Abstract
We propose a singlelead ECGbased heart rate variability (HRV) analysis algorithm to quantify autonomic nervous system activity during meditation. Respiratory sinus arrhythmia (RSA) induced by breathing is a dominant component of HRV, but its frequency depends on an individual’s breathing speed. To address this RSA issue, we designed a novel HRV tachogram decomposition algorithm and new HRV indices. The proposed method was validated by using a simulation, and applied to our experimental (mindfulness meditation) data and the WESAD opensource data. During meditation, our proposed HRV indices related to vagal and sympathetic tones were significantly increased (p < 0.000005) and decreased (p < 0.000005), respectively. These results were consistent with selfreports and experimental protocols, and identified parasympathetic activation and sympathetic inhibition during meditation. In conclusion, the proposed method successfully assessed autonomic nervous system activity during meditation when respiration influences disrupted classical HRV. The proposed method can be considered a reliable approach to quantify autonomic nervous system activity.
Similar content being viewed by others
Introduction
Due to the increase in the number of mental disorders, various mental healthcare therapies (e.g., psychotherapy, digital therapeutics, and meditation) have been widely used to improve emotional wellness^{1}. Over the past few decades, mindfulness meditation has drawn much attention because of its benefits, such as emotion regulation, increased awareness, and improved cognitive performance^{2,3,4}, and it has been clinically used to reduce chronic pain, sleep disturbance, anxiety, distress, and depression^{5,6,7}. Many neuroscientific studies have investigated changes in the brain during meditation^{2}, and the activation of several brain regions, such as the frontopolar cortex, sensory cortex, insula, hippocampus, anterior cingulate cortex, midcingulate cortex, and orbitofrontal cortex, has been observed^{8}. These brain regions are related to metaawareness, body awareness, memory processes, and emotion regulation. In particular, the insular and midcingulate cortices play central roles in the central autonomic network, which controls the activity of parasympathetic and sympathetic nerves^{9}. Through this neurological pathway, it seems that body relaxation is induced during meditation.
As a technologybased therapy, biofeedback via a wearable device has been developed^{10}. Biofeedback devices help users carry out a therapeutic protocol (e.g., meditation) by means of their physiological signals as feedback. Electroencephalogrambased neurofeedback training is a representative biofeedback therapy that is used to treat various mental disorders, such as insomnia, anxiety, depression, and addiction^{11}. On the other hand, electroencephalogrambased digital meditation was proposed to enhance attention and working memory^{12,13}. Although neurofeedback treatment shows various clinical effects, an electroencephalogram sensor is inconvenient. Recently, a new modality called skin sympathetic nerve activity has been developed to noninvasively observe the activity of sympathetic nerves^{14,15}. It can be simply recorded from the high frequency component (500–1000 Hz) of an analog electrocardiogram (ECG), but it is very vulnerable to motion artifacts due to its weak amplitude^{16,17}. Because it is easy to measure ECG signals with a high signaltonoise ratio in daily life, heart rate variability (HRV)based biofeedback training is drawing attention^{18,19}. In this study, we propose a singlelead ECGbased HRV analysis algorithm to investigate the changes in vagal and sympathetic tones during meditation (see Fig. 1).
ECG based cardiopulmonary coupling analysis and HRV analysis have been used to assess psychiatric conditions. Cardiopulmonary coupling analysis measures the synchronization between the heartbeat interval and respiration by using the spectral coherence, cross entropy and phase locking value (PLV)^{20,21,22}. This synchronization phenomenon is called respiratory sinus arrhythmia (RSA), which is the modulation of efferent parasympathetic nerves caused by baroreflexes (arterial baroreflex, lung stretch reflex, and Bainbridge reflex) during the inspiration phase^{23,24,25,26}. RSA is affected by the tidal volume and respiratory rate (RR)^{27,28,29}, and its quantity reflects sleep quality and apnea^{20,22}. To capture other neurophysiological changes, HRV is widely used, and its frequency analysis reflects the balance between vagal and sympathetic tones. The former shows a fast response, and the latter dominates the lowfrequency band of HRV and seems to be independent of the baroreflex^{30}.
Meditation (or relaxation) training leads to parasympathetic activation and sympathetic inhibition, and it is supported by several modalities, such as muscle sympathetic nerve activity^{31}, functional MRI^{2}, and electroencephalogram^{32}. Some HRV studies on meditation also reported an increase in highfrequency power^{32,33,34,35}, which indicates vagal tone, but other studies showed the opposite results^{36,37,38,39,40}. These contradictory results seem to be due to the influence of RSA. When the RR < 0.15 Hz, the RSA component moves from the high to lowfrequency region, and then the low and highfrequency powers can be overestimated and underestimated, respectively^{28}. Hence, the fluctuation of an individual’s RR led to significant variation in HRV analysis^{41}, and there was no report on the identification of cardiac sympathetic inhibition during meditation.
To separate the RSA effect and other autonomic nervous system activities in HRV analysis, various decomposition algorithms have been utilized, such as independent component analysis^{42}, adaptive noise cancellers^{43}, autoregressive moving average with exogenous input model (ARMAx)^{44}, and orthogonal subspace projection (OSP)^{45,46}. The adaptive noise canceller, ARMAx, and OSP share the same algorithmic structure, with OSP exhibiting the best performance^{45}. Even though these algorithms significantly enhance the reliability of HRV, they assume a linear relationship between the RSA component and a respiration signal, while in reality the relationship is nonlinear^{47}. As an alternative, a notch filter with a Gaussian bell shape (Gauss) was employed, but it assumed that the instantaneous RR is constant^{48}. In this study, to overcome the nonlinearity and nonstationarity of RSA, we proposed a novel HRV tachogram decomposition algorithm, and new HRV indices to quantify vagal and sympathetic tones during meditation (see Fig. 2).
