Instantaneous Lung Volume Research Articles

Estimation of ARMA Model Order via Artificial Neural Network for Modeling Physiological Systems

Model order estimation is the most important but challenging step for system identification using an autoregressive moving average (ARMA) model. In this paper, we propose an artificial neural network (ANN) structure to estimate the best model order for ARMA modeling of linear, time-invariant systems using the system’s input and output data. The proposed algorithm creates an equivalent ANN structure corresponding to an ARMA model and chooses the best model order using the neural network’s mean squared error (MSE) loss function. The proposed method is validated on simulated ARMA model data and the performance is compared with the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). We considered three hypothetical linear systems and performed 100 Monte Carlo simulations for each model, with different data lengths, and with additive noise. For each of the three simulation models, the proposed method significantly outperformed the AIC and BIC in terms of the correct model order selection. Finally, the proposed ANN-based model order estimator was successfully applied to determine the dynamic relationship between heart rate (HR) and instantaneous lung volume (ILV) using an ARMA model. The results indicate that physiological and biological systems can be modeled with appropriate ARMA models obtained by the proposed algorithm to better understand the system dynamics.

CRSIDLab: A Toolbox for Multivariate Autonomic Nervous System Analysis Using Cardiorespiratory Identification.

This paper presents the Cardiorespiratory System Identification Lab (CRSIDLab), a MATLAB-based software tool for multivariate autonomic nervous system (ANS) evaluation through heart rate variability (HRV) analysis and cardiorespiratory system identification. Based on a graphical user interface, CRSIDLab provides a complete set of tools including pre-processing cardiorespiratory data (electrocardiogram, continuous blood pressure, airflow, and instantaneous lung volume), power spectral density estimation, and multivariable cardiorespiratory system model identification. Parametrized multivariate models can assess both HRV and baroreflex sensitivity (BRS) by considering the causal relationship from respiration to heart rate (or its reciprocal, R-to-R interval - RRI) and from systolic blood pressure to RRI, for instance. The impulse response, estimated from the model, is used as a mathematical tool to effectively open the inherently closed-loop nature of the cardiorespiratory system, allowing the investigation of the dynamic response between pairs of cardiorespiratory variables. This system modeling approach provides information on gain and temporal behavior regarding dynamics, such as the baroreflex, complementing traditional HRV, and BRS indices. The toolbox is presented and used to investigate autonomic function in sleep apnea. The results show that, while traditional HRV indices were unable to differentiate between apneic and non-apneic subjects, the autonomic descriptors obtained from the multivariate system identification techniques were able to show vagal impairment in apneic compared to non-apneic subjects. Thus, CRSIDLab can help promote the use of cardiorespiratory system identification as a potentially more sensitive measure of ANS activity than classical HRV analysis.

RR interval–respiratory signal waveform modeling in human slow paced and spontaneous breathing

Our aim was to model the dependence of respiratory sinus arrhythmia (RSA) on the respiratory waveform and to elucidate underlying mechanisms of cardiorespiratory coupling. In 30 subjects, RR interval and respiratory signal were recorded during spontaneous and paced (0.1Hz/0.15Hz) breathing and their relationship was modeled by a first order linear differential equation.This model has two parameters: a0 (related to the instantaneous degree of abdominal expansion) and a1 (referring to the speed of abdominal expansion). Assuming that a0 represents slowly adapting pulmonary stretch receptors (SARs) and a1 SARs in coordination with other stretch receptors and central integrative coupling; then pulmonary stretch receptors relaying the instantaneous lung volume are the major factor determining cardiovagal output during inspiration.The model's results depended on breathing frequency with the least error occurring during slow paced breathing. The role of vagal afferent neurons in cardiorespiratory coupling may relate to neurocardiovascular diseases in which weakened coupling among venous return, arterial pressure, heart rate and respiration produces cardiovagal instability.

