Accurate Heart Rate Variability Research Articles

Achieving Real-Time Prediction of Paroxysmal Atrial Fibrillation Onset by Convolutional Neural Network and Sliding Window on R-R Interval Sequences.

Early diagnosis of paroxysmal atrial fibrillation (PAF) could prompt patients to receive timely interventions in clinical practice. Various PAF onset prediction algorithms might benefit from accurate heart rate variability (HRV) analysis and machine learning classification but are challenged by real-time monitoring scenarios. The aim of this study is to present an end-to-end deep learning-based PAFNet model that integrates a sliding window technique on raw R-R intervals of electrocardiogram (ECG) segments to achieve a real-time prediction of PAF onset. This integration enables the deep convolutional neural network (CNN) to be customized as a light-weight architecture that accommodates the size of sliding windows simply by altering the input layer, and specifically its effectiveness in making a new prediction with each new heartbeat. Catering to the potential influence of input sizes, three CNN models were trained using 50, 100, and 200 R-R intervals, respectively. For each model, the performance of the automated algorithms was evaluated for PAF prediction using a ten-fold cross-validation. As a results, a total of 56,381 PAFN-type and 56,900 N-type R-R interval segments were collected from publicly accessible ECG databases, and a promising prediction performance of the automated algorithm with 100 R-R intervals was achieved, with a sensitivity of 97.12%, a specificity of 97.77%, and an accuracy of 97.45%, respectively. Importantly, the automated algorithm with a sliding window step of 1 could process one sample in only 23.1 milliseconds and identify the onset of PAF at least 45 min in advance. The present results suggest that the sliding window technique on raw R-R interval sequences, along with deep learning-based algorithms, may offer the possibility of providing an accurate, real-time, end-to-end clinical tool for mass monitoring of PAF.

Heart Rate Variability Estimation based on RFID Tag-Pair in Dynamic Environments

With the rapid development of smart healthcare, accurate Heart Rate Variability (HRV) estimation for the early detection of diseases has become a hot research topic. Advanced work uses the wireless signal to estimate the heartbeat in a contact-free way, which usually cannot separate multiple users or work in a dynamic environment. In this paper, we propose a lightweight heartbeat-sensing method based on RFID tag pairs, which focuses on HRV extraction in a more general sensing scenario. Based on the tag-pair design, we build a novel heartbeat and respiration model to describe the signal relationship between the two tags from the time and space domains. Based on the model, we propose a Calibrated Temporal-Spatial IQ-Shaping-based signal cancellation algorithm to cancel the respiration and extract the heartbeat. To remove the interference in dynamic measurement, we build an IQ-based signal model via a Principal Component Analysis-based interference estimation. To reduce the statistical error in HRV extraction, we further design a neural network to predict the HRV index. We have implemented a system prototype in a real environment with COTS RFID devices. Extensive experiments show that our system can achieve a median RMSSD error of 7.51ms, which satisfies the medical demand in HRV measurement.

Resampling Strategies and their Influence on Heart Rate Variability Features in Low Sampling Rate Electrocardiogram Data

Heart rate variability (HRV) is a parameter to measure fluctuations in the interval between heartbeats. HRV provides essential insights into the cardiovascular function and autonomic nervous system. Electrocardiograms (ECG) on wearable devices are often recorded at low sampling rates, limiting temporal resolution and information. Resampling is a technique of changing the sampling rate from a high sampling rate to a lower sampling rate and vice versa. This research aims to evaluate the effect of resampling ECG data with a low sampling rate on HRV features. ECG data consists of 50 Hz and 100 Hz sampling rates. Data with a 50 Hz sampling rate is up-sampled up to 100 Hz, while 100 Hz data is down-sampled up to 50 Hz and up-sampled up to 250 Hz using the Fast Fourier Transform Interpolation Method. Upsampling from 50 Hz to 100 Hz shows unsatisfactory results, except for some HRV features such as NN20, pNN20, and CVI. Better results were found when up sampling from 100 Hz up to 250 Hz, with some HRV features showing good concordance values. However, downsampling from 100 Hz up to 50 Hz is unsuitable for HRV feature analysis. To obtain accurate HRV analysis results in all domains, it is highly recommended to use a sampling rate above 100 Hz.