The proposed method is described in the “Materials and methods” section. We compared the performance of decomposition algorithms using simulation, and observed the changes in the vagal and sympathetic tones during meditation through real ECG data in “Results” section. In “Discussion” section, the results are discussed.
Materials and methods
Since HRV reflects the balance of the autonomic nervous system, HRV analysis is frequently used to assess stress and psychiatric conditions. The first step of HRV analysis is QRS detection on the ECG. In this study, we bandpass filtered an ECG with 0.5–50 Hz by a 3rdorder Butterworth, and we employed the PanTompkins algorithm to detect Rpeaks of the ECG^{49}. The interval between successive Rpeaks (RRI) is accumulated over a predefined time window. In this study, we adopted a 5 min time window for shortterm HRV analysis. To detect ectopic beats and incorrect peaks, we calculated the kth derivative of the instantaneous heart rate (HR)^{50} as follows:
where R_{k} is the kth Rpeak. When an HR’_{k} was greater than a predefined threshold, R_{k} was excluded from the HRV analysis, and the corrected Rpeaks were doublechecked by human experts. Because the RRI is sampled at irregular time points (heartbeats), the RRI must be converted into an evenly sampled time series for frequency analysis. A uniformly sampled signal of the RRI, called a tachogram, was reconstructed using cubic spline interpolation, and we resampled the tachogram to 4 Hz, as in previous studies^{44,45,46}. For precise HRV analysis, it is necessary to distinguish the cardiac vagal tone modulated by respiration and other autonomic nervous system activities. Hence, we consider the original tachogram x_{RAW} as follows:
where x_{RSA} and x_{Rf} are the RSA and RSAfree tachograms, respectively. x_{Rf} is obtained by removing x_{RSA} from x_{RAW}. The tachogram decomposition algorithm will be described in the next subsection. In this study, since we focus on the tachogram decomposition algorithm, we adopt only frequency domain HRV indices.
ECGderived respiration
To measure the respiration signal, various sensors have been utilized, such as a pressure transducer and a thermistor. These sensors provide accurate information but are both expensive and inconvenient to wear. As an alternative, several algorithms to estimate the respiration signal from biomedical signals, such as photoplethysmogram and ECG, have been proposed^{51,52,53,54}. In this study, we employed an ECGderived respiration (EDR) algorithm that captures the change in electrical impedance and the cardiac vector according to chest expansion. Here, we briefly review 4 frequently used EDR algorithms: EDR_{A} (Rwave amplitude), EDR_{QR} (QR slope), EDR_{RS} (RS slope) and EDR_{SR} (slope range). Each EDR algorithm uses a respirationrelated sample r(i) at the ith heartbeat as follows:

EDR_{A}: r_{A}(i) = Rpeak amplitude

EDR_{QR}: r_{QR}(i) = steepest ascent slope on QR wave

EDR_{RS}: r_{RS}(i) = steepest descent slope on RS wave

EDR_{SR}: r_{SR}(i) = r_{QR}(i) − r_{RS}(i)
By interpolating the respirationrelated samples, the EDR signal is generated. The overall procedure is explained in detail in^{51}.
Tachogram decomposition
Previous algorithms
To decompose raw tachogram x_{RAW} into x_{RSA} and x_{Rf}, various algorithms have been proposed, such as Gauss, ARMAx and OSP. Both the ARMAx and OSP algorithms assume x_{RSA} as follows:
where RESP is a respiratory basis matrix. The trajectory matrix^{46} or wavelet coefficients^{45} of EDR have been utilized as the RESP matrix. In this study, the trajectory matrix was employed for the RESP. We assigned 3 s to the lag size of the trajectory matrix of the ARMAx, while the lag size of OSP was set to the minimum value between the minimum description length and the Akaike information criterion, with a maximum lag of 10 s. The RSAfree tachogram is estimated as follows:
The ARMAx and the OSP are described in^{44,45,46} in detail. As another tachogram decomposition algorithm, a notch filter with Gaussian bell shape (Gauss) was proposed. Its magnitudesquared frequency response is:
where θ represents the notch frequency and corresponds to the average RR. α and β are fitting parameters and it can be computed from the spectrum of x_{RAW} by using a least square method. The filtered signal will be \({\overline{{\varvec{x}}} }_{Rf}\), and \({\overline{{\varvec{x}}} }_{RSA}\) is acquired by subtracting \({\overline{{\varvec{x}}} }_{Rf}\) from x_{RAW}. The overall procedure of Gauss is described in^{48}.
Proposed algorithm
Based on the synchronization between the frequency of x_{RSA} and the instantaneous RR, we designed a novel tachogram decomposition algorithm. Because it has a zerophase response, we call it the zerophase line enhancer (ZLE). The ZLE can be interpreted as an adaptive filter without phase distortion^{55}. As shown in Fig. 3, the ZLE has two parts: (a) instantaneous RR estimation and (b) timevariant forward–backward IIR notch filtering. To estimate the instantaneous RR, we employed a smoothed pseudo Wigner–Ville distribution of the EDR signal^{56,57} as follows:
where N represents the length of the EDR (N = 5 min∙60∙4 Hz). g[] and v[] are the time and frequency windows, respectively; a Kaiser window (γ = 20) was employed for g[] and v[]. From the time–frequency representation ω[k, n], we estimated the instantaneous RR θ[n] by using a recursive algorithm to minimize the problem of outliers. The maximum point among frequency bins adjacent to θ[n1] was assigned to θ[n], as described in Algorithm 1.