Compression on trigger points in the leg muscle increases parasympathetic nervous activity based on heart rate variability

Massotherapy, the therapeutic use of massage, is used to treat various chronic pain syndromes. One type of massotherapy, pressure stimulus applied over trigger points (TPs), is reported to have excellent therapeutic effects. Its effect is possibly mediated through changes in the autonomic nervous system although little research has been conducted to assess autonomic activity during TP compression. We have investigated how compression applied over TPs affects the autonomic nervous system. Six healthy young adult females whose daily working routine was carried out predominantly in a standing position were enrolled in the study cohort. After a day's work, the subjects were asked to rest supine, and electrocardiograms (ECGs), instantaneous lung volume (ILV) and systolic and diastolic blood pressures (SBP, DBP) were measured before and after pressure application over the TPs in those lower limb muscles where the subjects felt muscle fatigue or discomfort. The subjects were also asked to coordinate breathing with the beeping sounds. The therapeutic effects of TP compression were assessed by a subjective fatigue scale. Parasympathetic nervous activity was also assessed by spectral analysis of heart rate (HR) variability. The transfer function from ILV to HR was evaluated using linear analysis. The results indicated that TP compression (1) decreased HR, SBP and DBP, (2) increased parasympathetic activity, (3) increased the gain from ILV to HR, and (4) improved the fatigue scores. These findings suggest that an increase in parasympathetic nervous activity after the TP compression induced a reduction of fatigue. The therapeutic mechanisms of TP compression to enhance parasympathetic nervous system are discussed.

Postural-induced phase shift of respiratory sinus arrhythmia and blood pressure variations: insight from respiratory-phase domain analysis

The purpose of this study is to evaluate the multiple effects of respiration on cardiovascular variability in different postures, by analyzing respiratory sinus arrhythmia (RSA) and respiratory-related blood pressure (BP) variations for systolic BP (SBP), diastolic BP (DBP), and pulse pressure (PP) in the respiratory-phase domain. The measurements were conducted for 420 s on healthy humans in the sitting and standing positions, while the subjects were continuously monitored for heart rate and BP variability and instantaneous lung volume. The waveforms of RSA and respiratory-related BP variations were extracted as a function of the respiratory phase. In the standing position, the waveforms of the BP variations for SBP, DBP, and PP show their maxima at around the end of expiration (pi rad) and the minima at around the end of inspiration (2 pi rad), while the waveform of RSA is delayed by approximately 0.35 pi rad compared with the BP waveforms. On the other hand, in the sitting position, the phase of the DBP waveform (1.69 pi rad) greatly and significantly (P < 0.01) differs from that in the standing position (1.20 pi rad). Also, the phase of PP is delayed and that of RSA is advanced in the sitting position (P < 0.01). In particular, the phase shift of the DBP waveform is sufficiently large to alter whole hemodynamic fluctuations, affecting the amplitudes of SBP and PP variations. We conclude that the postural change associated with an altered autonomic balance affects not only the amplitude of RSA, but also the phases of RSA and BP variations in a complicated manner, and the respiratory-phase domain analysis used in this study is useful for elucidating the dynamic mechanisms of RSA.

Selective quantification of the cardiac sympathetic and parasympathetic nervous systems by multisignal analysis of cardiorespiratory variability

Heart rate (HR) power spectral indexes are limited as measures of the cardiac autonomic nervous systems (CANS) in that they neither offer an effective marker of the beta-sympathetic nervous system (SNS) due to its overlap with the parasympathetic nervous system (PNS) in the low-frequency (LF) band nor afford specific measures of the CANS due to input contributions to HR [e.g., arterial blood pressure (ABP) and instantaneous lung volume (ILV)]. We derived new PNS and SNS indexes by multisignal analysis of cardiorespiratory variability. The basic idea was to identify the autonomically mediated transfer functions relating fluctuations in ILV to HR (ILV-->HR) and fluctuations in ABP to HR (ABP-->HR) so as to eliminate the input contributions to HR and then separate each estimated transfer function in the time domain into PNS and SNS indexes using physiological knowledge. We evaluated these indexes with respect to selective pharmacological autonomic nervous blockade in 14 humans. Our results showed that the PNS index derived from the ABP-->HR transfer function was correctly decreased after vagal and double (vagal + beta-sympathetic) blockade (P < 0.01) and did not change after beta-sympathetic blockade, whereas the SNS index derived from the same transfer function was correctly reduced after beta-sympathetic blockade in the standing posture and double blockade (P < 0.05) and remained the same after vagal blockade. However, this SNS index did not significantly decrease after beta-sympathetic blockade in the supine posture. Overall, these predictions were better than those provided by the traditional high-frequency (HF) power, LF-to-HF ratio, and normalized LF power of HR variability.