Robust PPG-Based Mental Workload Assessment System Using Wearable Devices.

Heart rate variability (HRV) has been used in assessing mental workload (MW) level. Compared with ECG, photoplethysmogram (PPG) provides convenient in assessing MW with wearable devices, which is more suitable for daily usage. However, PPG collected by smartwatches are prone to suffer from artifacts. Those signal corruptions cause invalid Inter-beat Intervals (IBI), making it challenging to evaluate the HRV feature. Hence, the PPG-based MW assessment system is difficult to obtain a sustainable and reliable assessment of MW. In this paper, we propose a pre- and post- processing technique, called outlier removal and uncertainty estimation, respectively, to reduce the negative influences of invalid IBIs. The proposed method helps to acquire accurate HRV features and evaluate the reliability of incoming IBIs, rejecting possibly misclassified data. We verified our approach in two open datasets, which are CLAS and MAUS. Experiment results show proposed method achieved higher accuracy (66.7% v.s. 74.2%) and lower variance (11.3% v.s. 10.8%) among users, which has comparable performance to an ECG-based MW system.

Measuring heart rate variability using a heart rate monitor in horses (Equus caballus) during groundwork.

Measuring physiological parameters of stress in horses during groundwork, for example when involved in equine-assisted interventions, is important to gain insight into the stress levels of the horses. Heart rate and heart rate variability can be used as physiological indicators of stress in horses. Heart rate monitors could be easily incorporated into practice, as they are not expensive and easy to use. However, it is questionable whether heart rate monitors present accurate heart rate variability results in exercising horses, similar to electrocardiograms. The aim of this study was to determine the accuracy of heart rate monitors for the assessment of heart rate variability in horses during groundwork exercise. Simultaneous telemetric electrocardiograms (Televet) and heart rate monitor (Polar H10 transmitter and M430 receiver, Hylofit electrodes) recordings were performed on 28 horses (15 mares and 13 geldings). Results indicate that the heart rate monitor accurately determined heart rate and time-domain heart rate variability parameters when compared to electrocardiograms during both baseline and groundwork conditions. As expected, heart rate significantly increased and the heart rate variability significantly decreased during groundwork compared to baseline conditions. This indicates that the heart rate monitor can be used to accurately determine heart rate variability during groundwork.

RMSSD Is More Sensitive to Artifacts Than Frequency-Domain Parameters: Implication in Athletes' Monitoring.

Easy-to-use and accurate heart rate variability (HRV) assessments are essential in athletes' follow-up, but artifacts may lead to erroneous analysis. Artifact detection and correction are the purpose of extensive literature and implemented in dedicated analysis programs. However, the effects of number and/or magnitude of artifacts on various time- or frequency-domain parameters remain unclear. The purpose of this study was to assess the effects of artifacts on HRV parameters. Root mean square of the successive differences (RMSSD), standard deviation of the normal to normal inter beat intervals (SDNN), power in the low- (LF) and high-frequency band (HF) were computed from two 4-min RR recordings in 178 participants in both supine and standing positions, respectively. RRs were modified by (1) randomly adding or subtracting 10, 30, 50 or 100 ms to the successive RRs; (2) a single artifact was manually inserted; (3) artifacts were automatically corrected from signal naturally containing artifacts. Finally, RR recordings were analyzed before and after automatic detection-correction of artifacts. Modifying each RR by 10, 30, 50 and 100 ms randomly did not significantly change HRV parameters (range -6%, +6%, supine). In contrast, by adding a single artifact, RMSSD increased by 413% and 269%, SDNN by 54% and 47% in supine and standing positions, respectively. LF and HF changed only between -3% and +8% (supine and standing) in the artifact condition. When more than 0.9% of the signal contained artifacts, RMSSD was significantly biased, whilst when more than 1.4% of the signal contained artifacts LF and HF were significantly biased. RMSSD and SDNN were more sensitive to a single artifact than LF and HF. This indicates that, when using RMSSD only, a single artifact may induce erroneous interpretation of HRV. Therefore, we recommend using both time- and frequency-domain parameters to minimize the errors in the diagnoses of health status or fatigue in athletes.