The frequency component of x_{RAW} corresponding to the instantaneous RR θ[n] will be x_{RSA}, and the ZLE removes x_{RSA} by means of a tunable IIR notch filter^{58} as follows:
where r is the polezero contraction factor, which determines the stop band. When r is closer to 1, the frequency response of the notch filter becomes narrower. We set r to 0.93, and then the stop band of H(z, θ[n]) is 0.047 Hz. As shown in Fig. 3, the timevariant forward–backward IIR notch filtering performs four sequential steps: 1—forward IIR notch filtering, 2—time reversal, 3—backward IIR notch filtering, and 4—time reversal. The zerophase response can be explained in the zdomain as follows:
where F(z) and B(z) mean forward and backwardfiltered signals, respectively. P(z) represents the timereversed F(z). The filter output is the estimated RSAfree tachogram \({\overline{{\varvec{x}}} }_{Rf}\), and the RSA tachogram is estimated as follows:
RSA indices
RSA has been utilized as a measure of cardiac vagal function. RSA can be quantified by cardiopulmonary coupling analysis. In this study, we utilized the PLV^{22,39,59}. The PLV is broadly used to measure a phase interaction between two signals and is defined as follows:
where φ_{1}(n) and φ_{2}(n) represent the phases of the respiration signal (or EDR) and tachogram x_{RAW}, respectively. The phases are calculated by a Hilbert transform. As another measure for RSA, the extent of RSA (RSA_{E})^{27,39,60} was calculated by the median value of RRI variation (longest RRI—shortest RRI) within each breathing cycle as follows:
where RRI^{k} represents the subset of RRIs within the kth breathing cycle. max and min are maximum and minimum functions, respectively. K indicates the number of breaths taken in the time window (5 min). Because RSA_{E} is based on the RRI, the unit of RSA_{E} is ms.
HRV indices
In HRV analysis, the power spectrum of x_{RAW} is segmented into lowfrequency (0.04–0.15 Hz) and highfrequency (0.15–0.4 Hz) bands, and each band power is called LF (low frequency) and HF (high frequency)^{61}. HF is regarded as an index for vagal tone, and LF reflects both sympathetic and vagal tones. To quantify sympathetic tone, LF/HF is generally used. Since these HRV indices were insufficient to cover the aforementioned RSA issue, it is necessary to simultaneously use both power spectra of \({\overline{{\varvec{x}}} }_{Rf}\) and \({\overline{{\varvec{x}}} }_{RSA}\); Subscripts Rf and RSA were used to distinguish each band power (\({\overline{{\varvec{x}}} }_{Rf}\): LF_{Rf} and HF_{Rf}, \({\overline{{\varvec{x}}} }_{RSA}\): LF_{RSA} and HF_{RSA}). To assess sympathovagal balance, Varon et al. designed an HRV index SB_{U} based on the separated band powers^{46} as follows:
Since x_{RSA} and HF_{Rf} purely indicate vagal tone but LF_{Rf} is related to both vagal and sympathetic tones, we proposed the new HRV indices related to vagal (Eq. 17) and sympathetic (Eq. 18) tones as follows:
We employed Welch’s method to compute the power spectra.
Simulation
To compare the tachogram decomposition algorithms, we designed a simulated tachogram \({\widetilde{{\varvec{x}}}}_{RAW}\). First, to cover various situations, we considered three respiration types as follows:
where \(\dot{\psi }\left(n\right)\), the derivative of \(\psi (n)\), means the simulated RR. ①, ②, and ③ represent the frequencies of a sine wave (0.15 Hz), linear chirp (0.04–0.4 Hz), and sinusoidal frequency modulated signal (0.04–0.4 Hz). Since not only the frequency but also the amplitude of the tachogram is modulated by RR^{27,28,29}, we devised an amplitude modulation function for the simulated RSA component \({\widetilde{{\varvec{x}}}}_{RSA}\). Specifically, because the amplitude of the tachogram with respect to RR in^{29} was skewed to the right, we utilized the gamma distribution function as the amplitude modulation function m() as follows:
where Г(α) is the gamma function. α and β are shape and rate parameters (α = 2, β = 15). Note that when \(\dot{\psi }\left(n\right)\) is constant the amplitude modulation function m() becomes constant. On the other hand, as the simulated RSAfree tachogram \(\widetilde{{\varvec{x}}}\)_{Rf}, we employed white Gaussian noise as follows:
where \({\sigma }^{2}\) is the variance of the white Gaussian noise. We adjusted \({\sigma }^{2}\) so that the power ratio of \(\widetilde{{\varvec{x}}}\)_{RSA} and \(\widetilde{{\varvec{x}}}\)_{Rf} was equal (0 dB), and then the simulated tachogram \(\widetilde{{\varvec{x}}}\)_{RAW} was given by their summation.
For rigorous comparison, we performed Monte Carlo simulation with 1000 independent simulations.
Database
To investigate the changes in HRV indices according to the tachogram decomposition algorithms during meditation, we utilized two real ECG databases: mindfulness meditation data and WESAD data.
Mindfulness meditation data
We collected a singlelead ECG and a respiration signal from 8 male and 8 female subjects (age = 28.5 ± 4.2 years) during mindfulness meditation. All subjects were novice meditators who had engaged in meditation within 1–2 times. All subjects rested for 10 min and then performed mindfulness meditation for 40 min under the guidance of a meditation teacher. During mindfulness meditation, subjects were guided to maintain awareness of their breathing and sensations without judgment. Breathing speed was not restricted. All subjects closed their eyes and remained in a sitting position to minimize motion artifacts in the ECG. They completed a visual analog scale (VAS) for stress, tension, and concentration before and after mindfulness meditation. Each ECG was measured by a wearable ECG patch (EP200, Life Science Technology^{®}, Seoul, Korea), and the sampling rate was 250 Hz. The reference respiration signal was recorded by a respiratory transducer (RSPECR, Biopac^{®}, California, USA) with a sampling rate of 250 Hz. We used the reference respiration signal to evaluate EDR and measure vital signs and RSA indices, and we employed EDR for the tachogram decomposition algorithms. Informed consent was obtained from all subjects, and the institutional review board of the Korea Institute of Science and Technology (KIST) approved all procedures in this study (KIST2019015, July. 19, 2019). Informed consent was obtained from all participants prior to the experiment, and all experiments were conducted in strict accordance with KIST ethics guidelines and the declaration of Helsinki.