Multiscale analysis of short term heart beat interval, arterial blood pressure, and instantaneous lung volume time series

Physiological systems are ruled by mechanisms operating across multiple temporal scales. A recently proposed approach, multiscale entropy analysis, measures the complexity at different time scales and has been successfully applied to long term electrocardiographic recordings. The purpose of this work is to show the applicability of this methodology, rooted on statistical physics ideas, to short term time series of simultaneously acquired samples of heart rate, blood pressure and lung volume, from healthy subjects and from subjects with chronic heart failure. In the same spirit, we also propose a multiscale approach, to evaluate interactions between time series, by performing a multivariate autoregressive (AR) modeling of the coarse grained time series. We apply the multiscale entropy analysis to our data set of short term recordings. Concerning the multiscale version of the multivariate AR approach, we apply it to the four dimensional time series so as to detect scale dependent patterns of interactions between the physiological quantities. Evaluating the complexity of signals at the multiple time scales inherent in physiologic dynamics, we find new quantitative indicators which are statistically correlated with the pathology. Our results show that multiscale entropy calculated on all the measured quantities significantly differs (P<10(-2) and less) in patients and control subjects, and confirms the complexity-loss theory of aging and disease. Also applying the multiscale autoregressive approach significant differences were found between controls and patients; in the sight of finding a possible diagnostic tools, satisfactory results came also from a receiver-operating-characteristic curve analysis (with some values above 0.8). The multiscale entropy analysis can give useful information also when only short term physiological recordings are at disposal, thus enlarging the applicability of the methodology. Also the proposed multiscale version of the multivariate regressive analysis, applied to short term time series, can shed light on patterns of interactions between cardiorespiratory variables.

Respiratory-phase Domain Analysis of Heart Rate Variability Can Accurately Estimate Cardiac Vagal Activity during a Mental Arithmetic Task

The objectives of this paper were to present a method to extract the amplitude of RSA in the respiratory-phase domain, to compare that with subjective or objective indices of the MWL (mental workload), and to compare that with a conventional frequency analysis in terms of its accuracy during a mental arithmetic task. HRV (heart rate variability), ILV (instantaneous lung volume), and motion of the throat were measured under a mental arithmetic experiment and subjective and objective indices were also obtained. The amplitude of RSA was extracted in the respiratory-phase domain, and its correlation with the load level was compared with the results of the frequency domain analysis, which is the standard analysis of the HRV. The subjective and objective indices decreased as the load level increased, showing that the experimental protocol was appropriate. Then, the amplitude of RSA in the respiratory-phase domain also decreased with the increase in the load level. The results of the correlation analysis showed that the respiratory-phase domain analysis has higher negative correlations, -0.84 and -0.82, with the load level as determined by simple correlation and rank correlation, respectively, than does frequency analysis, for which the correlations were found to be -0.54 and -0.63, respectively. In addition, it was demonstrated that the proposed method could be applied to the short-term extraction of RSA amplitude. We proposed a simple and effective method to extract the amplitude of the respiratory sinus arrhythmia (RSA) in the respiratory-phase domain and the results show that this method can estimate cardiac vagal activity more accurately than frequency analysis.

Tutorial on Multivariate Autoregressive Modelling

In the present paper, the theoretical background of multivariate autoregressive modelling (MAR) is explained. The motivation for MAR modelling is the need to study the linear relationships between signals. In biomedical engineering, MAR modelling is used especially in the analysis of cardiovascular dynamics and electroencephalographic signals, because it allows determination of physiologically relevant connections between the measured signals. In a MAR model, the value of each variable at each time instance is predicted from the values of the same series and those of all other time series. The number of past values used is called the model order. Because of the inter-signal connections, a MAR model can describe causality, delays, closed-loop effects and simultaneous phenomena. To provide a better insight into the subject matter, MAR modelling is here illustrated with a model between systolic blood pressure, RR interval and instantaneous lung volume.

Non-invasive monitoring of left ventricular contractility and ventilatory mechanics.

The maximum left ventricular elastance (Emax<inf>iv</inf>) is a valuable index of ventricular contractility. However, conventional methods for its measurements are too invasive for routine use. We propose a non-invasive technique for tracking changes in Emax<inf>iv</inf>by mathematically determining the dynamic coupling between beat-to-beat fluctuations in instantaneous lung volume and arterial blood pressure measured during random-interval breathing. The mathematical technique also provides an estimate of the dominant time constant of the ventilatory system (), which is equal to the product of the airway resistance and lung compliance. We then describe a theoretical evaluation of the technique with respect to realistic beat-to-beat variability generated by a cardiovascular simulator. Our results show that the technique is able to detect actual changes in the Emax<inf>iv</inf>and values of the simulator. With successful experimental testing, the technique may ultimately be employed to help guide therapy in heart failure patients.