AFibNet: an implementation of atrial fibrillation detection with convolutional neural network

BackgroundGeneralization model capacity of deep learning (DL) approach for atrial fibrillation (AF) detection remains lacking. It can be seen from previous researches, the DL model formation used only a single frequency sampling of the specific device. Besides, each electrocardiogram (ECG) acquisition dataset produces a different length and sampling frequency to ensure sufficient precision of the R–R intervals to determine the heart rate variability (HRV). An accurate HRV is the gold standard for predicting the AF condition; therefore, a current challenge is to determine whether a DL approach can be used to analyze raw ECG data in a broad range of devices. This paper demonstrates powerful results for end-to-end implementation of AF detection based on a convolutional neural network (AFibNet). The method used a single learning system without considering the variety of signal lengths and frequency samplings. For implementation, the AFibNet is processed with a computational cloud-based DL approach. This study utilized a one-dimension convolutional neural networks (1D-CNNs) model for 11,842 subjects. It was trained and validated with 8232 records based on three datasets and tested with 3610 records based on eight datasets. The predicted results, when compared with the diagnosis results indicated by human practitioners, showed a 99.80% accuracy, sensitivity, and specificity.ResultMeanwhile, when tested using unseen data, the AF detection reaches 98.94% accuracy, 98.97% sensitivity, and 98.97% specificity at a sample period of 0.02 seconds using the DL Cloud System. To improve the confidence of the AFibNet model, it also validated with 18 arrhythmias condition defined as Non-AF-class. Thus, the data is increased from 11,842 to 26,349 instances for three-class, i.e., Normal sinus (N), AF and Non-AF. The result found 96.36% accuracy, 93.65% sensitivity, and 96.92% specificity.ConclusionThese findings demonstrate that the proposed approach can use unknown data to derive feature maps and reliably detect the AF periods. We have found that our cloud-DL system is suitable for practical deployment

The minimal sampling frequency of the photoplethysmogram for accurate pulse rate variability parameters in healthy volunteers

ObjectiveToday mobile health-monitoring devices calculate heart rate and its variability mostly from the photoplethysmogram (PPG). Minimizing the power consumption is crucial, one option is optimizing the signal sampling rate. Present study aimed to determine the minimal sampling frequency of the PPG signal, which is sufficient for accurate heart rate variability (HRV) analysis. Methods57 high-quality, 5-minute PPG signals from healthy volunteers sampled at 1 kHz (master) were decimated by a factor of 2, 5, 10, 20, 50, 100, 200, 500, then cubic spline and parabola interpolated back to 1 ms resolution. The mean pulse rate, its standard deviation (SDNN), root mean square of successive RR-differences (RMSSD), Porta and Guzik indices (PI, GI) were calculated. Their relative accuracy error (RAE) was determined, RAE<5% was acceptable. Also, the processing times were measured. Results200ms sampling interval without interpolation is sufficient to calculate mean pulse rate with RAE<0.05%. SDNN and RMSSD require at least 20 ms sampling interval without interpolation (RAE: 1.28±0.96% and 4.25±3.59%, respectively). By both interpolations, these sampling intervals can be increased to 100 ms and 50 ms, respectively. Also, the accuracy of GI and PI can be improved by interpolation, here parabola approximation was better (GI: 20 ms versus 100 ms, PI: 50 ms versus 100 ms). Parabola approximation needs significantly less computation than cubic spline interpolation. ConclusionsFor monitoring the average heart rate, 5 Hz sampling frequency can be sufficient without interpolation in healthy subjects. Correct HRV analysis requires higher sampling rates depending on the parameter. Interpolation can improve HRV accuracy from lower temporal resolution PPGs.