WESAD data
We used opensource data (WESAD, Wearable Stress and Affect Detection) to verify the proposed method in a public domain. This dataset contains the ECGs of 15 subjects (age = 27.5 ± 2.4 years), and its sampling rate is 700 Hz. The experimental protocol consisted of four different affective states (neutral, stress, amusement, meditation), and each emotion was mapped onto a twodimensional affective space (arousal and valence) by using selfreports (SAM, SelfAssessment Manikins). The WESAD data are described in detail in^{62}.
Statistical evaluation
Since the simulation was performed with white Gaussian noise, ANOVA and t tests were used to compare the tachogram decomposition algorithms. In the case of real data (mindfulness meditation and WESAD data), nonparametric statistical tests were utilized because the number of subjects was not large. Specifically, the Wilcoxon signed rank test was employed when the two groups had the same sample size, and the Mann–Whitney U test was used otherwise. We defined p < 0.000005 as statistically significant. To depict the distribution, we utilized a box plot where the top and bottom boxes represent the 75th and 25th percentiles, and the center, top, and bottom lines represent the 50th, 90th, and 10th percentiles, respectively.
Results
Simulation
For each simulation type (RR: constant, linear function, and sinusoidal function), we calculated the Pearson correlation coefficient and root mean square error (RMSE) between \(\widetilde{{\varvec{x}}}\)_{Rf} and \(\overline{{\varvec{x}} }\)_{Rf} of the tachogram decomposition algorithms (Gauss, ARMAx, OSP, and ZLE). Figure 4 depicts the box plots of correlation coefficients and RMSEs ((a): a sine wave, (b): linear chirp, and (c): sinusoidal frequency modulated signal). For each simulation type, we conducted oneway ANOVA with Bonferroni correction to compare four algorithms (Gauss, ARMAx, OSP, and ZLE), and significant differences (p < 0.000005) were observed in both the Pearson correlation coefficients and RMSEs for all simulation types. For pairwise comparisons, post hoc tests were performed by using paired t tests, and all algorithm comparisons showed significant differences (p < 0.000005). The OSP showed the lowest RMSE and highest correlation coefficient when the simulated RR was constant (see Fig. 4a), but the ZLE had the lowest RMSE and highest correlation coefficient for other RR types (see Fig. 4b,c). It is notable that ZLE was robust to the fluctuation of instantaneous RR but contrastively Gauss did not work well in timevarying RR.
EDR evaluation
We evaluated the performance of the aforementioned EDRs (EDR_{A}, EDR_{QR}, EDR_{RS}, and EDR_{SR}) by using the reference respiration signal of the mindfulness meditation data. Because there was a phase difference between EDR and reference respiration signal, we utilized two measures that were not significantly affected by the phase difference. One is the PLV between each EDR and the reference respiration signal, and the other is the mean absolute error (MAE) between RRs calculated by each EDR and the reference respiration signal. The PLV and MAE reflect the similarity of phases and the accuracy of RR estimation, respectively. As shown in Table 1, EDR_{SR} showed the best performance as in previous study^{51}, so we employed EDR_{SR} in the following analysis.
Mindfulness meditation data
Before applying the tachogram decomposition algorithms, we observed the effect of mindfulness meditation through various measures: VAS for stress, RSA indices, and vital signs. First, we confirmed the stress reduction after meditation through the VAS for stress (5.35 ± 1.77 → 2.47 ± 1.01). Second, we calculated the vital signs and RSA indices with 5 min intervals, and their box plots are depicted in Fig. 5. The horizontal lines represent the median values of the (a) HR, (b) RR, (c) RSA_{E}, and (d) PLV. As meditation progressed, vital signs (HR and RR) decreased, and both the RSA_{E} and PLV increased. To assess the changes in the HR, RR, RSA_{E} and PLV, we performed statistical tests for these indices between the rest and meditation phases with the Mann–Whitney U test. Statistically significant differences (p < 0.000005) were observed in the instantaneous RR and RSA_{E}. For each tachogram decomposition algorithm (ARMAx, OSP, and ZLE), we calculated Pearson correlation coefficients between x_{RAW} and \(\overline{{\varvec{x}} }\)_{Rf}, and their box plots and median values are shown in Fig. 6 (green boxes and dotted line: Gauss, blue boxes and dashed line: ARMAx, yellow boxes and dash dotted line: OSP, and red boxes and solid line: ZLE). A higher correlation coefficient indicates that \(\overline{{\varvec{x}} }\)_{Rf} is less filtered or less distorted.
We computed the power spectra of tachograms and calculated HRV indices. Figure 7 shows the box plots and median values of the proposed HRV indices: (a) IDX_{PNA} and (b) IDX_{SNA}. The boxes and lines represent (a) IDX_{PNA}s and (b) IDX_{SNA}s of Gauss (green boxes and dotted line), ARMAx (blue boxes and dashed line), OSP (yellow boxes and dashdotted line), and ZLE (red boxes and solid line). To assess the trend of the HRV indices, we divided each record (50 min) into 1 rest phase and 4 meditation phases of equal length (10 min) and then compared the rest and each meditation phase by using a Wilcoxon’s twosample signed rank test (see Table 2). Table 2 depicts the mean and standard deviation of HRV indices. HF and IDX_{PNA} are the HRV indices for vagal tone, and other indices (LF/HF, SB_{U} and IDX_{SNA}) indicate sympathetic tone. No statistically significant differences were observed in HFs, but all IDX_{PNA}s of ZLE were significantly increased during meditation (p < 0.000005). Unlike ZLE, IDX_{PNA}s of previous tachogram decomposition algorithms did not significantly decrease at 20–30 min. For the HRV indices related to sympathetic tone, the LF/HF of the raw tachogram (x_{RAW}) significantly increased at 30–50 min, which seems to be caused by RSA, but all SB_{U}s and IDX_{SNA}s of ZLE were significantly diminished (see Fig. 7; Table 2). Although, after 40 min, all SB_{U}s and IDX_{SNA}s of all algorithms except for Gauss were significantly decreased, but no statistically significant differences were observed in most SB_{U}s and IDX_{SNA}s of previous tachogram decomposition algorithms (Gauss, ARMAx, and OSP).