Comparison of Closed-loop and Open-loop Models in the Assessment of Cardiopulmonary and Baroreflex Gains

Both open- and closed-loop models of beat-to-beat cardiovascular control have been suggested. We tested whether the modelling yields different results with real data while assessing cardiopulmonary and baroreflex gains. Two autoregressive models are described to resolve causal relationships between systolic blood pressure (SBP), RR-interval (RRI) and instantaneous lung volume (ILV): a closed-loop model which takes into account both the RRI changes induced by changes in SBP and the SBP changes mediated by changes in RRI, and an open-loop model which does not have a link from RRI to SBP. The performance of the models was compared in 14 healthy men in supine and standing positions under control conditions and under conditions of beta -sympathetic and parasympathetic pharmacological blockades. Transfer function gains were computed from ILV to RRI (cardiopulmonary gain) and from SBP to RRI (baroreflex gain). The measurements were done under controlled random-interval breathing. The gains identified by the open-loop model tended to be higher than those from the closed-loop model, but the differences did not reach statistical significance. Importantly, the two models discriminated the changes in transfer gains between different interventions equally well. Because the interactions between SBP and RRI occur physiologically in a closed-loop condition, the closed-loop model provides a theoretical advantage over the open-loop model. However, in practise, it seems to be little reason to select one over the other due to methodological errors when estimating cardiopulmonary or baroreflex transfer gains.

Effects of simulated microgravity on closed-loop cardiovascular regulation and orthostatic intolerance: analysis by means of system identification.

Microgravity-induced orthostatic intolerance (OI) continues to be a primary concern for the human space program. To test the hypothesis that exposure to simulated microgravity significantly alters autonomic nervous control and, thus, contributes to increased incidence of OI, we employed the cardiovascular system identification (CSI) technique to evaluate quantitatively parasympathetic and sympathetic regulation of heart rate (HR). The CSI method analyzes second-to-second fluctuations in noninvasively measured HR, arterial blood pressure, and instantaneous lung volume. The coupling mechanisms between these signals are characterized by using a closed-loop model. Parameters reflecting parasympathetic and sympathetic responsiveness with regard to HR regulation can be extracted from the identified coupling mechanisms. We analyzed data collected from 29 human subjects before and after 16 days of head-down-tilt bed rest (simulated microgravity). Statistical analyses showed that parasympathetic and sympathetic responsiveness was impaired by bed rest. A lower sympathetic responsiveness and a higher parasympathetic responsiveness measured before bed rest identified individuals at greater risk of OI before and after bed rest. We propose an algorithm to predict OI after bed rest from measures obtained before bed rest.

A forward model-based validation of cardiovascular system identification.

We present a theoretical evaluation of a cardiovascular system identification method that we previously developed for the analysis of beat-to-beat fluctuations in noninvasively measured heart rate, arterial blood pressure, and instantaneous lung volume. The method provides a dynamical characterization of the important autonomic and mechanical mechanisms responsible for coupling the fluctuations (inverse modeling). To carry out the evaluation, we developed a computational model of the cardiovascular system capable of generating realistic beat-to-beat variability (forward modeling). We applied the method to data generated from the forward model and compared the resulting estimated dynamics with the actual dynamics of the forward model, which were either precisely known or easily determined. We found that the estimated dynamics corresponded to the actual dynamics and that this correspondence was robust to forward model uncertainty. We also demonstrated the sensitivity of the method in detecting small changes in parameters characterizing autonomic function in the forward model. These results provide confidence in the performance of the cardiovascular system identification method when applied to experimental data.

Decreased fractal component of human heart rate variability during non-REM sleep.

The physiological significance of the fractal component of short-term, spontaneous heart rate variability (HRV) in humans remains unclear. The aim of the present study was to gain further information about the respective fractal components by extracting them from HRV, blood pressure variability (BPV), and instantaneous lung volume (ILV) time series via coarse graining spectral analysis in nine healthy subjects during waking and sleep states. The results show that the contribution made by the fractal component to the total variance in the beat-to-beat R-R interval declined significantly as the depth of non-rapid eye movement (non-REM) sleep increased, that the ILV time series was largely periodic (i.e., nonfractal), and that BPV was unaffected by sleep stage. Finally, the fractal component of HRV during REM sleep was found to be quite similar to that seen during waking. These results suggest that mechanisms involving electroencephalographic desynchronization and/or conscious states of the brain are reflected in the fractal component of HRV.