Outliers detection for accurate HRV-seizure baseline estimation using modern numerical algorithms

Due to inevitable measurement artifacts, accurate baseline estimation remains an important issue in modern heart rate variability (HRV) analysis with partial epilepsy signals. To detect seizures, the HRV is commonly analyzed as a quasi stationary signal, which is correlated with neuroautonomic activity and considered to be noninvasive. A sudden seizure induces outliers and disturbances to the normal baseline that makes it difficult to provide accurate HRV estimation and outliers classification using the traditional boxplot methodology. A standard strategy to detect outliers implies computing the residuals for the estimated baseline and setting thresholds to extract the first and third quartiles from a histogram. In this work, we analyze modern numerical algorithms developed for HRV-seizure baseline estimation. We also propose a new iterative unbiased finite impulse response (I-UFIR) smoothing algorithm developed for colored measurement noise (CMN) using the backward Euler-based state-space model and show its advantages and shortcomings. A comparison of the estimation algorithms is provided using simulated synthetic and real data. It is demonstrated that the I-UFIR smoother is highly robust against sharp HRV changes that allows removing the outliers with a minimum loss in information. It is also shown that the time-frequency analysis allows analyzing accurately the HRV frequencies, provided that the outliers are removed. All methods are tested by partial seizures records taken from patients during continuous electroencephalography, electrocardiography, and video monitoring.

R-R Interval Estimation for Wearable Electrocardiogram Based on Single Complex Wavelet Filtering and Morphology-Based Peak Selection

Recent innovations in wearable electrocardiogram (ECG) devices have enabled various personal healthcare applications based on heart rate variability (HRV). However, wearable ECGs rarely undergo visual inspection by medical experts, hence may contain noise and artifacts. Because apparent changes in the recorded ECGs caused by noise and artifacts may hamper the extraction of QRS complexes, an R-R interval (RRI) estimation algorithm tolerant to these measurement faults is required as the initial step toward HRV analysis using wearable ECGs. This paper proposes a semi-real-time RRI estimation for wearable ECGs utilizing a two-stage structure. In the preprocessing stage, we use a complex-valued wavelet that can adaptively fit to morphological variations of the QRS complex while retaining computing resources for extracting the QRS complex features. In the decision stage, we make use of complex-valued features and select appropriate QRS complexes in consideration of three features: peak magnitude, peak location, and peak morphology (phase). Initial evaluations show that the QRS complex detection performance of the proposed method achieved the F1 score of 0.952 ± 0.040 when targeting pseudo ECG data created from open data assuming wearable ECGs, and of 0.986 ± 0.018 when targeting actual ECG data recorded by a shirt-type wearable ECG device during an exercise activity. Furthermore, the proposed method was able to suppress overlook or misdetection of QRS complexes, so the obtained RRIs are closer to the reference RRIs. The proposed method therefore contributes to achieving accurate HRV analysis using wearable ECGs in terms of obtaining accurate RRIs.

Detecting stress from imaging photoplethysmography using high frame rate video and a yellow-green filter: A pilot study

We investigate the use of a yellow-green filter to increase the signal-to-noise ratio (snr) in imaging photoplethysmography (iPPG) and test if high frame rate (HFR) video improves the accuracy of the derived heart rate variability (HRV). This pilot study is associated with a broader program to use iPPG to detect and monitor stress levels using HRV. To improve the snr of the iPPG signal, we employ two HFR colour video cameras of which one was fitted with a yellow-green filter (corresponding to the haemoglobin absorption peak within the visible spectrum). To our knowledge, the benefit of a yellow-green filter has never been explored. The predominant influence on HRV comes from the autonomic nervous system (ANS), which connects directly to the heart and cues the human body to relax or to stress. The linkage of HRV to the ANS makes HRV a proxy for stress levels. The HRV is derived from the iPPG signal by first using a cubic spline interpolation for more precise peak detection, and then calculating the inter-beat intervals from the peak-to-peak time differences. Instead of interpolating the signal, we hypothesise that a more accurate HRV measurement can be obtained using a HFR video camera, in our case at 200 frames per second.