WESAD data
The WESAD data contains ECGs with 4 affective states: neutral, stress, amusement, and meditation. The SAM selfreports showed that the lowest arousal scores (2.3 ± 1.4) occurred in the meditation state. As shown in Fig. 8, the RSA indices (RSA_{E} and PLV) and vital signs (HR and RR) had the highest and lowest values in the meditation state, respectively. For HR, RR, RSA_{E} and PLV, we performed statistical tests between the neutral and meditation states by using Mann–Whitney U tests. Statistically significant differences (p < 0.000005) were observed in all comparisons except HR. This is similar to the results of the previous subsection. Figure 9 depicts the box plots and median values of the proposed HRV indices: (a) IDX_{PNA} and (b) IDX_{SNA}. The boxes and lines depict (a) IDX_{PNA}s and (b) IDX_{SNA}s of Gauss (green boxes and dotted line), ARMAx (blue boxes and dashed line), OSP (yellow boxes and dashdotted line), and ZLE (red boxes and solid line). We evaluated the change between the neutral and other emotions of the HRV indices by using the Mann–Whitney U test. The mean and standard deviation of HRV indices are shown in Table 3. No statistically significant differences were observed in HFs, but IDX_{PNA} of ARMAx and ZLE significantly increased during meditation, in parallel with the RSA indices. Similar to the results of the previous subsection, SB_{U} and IDX_{SNA} of the ZLE significantly (p < 0.000005) decreased during meditation, but no statistically significant differences were observed in SB_{U} and IDX_{SNA} of other decomposition algorithms (Gauss, ARMAx and OSP) as shown in Fig. 9 and Table 3.
Discussion
The tachogram of HRV consists of various oscillations caused by complicated interactions between the heart and brain. In particular, the oscillation caused by RSA often forms most of the raw tachogram x_{RAW}. For the precise quantification of autonomic nervous system activity, it is necessary to decompose the raw tachogram into RSA and RSAfree tachograms. In this study, we proposed the novel tachogram decomposition algorithm (ZLE) and new HRV indices (IDX_{PNA} and IDX_{SNA}). We evaluated the performance of the proposed method through simulations, mindfulness meditation data, and WESAD data.
RSA, which is a synchrony between the heartbeat interval and breathing, reflects cardiac vagal control. To quantify RSA, we employed the PLV and RSA_{E}. Although the RSA_{E} and PLV capture different features (phase interaction and RRI variation), these indices increased during meditation (see Figs. 6, 9) as in previous study^{39}, which indicates parasympathetic activation. In HRV analysis, all IDX_{PNA}s of ZLE showed significant increases in contrast to HFs (see Tables 2, 3), and IDX_{PNA} appeared to be better than HF in identifying parasympathetic activation.
Sympathetic inhibition during meditation is supported not only by literature but also by experimental analysis. Breathing meditation and slow breathing have been reported to improve pulmonary gas exchange performance^{63} and lower blood pressure^{64}, galvanic skin response^{65} and muscle sympathetic nerve activity^{66,67,68}. Reductions in vital signs (HR and RR) and selfreport scores (stress and arousal) were also observed during meditation in our experiments, as shown in Figs. 6 and 9. Despite much evidence for sympathetic inhibition, the LF/HF of the classical HRV was noticeably increased during meditation^{35,36,37,38,39,40} (see Tables 2, 3); it seems to be due to the influence of RSA. To cancel the interference of RSA, C. Varon et al. suggested SB_{U}^{46} and we proposed IDX_{SNA}. Although SB_{U} (LF_{Rf}/(LF_{RSA} + HF_{RSA})) is similar to IDX_{SNA} (LF_{Rf}/(LF_{RSA} + HF_{RSA} + HF_{Rf})), IDX_{SNA} includes RSAfree vagal tone information (HF_{Rf}). Since RSA is insufficient to represent a whole cardiac vagal tone^{69}, the proposed IDX_{SNA} can be considered a more reasonable index to quantify sympathetic tone. In real meditation data, both SB_{U} and IDX_{SNA} of ZLE successfully identified sympathetic inhibition, and it may be the first report on the identification of cardiac sympathetic inhibition during meditation. As a result, both SB_{U} and IDX_{SNA} appear to be more reliable than classical LF/HF to identify sympathetic inhibition.
ARMAx and OSP share the same scheme, which finds a weight vector (w = (RESP^{T}∙RESP)^{−1}RESP∙X) and subtracts \(\overline{{\varvec{x}} }\)_{RSA} from the original tachogram x_{RAW}^{45}. When the RESP is the trajectory matrix of EDR, the ARMAx and OSP can be interpreted as adaptive noise cancellers with a least square solution; then, the EDR and the RSA component (x_{RSA}) can be regarded as a noise reference signal and true noise, respectively. The algorithm performance is determined by a linear correlation between x_{RSA} and the EDR^{70}, but their true relationship is nonlinear^{47}. Specifically, both the amplitude and frequency of the original tachogram are modulated by the RR and tidal volume^{27,28,29}, and such modulations cannot be expressed by a linear transform of EDR. As an alternative, Gauss was suggested, which assumes paced respiration^{48}, but instantaneous RR generally changes with time. In the simulation, previous decomposition algorithms (Gauss, ARMAx, and OSP) performed well when RR was constant, but their performance was insufficient in timevarying RR. However, ZLE was robust to the fluctuations of simulated RR (see Fig. 4). Note that abrupt changes in RR often occur during breathing meditation. In fact, the HRV indices (IDX_{PNA}, SB_{U}, and IDX_{SNA}) of ZLE certainly identified both parasympathetic activation and sympathetic inhibition during meditation when no statistically significant differences were observed in some IDX_{PNA}s and most IDX_{SNA} of both ARMAx and OSP (see Tables 2, 3). Therefore, the ZLE seems to be the most reliable tachogram decomposition algorithm when RR dynamically changes.