A dynamic analysis of chest wall motions with MRI in healthy young subjects.

The objective of this study was to analyse respiratory-related motion of the chest wall with non-invasive method. Using magnetic resonance image (MRI), 30 sequential images (scanning time, 0.4 s per image) on sagittal, axial and coronal planes were obtained in nine healthy young subjects during quiet breathing (QB) and maximal deep breathing (MDB). The coronal planes were obtained in five of nine subjects during MDB. Ventilation was simultaneously measured with pneumotachometer. There was a linear correlation between instantaneous lung volume and lung cross-sectional area. Motion of the diaphragm and rib cage was also linearly related to instantaneous lung volume. The exception was lower anteroposterior (AP) diameter of the rib cage. The contribution of individual part of the chest wall motion to a unit lung volume change was assessed by slope (S) of the linear regression line. The S at the anterior diaphragm was significantly smaller than those at middle and posterior parts during MDB. The S of middle and posterior diaphragmatic motion was approximately five times that of AP motion of upper rib cage. The S of AP motion of upper rib cage was twice that of transverse motion during either QB or MDB. We concluded that dynamic MRI study with concurrent ventilation measurement is a simple and reliable method for evaluation of local chest wall motion, and that neither diaphragm nor rib cage works as a single functional unit during active ventilation.

Cardiorespiratory interactions during periodic breathing in awake chronic heart failure patients.

We applied spectral techniques to the analysis of cardiorespiratory signals [instantaneous lung volume (ILV), instantaneous tidal volume (ITV), arterial O(2) saturation (Sa(O(2))) at the ear, heart rate (HR), systolic (SAP), and diastolic (DAP) arterial pressure] during nonapneic periodic breathing (PB) in 29 awake chronic heart failure (CHF) patients and estimated the timing relationships between respiratory and slow cardiovascular (<0.04 Hz) oscillations. Our aim was 1) to elucidate major mechanisms involved in cardiorespiratory interactions during PB and 2) to test the hypothesis of a central vasomotor origin of PB. All cardiovascular signals were characterized by a dominant (>/=84% of total power) oscillation at the frequency of PB (mean +/- SE: 0.022 +/- 0.0008 Hz), highly coherent (>/=0.89), and delayed with respect to ITV (ITV-HR, 2.4 +/- 0.72 s; ITV-SAP, 6.7 +/- 0.65 s; ITV-DAP, 3.2 +/- 0.61 s; P < 0.01). Sa(O(2)) was highly coherent with (coherence function = 0.96 +/- 0. 009) and almost opposite in phase to ITV. These findings demonstrate the existence of a generalized cardiorespiratory rhythm led by the ventilatory oscillation and suggest that 1) the cyclic increase in inspiratory drive and cardiopulmonary reflexes and 2) mechanical effects of PB-induced changes in intrathoracic pressure are the more likely sources of the HR and blood pressure oscillations, respectively. The timing relationship between ITV and blood pressure signals excludes the possibility that PB represents the effect of a central vasomotor rhythm.

Heart Rate Signal Decomposition

This paper proposes a method for decomposing heart rate fluctuations into background, respiratory and blood pressure oriented fluctuations. A signal cancellation scheme using the adaptive RLS algorithm has been introduced for cancelling respiration and blood pressure oriented changes in the heart rate fluctuations. The computer simulation confirmed the validity of the proposed method. Then, heart rate fluctuations, instantaneous lung volume and blood pressure changes are simultaneously recorded from eight normal subjects aged 20-24 years. It was shown that after signal decomposition, the power spectrum of the heart rate showed a consistent monotonic 1/fa type pattern. The proposed method enables a clear interpretation of heart rate spectrum removing uncertain large individual variations due to the respiration and blood pressure change.

Determination of baroreflex gain using auto-regressive moving-average analysis during spontaneous breathing.