Validation of an equine fitness tracker: heart rate and heart rate variability

The adoption of fitness tracker devices to monitor training in the equine market is in full expansion. However, the validity of most of these devices has not been assessed. The aim of this study was to examine the validity of heart rate (HR) and heart rate variability (HRV) measurements during high-intensity exercise by an integrated equine fitness tracker with an electrocardiogram (ECG) (Equimetre) in comparison to an ECG device (Televet). Twenty Thoroughbred racehorses were equipped with the two devices and completed a training session at the track. Data from 18 horses was readable to be analysed. Equimetre HR was compared to Televet HR derived from the corrected Televet ECG. HRV parameters were computed in a dedicated software (Kubios) on uncorrected and manually corrected ECG from both devices, and compared to the Televet corrected data. The HR was recorded on the entire training session and HRV parameters were calculated during the exercise and recovery periods. A strong correlation between the Equimetre HR and Televet HR on corrected data was found (Pearson correlation: r=0.992, P&lt;0.001; root mean square error = 4.06 bpm). For HRV, the correlation was good for all parameters when comparing corrected Equimetre to corrected Televet data (Lin’s coefficient = 0.998). When comparing data obtained from uncorrected Equimetre data to the corrected Televet data, the correlation for HR was still good (Lin’s coefficient = 0.995) but the correlation for all HRV parameters was poor, except for the triangular index (Lin’s coefficient = 0.995). However, correlation between the uncorrected Televet HRV data and the corrected Televet data was equally poor (Lin’s coefficient &lt;0.9). In conclusion, the integrated equine fitness tracker Equimetre satisfies validity criteria for HR monitoring in horses during high intensity exercise. When using corrected ECG data, it provides accurate HRV parameters as well.

Effect of Motion Artifact on Variation in Heart Rate Variability Parameters

The heart rate variability (HRV) is a noninvasive way properly for investigating the activity of the autonomic nervous system (ANS) as well as to predict cardiovascular diseases. To guarantee an accurate HRV analysis, a motion artifact-free HRV recording must be obtained. However, complete removal of a motion artifact is impossible when measuring heartbeats for 5 min, and the motion artifact due to sudden ANS activity must be taken into consideration for the HRV parameters. And, the ANS balance has thus far been evaluated by each individual HRV parameter calculated for a single 5 min HRV segment, leading to the dynamic activity of the ANS within the same period being ignored. Therefore, to resolve this problem, HRV parameters for ultra-short-term segments that are short enough to reflect a sudden motion artifact must be analyzed. The aim of the present study was to evaluate the effects of a motion artifact on the variation in HRV parameters to provide detailed information on ANS activity. The 121 ultra-short-term HRV segments were created by moving a 1-min window forward by a time shift interval of 2 s for the entire 5 min HRV segment. The ratios of Ln LF to Ln HF in these ultra-short-term segments and a single 5 min segment with a motion artifact were 0.89 and 1.06, respectively, while those in a motion artifact-free HRV segment were 0.75 and 0.93, respectively. This variation test for a short-term motion artifact and motion artifact-free HRV dataset was found to affect the SDNN (7.73 and 2.68), SD2 (11.44 and 4.42), TINN (40.33 and 9.92), and Ln HF (0.37 and 0.13) the most in terms of the standard deviation, respectively. Taken together, the mean HRV parameters of many ultra-short-term segments might play an important role in evaluating dynamic ANS activities within a short-term segment, avoiding the false conclusions made by the traditional HRV analysis.