Despite the promising performance of the proposed ZLE and HRV indices, they have limitations. First, if some oscillation of the RSAfree tachogram (x_{Rf}) overlaps with the stop band of the notch filter in the frequency domain, then the corresponding oscillation will be suppressed. This filtering loss can be observed through the correlation coefficient between x_{RAW} and \(\overline{{\varvec{x}} }\)_{Rf} when the influence of RSA was small (e.g., resting state). As shown in Fig. 6, the loss of ZLE seems to be insignificant. Second, the proposed method did not identify sympathetic activation. In the stress state, sympathetic nerves are activated, but none of the SB_{U}s or IDX_{SNA}s show a significant increase. In future works, we plan to advance the proposed method for stress assessment.
Conclusions
To precisely assess autonomic nervous system activity, we proposed a novel tachogram decomposition algorithm (ZLE) and new HRV indices (IDX_{PNA} and IDX_{SNA}). ZLE clearly decomposed x_{RAW} into \(\overline{{\varvec{x}} }\)_{RSA} and \(\overline{{\varvec{x}} }\)_{Rf}, and IDX_{PNA} and IDX_{SNA} identified parasympathetic activation and sympathetic inhibition during meditation, respectively. This study may be the first report on the identification of cardiac sympathetic inhibition during meditation (or slow breathing). Although the overlapping issue of the ZLE still remains, the ZLE was more robust than previous tachogram decomposition algorithms (Gauss, ARMAx and OSP) when RR dynamically fluctuates. Since the proposed approach requires only a singlelead ECG, we expect that it will be used in various fields, such as the internet of Medical Things (IoMT), digital healthcare, digital meditation, and digital therapeutics.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
References
Kutz, I., Borysenko, J. Z. & Benson, H. Meditation and psychotherapy: A rationale for the integration of dynamic psychotherapy, the relaxation response, and mindfulness meditation. Am. J. Psychiatry 142, 1–8 (1985).
Tang, Y. Y., Hölzel, B. K. & Posner, M. I. The neuroscience of mindfulness meditation. Nat. Rev. Neurosci. 16, 213–225. https://doi.org/10.1038/nrn3916 (2015).
Wu, R. et al. Brief mindfulness meditation improves emotion processing. Front. Neurosci. 13, 1074 (2019).
Eberth, J. & Sedlmeier, P. The effects of mindfulness meditation: A metaanalysis. Mindfulness (N Y) 3, 174–189 (2012).
Salmon, P. et al. Mindfulness meditation in clinical practice. Cogn. Behav. Pract. 11, 434–446 (2004).
Carlson, L. E., Ursuliak, Z., Goodey, E., Angen, M. & Speca, M. The effects of a mindfulness meditationbased stress reduction program on mood and symptoms of stress in cancer outpatients: 6month followup. Support. Care Cancer 9, 112–123 (2001).
Winbush, N. Y., Gross, C. R. & Kreitzer, M. J. The effects of mindfulnessbased stress reduction on sleep disturbance: a systematic review. Explore J. Sci. Health 3, 585–591 (2007).
Fox, K. C. R. et al. Is meditation associated with altered brain structure? A systematic review and metaanalysis of morphometric neuroimaging in meditation practitioners. Neurosci. Biobehav. Rev. 43, 48–73. https://doi.org/10.1016/j.neubiorev.2014.03.016 (2014).
Beissner, F., Meissner, K., Bär, K. J. & Napadow, V. The autonomic brain: An activation likelihood estimation metaanalysis for central processing of autonomic function. J. Neurosci. 33, 10503–10511 (2013).
Hunkin, H., King, D. L. & Zajac, I. T. Wearable devices as adjuncts in the treatment of anxietyrelated symptoms: A narrative review of five device modalities and implications for clinical practice. Clin. Psychol. Sci. Pract. 26, e12290 (2019).
Brandmeyer, T. & Delorme, A. Meditation and neurofeedback. Front. Psychol. 4, 688 (2013).
Ziegler, D. A. et al. Closedloop digital meditation improves sustained attention in young adults. Nat. Hum. Behav. 3, 746–757 (2019).
Acabchuk, R. L., Simon, M. A., Low, S., Brisson, J. M. & Johnson, B. T. Measuring meditation progress with a consumergrade EEG device: Caution from a randomized controlled trial. Mindfulness (N Y) 12, 68–81 (2021).
Xing, Y. et al. An artifactresistant feature SKNAER for quantifying the burst of skin sympathetic nerve activity signal. Biosensors 12, 355 (2022).
Zhang, Y. et al. Dynamics of cardiac autonomic responses during hemodialysis measured by heart rate variability and skin sympathetic nerve activity: the impact of interdialytic weight gain. Front. Physiol. 13, 909 (2022).
Xing, Y. et al. Design and evaluation of an autonomic nerve monitoring system based on skin sympathetic nerve activity. Biomed. Signal. Process Control 76, 103681 (2022).
Kusayama, T. et al. Simultaneous noninvasive recording of electrocardiogram and skin sympathetic nerve activity (neuECG). Nat. Protoc. 15, 1853–1877 (2020).
Goessl, V. C., Curtiss, J. E. & Hofmann, S. G. The effect of heart rate variability biofeedback training on stress and anxiety: a metaanalysis. Psychol. Med. https://doi.org/10.1017/S0033291717001003 (2017).
DeschodtArsac, V., Lalanne, R., Spiluttini, B., Bertin, C. & Arsac, L. M. Effects of heart rate variability biofeedback training in athletes exposed to stress of university examinations. PLoS One 13, e0201388 (2018).