The heart rate component of the arterial baroreflex gain (BRG) was determined with auto-regressive moving-average (ARMA) analysis during each of spontaneous (SB) and random breathing (RB) protocols. Ten healthy subjects completed each breathing pattern on two different days in each of two different body positions, supine (SUP) and head-up tilt (HUT). The R-R interval, systolic arterial pressure (SAP) and instantaneous lung volume were recorded continuously. BRG was estimated from the ARMA impulse response relationship of R-R interval to SAP and from the spontaneous sequence method. The results indicated that both the ARMA and spontaneous sequence methods were reproducible (r = 0.76 and r = 0.85, respectively). As expected, BRG was significantly less in the HUT compared to SUP position for both ARMA (mean +/- SEM; 3.5 +/- 0.3 versus 11.2 +/- 1.4 ms mmHg-1; P < 0.01) and spontaneous sequence analysis (10.3 +/- 0.8 versus 31.5 +/- 2.3 ms mmHg-1; P < 0.001). However, no significant difference was found between BRG during RB and SB protocols for either ARMA (7.9 +/- 1.4 versus 6.7 +/- 0.8 ms mmHg-1; P = 0.27) or spontaneous sequence methods (21.8 +/- 2.7 versus 20.0 +/- 2.1 ms mmHg-1; P = 0.24). BRG was correlated during RB and SB protocols (r = 0.80; P < 0.0001). ARMA and spontaneous BRG estimates were correlated (r = 0.79; P < 0.0001), with spontaneous sequence values being consistently larger (P < 0.0001). In conclusion, we have shown that ARMA-derived BRG values are reproducible and that they can be determined during SB conditions, making the ARMA method appropriate for use in a wider range of patients.

Association between hemodynamic impairment and cheyne-stokes respiration and periodic breathing in chronic stable congestive heart failure secondary to ischemic or idiopathic dilated cardiomyopathy

Irregular breathing occurs frequently in patients with congestive heart failure (CHF) both during daytime and nighttime. Many factors are involved in the genesis of these breathing abnormalities, but the role of the hemodynamic impairment remains controversial. This study investigated the relation between worsening ventricular function and the frequency of respiratory disorders in patients with mild to severe CHF. One hundred fifty patients with CHF (mean age 53 ± 8 years, left ventricular (LV) ejection fraction 26 ± 7, in New York Heart Association [NYHA] classes II to IV, and who underwent stable therapy for ≥2 weeks) were studied. Analysis of instantaneous lung volume signal and arterial oxygen saturation during awake daytime revealed a normal respiratory pattern in 63 patients, whereas 87 had a persistent alteration of breathing, with a typical Cheyne-Stokes respiration (CSR) in 42 and periodic breathing (PB [oscillation of tidal volumes without apnea]) in 45 patients. Patients with PB and CSR showed a more pronounced hemodynamic impairment with a significantly reduced cardiac index, an increased pulmonary arterial wedge pressure, and a longer lung-to-ear circulation time (LECT) compared with patients with normal respiratory patterns. In a logistic regression model that included all of the variables significantly associated with breathing disorders, cardiac index and LECT emerged as the major determinants of CSR. In those patients with LECT ≥30 seconds (upper quartile) and cardiac index ≤1.9 L/min/m 2 (lower quartiles), the incidence of CSR was significantly higher (69%) than in patients with lower LECT and higher cardiac index (14%, p <0.001). In conclusion, abnormalities of breathing activity during daytime are significantly associated with a prolonged circulation time and a more severe impairment of systolic and diastolic LV indexes.

System identification of closed-loop cardiovascular control mechanisms: diabetic autonomic neuropathy.

We applied cardiovascular system identification (CSI) to characterize closed-loop cardiovascular regulation in patients with diabetic autonomic neuropathy (DAN). The CSI method quantitatively analyzes beat-to-beat fluctuations in noninvasively measured heart rate, arterial blood pressure (ABP), and instantaneous lung volume (ILV) to characterize four physiological coupling mechanisms, two of which are autonomically mediated (the heart rate baroreflex and the coupling of respiration, measured in terms of ILV, to heart rate) and two of which are mechanically mediated (the coupling of ventricular contraction to the generation of the ABP wavelet and the coupling of respiration to ABP). We studied 37 control and 60 diabetic subjects who were classified as having minimal, moderate, or severe DAN on the basis of standard autonomic tests. The autonomically mediated couplings progressively decreased with increasing severity of DAN, whereas the mechanically mediated couplings were essentially unchanged. CSI identified differences between the minimal DAN and control groups, which were indistinguishable based on the standard autonomic tests. CSI may provide a powerful tool for assessing DAN.