On the Minimal Adequate Sampling Frequency of the Photoplethysmogram for Pulse Rate Monitoring and Heart Rate Variability Analysis in Mobile and Wearable Technology

Abstract Recently there has been great interest in photoplethysmogram signal processing. However, its minimally necessary sampling frequency for accurate heart rate variability parameters is ambiguous. In the present paper frequency-modulated 1.067 Hz cosine wave modelled the variable PPG in silico. The five-minute-long, 1 ms resolution master-signals were decimated (D) at 2-500 ms, then cubic spline interpolated (I) back to 1 ms resolution. The mean pulse rate, standard deviation, root mean square of successive pulse rate differences (RMSSD), and spectral components were computed by Varian 2.3 and compared to the master-series via relative accuracy error. Also Poincaré-plot morphology was assessed. Mean pulse rate is accurate down to 303 ms (D) and 400 ms (I). In low-variability series standard deviation required at least 5 ms (D) and 100 ms (I). RMSSD needed 10 ms (D), and 303 ms (I) in normal, whereas 2 ms (D) and 100 ms (I) in low- variability series. In the frequency domain 5 ms (D) and 100 ms (I) are required. 2 ms (D) and 100 ms (I) preserved the Poincaré-plot morphology. The minimal sampling frequency of PPG for accurate HRV analysis is higher than expected from the signal bandwidth and sampling theorem. Interpolation improves accuracy. The ratio of sampling error and expected variability should be considered besides the inherent sensitivity of the given parameter, the interpolation technique, and the pulse rate detection method.

A Novel Algorithm for HRV Estimation from Short-Term Acoustic Recordings at Neck.

Heart rate variability (HRV) is an important noninvasive parameter to monitor the activity of the autonomic nervous system. This paper proposes an algorithm to analyze HRV by processing the acoustic data, recorded by placing a small, wearable sensor on the suprasternal notch (at neck) of an adult subject, primarily intended to record breathing sounds. The method used an empirical data analysis approach of the Hilbert-Huang transform (HHT) to construct an instantaneous energy envelope and segment the cardiac cycle by detecting S1 and S2 sounds using the K-means algorithm. The time-domain HRV analysis for the short-term recordings of 10 subjects demonstrated a close agreement with the reference ECG signal. The instantaneous heart rate (IHR) comparisons yielded an accuracy of 95.78% and 92.35% for S1 and S2 sounds respectively. The experimental results showed that the proposed algorithm can provide an accurate HRV analysis for the cardiac signals recorded at the neck.

Robust heart rate estimation using combined ECG and PPG signal processing

Heart rate variability (HRV) from recorded electrocardiograms (ECG) is a well-known diagnostic method for the assessment of autonomic nervous function of the heart, which is widely used to predict clinically relevant outcomes in the critical care setting, to risk stratify patients, and predict outcomes such as mortality. The morphological variations in the ECG waveform and the high degree of heterogeneity in the QRS complex often make it difficult to identify R waves, which may preclude the accurate analysis for HRV. Photoplethysmographic (PPG) signal can provide information about both the cardiovascular and respiratory systems and have extremely high degree of correlation with ECG during cardiac cycle. In this paper, we developed robust algorithm for high-resolution inter-beat waveform extraction using combined ECG and PPG analysis, which is highly needed for accurate HRV estimation. The simulation results showed high performance for inter-beat waveform detection in different cases that identifies missing/extra peaks in the QRS detection algorithm.

Reliability of the Parabola Approximation Method in Heart Rate Variability Analysis Using Low-Sampling-Rate Photoplethysmography.