Thomas, R. J., Mietus, J. E., Peng, C.K. & Goldberger, A. L. An electrocardiogrambased technique to assess cardiopulmonary coupling during sleep. Sleep 28, 1151–1161 (2005).
Zhao, L., Yang, L., Su, Z. & Liu, C. Cardiorespiratory coupling analysis based on entropy and crossentropy in distinguishing different depression stages. Front. Physiol. 10, 359 (2019).
Niizeki, K. & Saitoh, T. Association between phase coupling of respiratory sinus arrhythmia and slow wave brain activity during sleep. Front. Physiol. 9, 1338 (2018).
DeBoer, R. W., Karemaker, J. M. & Strackee, J. Hemodynamic fluctuations and baroreflex sensitivity in humans: A beattobeat model. Am. J. Physiol. Heart Circ. Physiol. 253, 25 (1987).
Triedman, J. K. & Saul, J. P. Blood pressure modulation by central venous pressure and respiration: Buffering effects of the heart rate reflexes. Circulation 89, 169–179 (1994).
Taha, B. H., Simon, P. M., Dempsey, J. A., Skatrud, J. B. & Iber, C. Respiratory sinus arrhythmia in humans: An obligatory role for vagal feedback from the lungs. J. Appl. Physiol. 78, 638–645 (1995).
Piepoli, M. et al. Origin of respiratory sinus arrhythmia in conscious humans: An important role for arterial carotid baroreceptors. Circulation 95, 1813–1821 (1997).
Hirsch, J. A. & Bishop, B. Respiratory sinus arrhythmia in humans: How breathing pattern modulates heart rate. Am. J. Physiol. Heart Circ. Physiol. 10, 620–629 (1981).
Brown, T. E., Beightol, L. A., Koh, J. & Eckberg, D. L. Important influence of respiration on human R–R interval power spectra is largely ignored. J. Appl. Physiol. 75, 2310–2317 (1993).
Song, H. S. & Lehrer, P. M. The effects of specific respiratory rates on heart rate and heart rate variability. Appl. Psychophysiol. Biofeedback 28, 13–23 (2003).
Cooley, R. L. et al. Evidence for a central origin of the lowfrequency oscillation in RRinterval variability. Circulation 98, 556–561 (1998).
Park, J., Lyles, R. H. & BauerWu, S. Mindfulness meditation lowers muscle sympathetic nerve activity and blood pressure in AfricanAmerican males with chronic kidney disease. Am. J. Physiol. Regul. Integrat. Comp. Physiol. 307, R93–R101 (2014).
Tang, Y. Y. et al. Central and autonomic nervous system interaction is altered by shortterm meditation. Proc. Natl. Acad. Sci. USA 106, 8865–8870 (2009).
Krygier, J. R. et al. Mindfulness meditation, wellbeing, and heart rate variability: A preliminary investigation into the impact of intensive vipassana meditation. Int. J. Psychophysiol. 89, 305–313 (2013).
Azam, M. A., Katz, J., Mohabir, V. & Ritvo, P. Individuals with tension and migraine headaches exhibit increased heart rate variability during poststress mindfulness meditation practice but a decrease during a poststress control condition—a randomized, controlled experiment. Int. J. Psychophysiol. 110, 66–74 (2016).
Sakakibara, M., Takeuchi, S. & Hayano, J. Effect of relaxation training on cardiac parasympathetic tone. Psychophysiology 31, 223–228 (1994).
ToroVelasco, C., ArroyoMorales, M., FernándezdelasPeñas, C., Cleland, J. A. & BarreroHernández, F. J. Shortterm effects of manual therapy on heart rate variability, mood state, and pressure pain sensitivity in patients with chronic tensiontype headache: a pilot study. J. Manipulat. Physiol. Ther. 32, 527–535 (2009).
Léonard, A., Clément, S., Kuo, C.D. & Manto, M. Changes in heart rate variability during heartfulness meditation: a power spectral analysis including the residual spectrum. Front. Cardiovasc. Med. 6, 62 (2019).
Phongsuphap, S., Pongsupap, Y., Chandanamattha, P. & Lursinsap, C. Changes in heart rate variability during concentration meditation. Int. J. Cardiol. 130, 481–484 (2008).
Cysarz, D. & Büssing, A. Cardiorespiratory synchronization during Zen meditation. Eur. J. Appl. Physiol. 95, 88–95 (2005).
DelgadoPastor, L. C., Perakakis, P., Subramanya, P., Telles, S. & Vila, J. Mindfulness (Vipassana) meditation: Effects on P3b eventrelated potential and heart rate variability. Int. J. Psychophysiol. 90, 207–214 (2013).
Bernardi, L. et al. Effects of controlled breathing, mental activity and mental stress with or without verbalization on heart rate variability. J. Am. Coll. Cardiol. 35, 1462–1469 (2000).
Tiinanen, S., Tulppo, M. & Seppänen, T. RSA component extraction from heart rate signal by independent component analysis. In 36th Annual Computers in Cardiology Conference (CinC) 161–164 (IEEE, 2009).
Tiinanen, S., Tulppo, M. & Seppänen, T. Reducing the effect of respiration in baroreflex sensitivity estimation with adaptive filtering. IEEE Trans. Biomed. Eng. 55, 51–59 (2008).
Choi, J. & GutierrezOsuna, R. Removal of respiratory influences from heart rate variability in stress monitoring. IEEE Sens. J. 11, 2649–2656 (2011).
Widjaja, D., Caicedo, A., Vlemincx, E., Van Diest, I. & Van Huffel, S. Separation of respiratory influences from the tachogram: A methodological evaluation. PLoS One 9, 101713 (2014).
Varon, C. et al. Unconstrained estimation of HRV indices after removing respiratory influences from heart rate. IEEE J. Biomed. Health Inform. 23, 2386–2397 (2019).
Kontaxis, S., Lázaro, J., Gil, E., Laguna, P. & Bailón, R. Assessment of quadratic nonlinear cardiorespiratory couplings during tilttable test by means of real wavelet biphase. IEEE Trans. Biomed. Eng. 66, 187–198 (2019).
