rats were equally divided into two groups: streptozotocin-induced diabetes and age-matched controls. RRIs and SBP were acquired in control rats and the diabetic rats at the onset of hyperglycemia and insulin-treated euglycemia to determine HRV by the ratio of low-frequency to high-frequency power (LF/HF) and Poincaré plot as SSR (SD1/SD2), BRS, and MS-CXApEn. SSR and BRS were not significantly different among the three groups. The LF/HF was significantly higher in the hyperglycemic diabetics than those in the controls and euglycemic diabetic rats. MS-CXApEn was higher in the diabetic hyperglycemic rats than the control rats from scales 2 to 10, and approached the values of controls in diabetic euglycemic rats at scales 9 and 10. Conclusions: We propose MS-CXApEn as a novel parameter to quantify the dynamic nonlinear interactions between SBP and RRIs that reveals more apparent changes in early diabetic rats. Furthermore, changes in this parameter were related to correction of hyperglycemia and could be useful for detecting and assessing CAN in early diabetes.

Evaluating the Effectiveness of Complexity Features of Eye Movement on Computer Activities Detection

“…The basic nature of complexity-based features that are containing “implicit” information about body responses to the human activity and states [ 46 , 47 ] has benefited AI models to help them to overcome the prediction confusedness (compare Table 4 with Table 5 ). As was expected, the findings in this study prove that complexity analysis is also suitable for eye-movement-based data, as useful as its usage in human heart rate [ 16 ], cerebral hemodynamics [ 17 ], blood pressure [ 18 ], and body movements [ 19 , 48 ] data. Moreover, the experimental procedure that resulted in moderate head movement also confirmed that AI models would be suitable for everyday use in distinguishing computer activities in daily life.…”

“…Meanwhile, complexity analysis [ 11 ] was recently used in certain human biometric data comprising heart rate [ 16 ], cerebral hemodynamics [ 17 ], blood pressure [ 18 ], and infants’ limb movements [ 19 ]. The use of the complexity of these biological data can describe the human states related to their health conditions [ 16 , 17 , 18 ] and activities [ 19 , 20 ]. The benefit is the potential application to eye-movement features.…”