Photoplethysmographic signals are useful for heart rate variability analysis in practical ambulatory applications. While reducing the sampling rate of signals is an important consideration for modern wearable devices that enable 24/7 continuous monitoring, there have not been many studies that have investigated how to compensate the low timing resolution of low-sampling-rate signals for accurate heart rate variability analysis. In this study, we utilized the parabola approximation method and measured it against the conventional cubic spline interpolation method for the time, frequency, and nonlinear domain variables of heart rate variability. For each parameter, the intra-class correlation, standard error of measurement, Bland-Altman 95% limits of agreement and root mean squared relative error were presented. Also, elapsed time taken to compute each interpolation algorithm was investigated. The results indicated that parabola approximation is a simple, fast, and accurate algorithm-based method for compensating the low timing resolution of pulse beat intervals. In addition, the method showed comparable performance with the conventional cubic spline interpolation method. Even though the absolute value of the heart rate variability variables calculated using a signal sampled at 20Hz were not exactly matched with those calculated using a reference signal sampled at 250Hz, the parabola approximation method remains a good interpolation method for assessing trends in HRV measurements for low-power wearable applications.

Minimal Window Duration for Accurate HRV Recording in Athletes.

Heart rate variability (HRV) is non-invasive and commonly used for monitoring responses to training loads, fitness, or overreaching in athletes. Yet, the recording duration for a series of RR-intervals varies from 1 to 15 min in the literature. The aim of the present work was to assess the minimum record duration to obtain reliable HRV results. RR-intervals from 159 orthostatic tests (7 min supine, SU, followed by 6 min standing, ST) were analyzed. Reference windows were 4 min in SU (min 3–7) and 4 min in ST (min 9–13). Those windows were subsequently divided and the analyses were repeated on eight different fractioned windows: the first min (0–1), the second min (1–2), the third min (2–3), the fourth min (3–4), the first 2 min (0–2), the last 2 min (2–4), the first 3 min (0–3), and the last 3 min (1–4). Correlation and Bland & Altman statistical analyses were systematically performed. The analysis window could be shortened to 0–2 instead of 0–4 for RMSSD only, whereas the 4-min window was necessary for LF and total power. Since there is a need for 1 min of baseline to obtain a steady signal prior the analysis window, we conclude that studies relying on RMSSD may shorten the windows to 3 min (= 1+2) in SU or seated position only and to 6 min (= 1+2 min SU plus 1+2 min ST) if there is an orthostatic test. Studies relying on time- and frequency-domain parameters need a minimum of 5 min (= 1+4) min SU or seated position only but require 10 min (= 1+4 min SU plus 1+4 min ST) for the orthostatic test.

Reliability evaluation of R-R interval measurement status for time domain heart rate variability analysis with wearable ECG devices.

Electrocardiograms (ECGs) captured by wearable ECG devices easily contain artifacts because of measurement faults. Since the frequency characteristics of artifacts are quite similar to those of R waves, it may result in R-R interval (RRI) miscalculations. To enable accurate heart rate variability (HRV) analysis in daily life, this paper proposes a method to reliably evaluate RRI measurement status, one that uses the electric potential characteristics of the QRS complex. Initial evaluation results show that it has the potential to distinguish miscalculated RRIs. Also proposed is a new RRI outlier exclusion method that uses the aforementioned method. Time domain measures of HRV derived from the proposed RRI outlier exclusion method are found to be more accurate than those derived from the conventional one.

Contactless heart rate variability measurement by IR and 3D depth sensors with respiratory sinus arrhythmia

Abstract: Heart rate variability (HRV) is known to be correlated with emotional arousal, cognitive depletion, and health status. Despite the accurate HRV detection by various body-attached sensors, a contactless method is desirable for the HCI purposes. In this research, we propose a non-invasive contactless HRV measurement by Microsoft Kinect 2 sensor with Respiratory Sinus Arrhythmia (RSA) correction. The Infrared and RGB cameras are used to measure the heart rate signal, and its 3D Depth sensor is employed to capture the human respiratory signal to correct the initially calculated HRV with RSA. The correlation analysis among the calculated HRVs by different methods and devices showed a significant improvement in reliable HRV measurements. This study enlightens the researchers and developers to choose a proper method for HRV calculations based on their required accuracy and application.