Lenis, G. et al. Separating the effect of respiration on the heart rate variability using Granger’s causality and linear filtering. Biomed. Signal Process. Control 31, 272–287 (2017).
Pan, J. & Tompkins, W. J. A realtime QRS detection algorithm. IEEE Trans. Biomed. Eng. 32, 230–236 (1985).
Mateo, J. & Laguna, P. Analysis of heart rate variability in the presence of ectopic beats using the heart timing signal. IEEE Trans. Biomed. Eng. 50, 334–343 (2003).
Varon, C. et al. A comparative study of ECGderived respiration in ambulatory monitoring using the singlelead ECG. Sci. Rep. 10, 1–14 (2020).
Lazaro, J., Alcaine, A., Gil, E., Laguna, P. & Bailon, R. Electrocardiogram derived respiration from QRS slopes. In Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS 3913–3916 (2013). https://doi.org/10.1109/EMBC.2013.6610400.
Lazaro, J. et al. Electrocardiogram derived respiratory rate using a wearable armband. IEEE Trans. Biomed. Eng. 20, 1. https://doi.org/10.1109/tbme.2020.3004730 (2020).
Park, C. & Lee, B. Realtime estimation of respiratory rate from a photoplethysmogram using an adaptive lattice notch filter. Biomed. Eng. Online 13, 170 (2014).
Park, C., Shin, H. & Lee, B. Blockwise PPG enhancement based on timevariant zerophase harmonic notch filtering. Sensors (Switzerland) 17, 25 (2017).
Hernando, A. et al. Inclusion of respiratory frequency information in heart rate variability analysis for stress assessment. IEEE J. Biomed. Health Inform. 20, 1016–1025 (2016).
Shui, P.L., Shang, H.Y. & Zhao, Y.B. Instantaneous frequency estimation based on directionally smoothed pseudoWignerVille distribution bank. IET Radar. Sonar Navig. 1, 317 (2007).
Mitra, S. Digital Signal Processing: A ComputerBased Approach 4th edn. (McgrawHill, 2011).
Aydore, S., Pantazis, D. & Leahy, R. M. A note on the phase locking value and its properties. Neuroimage 74, 231–244 (2013).
Pitzalis, M. V. et al. Effect of respiratory rate on the relationships between RR interval and systolic blood pressure fluctuations: A frequencydependent phenomenon. Cardiovasc. Res. 38, 332–339 (1998).
Malik, M. Heart rate variability. Ann. Noninvasive Electrocardiol. 1, 151–181 (1996).
Schmidt, P., Reiss, A., Duerichen, R., Marberger, C. & Van Laerhoven, K. Introducing WESAD, a multimodal dataset for wearable stress and affect detection. In Proceedings of the 20th ACM International Conference on Multimodal Interaction 400–408 (ACM, 2018).
Bernardi, L. et al. Effect of breathing rate on oxygen saturation and exercise performance in chronic heart failure. Lancet 351, 1308–1311 (1998).
Grossman, E., Grossman, A., Schein, M. H., Zimlichman, R. & Gavish, B. Breathingcontrol lowers blood pressure. J. Hum. Hypertens. 15, 263–269 (2001).
Singh, Y., Sharma, R. & Talwar, A. Immediate and longterm effects of meditation on acute stress reactivity, cognitive functions, and intelligence. Altern. Ther. Health 18, 46–53 (2012).
Oneda, B., Ortega, K. C., Gusmão, J. L., Araújo, T. G. & Mion, D. Sympathetic nerve activity is decreased during deviceguided slow breathing. Hypertens. Res. 33, 708–712 (2010).
Harada, D. et al. Slow and deep respiration suppresses steadystate sympathetic nerve activity in patients with chronic heart failure: from modeling to clinical application. Am. J. Physiol. Heart Circ. Physiol. 307, H1159–H1168 (2014).
Fonkoue, I. T. et al. Acute effects of deviceguided slow breathing on sympathetic nerve activity and baroreflex sensitivity in posttraumatic stress disorder. Am. J. Physiol. Heart Circ. Physiol. 315, H141–H149 (2018).
Kollai, M. & Mizsei, G. Respiratory sinus arrhythmia is a limited measure of cardiac parasympathetic control in man. J. Physiol. 424, 329–342 (1990).
Widrow, B. et al. Adaptive noise cancelling: Principles and applications. Proc. IEEE 63, 1692–1716 (1975).
Acknowledgements
This research was supported in part by the Korea Medical Device Development Fund Grant funded by the MSIT (the Ministry of Science and ICT, the Ministry of Trade, Industry and Energy, the Ministry of Health & Welfare, the Ministry of Food and Drug Safety) (1711138169, KMDF KMDF_PR_20200901_0100), the Smart HealthCare Program (www.kipot.or.kr) funded by the Korean National Police Agency (KNPA, Korea) (Development and commercialization of police officer’s lifelog acquisition and stress/health management system through artificial intelligence based on big data analysis/220222M03030), the National Research Council of Science and Technology (NST) Grant by the Korea Government (MSIT) (No. CAP18015000), and the Technology Innovation Program (20014477, Development of noncontact AI health monitoring system based on multimodal sensors) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea).
Author information
Authors and Affiliations
Contributions
Conceptualization, I.Y.; methodology, C.P.; data acquisition and analysis, C.P.; interpretation, C.P., I.Y., S.H.; writing original draft preparation, C.P.; supervision, I.Y., S.H. All authors reviewed the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Park, C., Youn, I. & Han, S. Singlelead ECG based autonomic nervous system assessment for meditation monitoring. Sci Rep 12, 22513 (2022). https://doi.org/10.1038/s4159802227121x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s4159802227121x
This article is cited by

A Narrative Review of the Effects of Mindfulness on Sleep and Hypertension
Current Hypertension Reports (2024)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.