Pulse Rate Variability Analysis Using Remote Photoplethysmography Signals

Objective. Monitoring changes in human heart rate variability (HRV) holds significant importance for protecting life and health. Studies have shown that Imaging Photoplethysmography (IPPG) based on ordinary color cameras can detect the color change of the skin pixel caused by cardiopulmonary system. Most researchers employed deep learning IPPG algorithms to extract the blood volume pulse (BVP) signal, analyzing it predominantly through the heart rate (HR). However, this approach often overlooks the inherent intricate time-frequency domain characteristics in the BVP signal, which cannot be comprehensively deduced solely from HR. The analysis of HRV metrics through the BVP signal is imperative. Approach. In this paper, the transformation invariant loss function with distance equilibrium (TIDLE) loss function is applied to IPPG for the first time, and the details of BVP signal can be recovered better. In detail, TIDLE is tested in four commonly used IPPG deep learning models, which are DeepPhys, EfficientPhys, Physnet and TS_CAN, and compared with other three loss functions, which are mean absolute error (MAE), mean square error (MSE), Neg Pearson Coefficient correlation (NPCC). Main results. The experiments demonstrate that MAE and MSE exhibit suboptimal performance in predicting LF/HF across the four models, achieving the Statistic of Mean Absolute Error (MAES) of 25.94% and 34.05%, respectively. In contrast, NPCC and TIDLE yielded more favorable results at 13.51% and 11.35%, respectively. Taking into consideration the morphological characteristics of the BVP signal, on the two optimal models for predicting HRV metrics, namely DeepPhys and TS_CAN, the Pearson coefficients for the BVP signals predicted by TIDLE in comparison to the gold-standard BVP signals achieved values of 0.627 and 0.605, respectively. In contrast, the results based on NPCC were notably lower, at only 0.545 and 0.533, respectively. Significance. This paper contributes significantly to the effective restoration of the morphology and frequency domain characteristics of the BVP signal.

This paper explores a novel method for anxiety detection in older adults using simple wristband sensors such as electrodermal activity (EDA) and photoplethysmogram (PPG) and a context-based feature. The proposed method for anxiety detection combines features from a single physiological signal with an experimental context-based feature to improve the performance of the anxiety detection model. The experimental data for this work are obtained from a year-long experiment on 41 healthy older adults (26 females and 15 males) in the age range 60–80 with mean age \(73.36 \pm 5.25\) during a Trier Social Stress Test (TSST) protocol. The anxiety level ground truth was obtained from State–Trait Anxiety Inventory (STAI), which is regarded as the gold standard to measure perceived anxiety. EDA and blood volume pulse (BVP) signals were recorded using a wrist-worn EDA and PPG sensor, respectively. 47 features were computed from EDA and BVP signal, out of which a final set of 24 significantly correlated features were selected for analysis. The phases of the experimental study are encoded as unique integers to generate the context feature vector. A combination of features from a single sensor with the context feature vector is used for training a machine learning model to distinguish between anxious and not-anxious states. Results and analysis showed that the EDA and BVP machine learning models that combined the context feature along with the physiological features achieved 3.37% and 6.41% higher accuracy, respectively, than the models that used only physiological features. Further, end-to-end processing of EDA and BVP signals was simulated for real-time anxiety level detection. This work demonstrates the practicality of the proposed anxiety detection method in facilitating long-term monitoring of anxiety in older adults using low-cost consumer devices.

Blood volume pulse (BVP) measurements obtained through the noninvasive photoplethysmography technique are often analyzed to derive physiological variables such as heart rate (HR) and HR variability, blood pressure, and respiratory rate. Physical activity is one of the predominant factors inducing changes in the BVP signal. The goal of the present study is to develop a mathematical model to characterize the exercise-induced changes in the BVP signal. A two-step procedure is developed that first models the HR using specified exercise intensity information, and then characterizes the change in the BVP waveform based on the estimated HR. The HR model parameters are identified using experiments involving 18 subjects participating in bouts of running and cycling exercise. Speed and incline for the treadmill and cycling power for the stationary bike are inputs to the HR model. The mean (standard deviation) for the root-mean-square error of the HR predictions are 16.1 (4.0) and 12.47 (4.6) BPM during running and cycling, respectively. The HR is transformed to the BVP signal through a model developed based on log-normal functions where model parameters change due to HR variability.

Pulse rate variability (PRV) can be extracted from functional near-infrared spectroscopy (fNIRS) (PRV(NIRS)) and photoplethysmography (PPG) (PRV(PPG)) signals. The present study compared the accuracy of simultaneously acquired PRV(NIRS) and PRV(PPG), and evaluated their different characterizations of the sympathetic (SNS) and parasympathetic (PSNS) autonomous nervous system activity. Ten healthy subjects were recorded during resting-state (RS) and respiratory challenges in two temperature conditions, i.e., room temperature (23°C) and cold temperature (4°C). PRV(NIRS) was recorded based on fNIRS measurement on the head, whereas PRV(PPG) was determined based on PPG measured at the finger. Accuracy between PRV(NIRS) and PRV(PPG), as assessed by cross-covariance and cross-sample entropy, demonstrated a high degree of correlation (r > 0.9), which was significantly reduced by respiration and cold temperature. Characterization of SNS and PSNS using frequency-domain, time-domain, and nonlinear methods showed that PRV(NIRS) provided significantly better information on increasing PSNS activity in response to respiration and cold temperature than PRV(PPG). The findings show that PRV(NIRS) may outperform PRV(PPG) under conditions in which respiration and temperature changes are present, and may, therefore, be advantageous in research and clinical settings, especially if characterization of the autonomous nervous system is desired.

Prophylactic treatment has been proved to reduce hypotension incidence after spinal anesthesia during cesarean labor. However, the use of pharmacological prophylaxis could carry out undesirable side-effects on mother and fetus. Thus, the prediction of hypotension becomes an important challenge. Hypotension events are hypothesized to be related to a malfunctioning of autonomic nervous system (ANS) regulation of blood pressure. In this work, ANS responses to positional changes of 51 pregnant women programmed for a cesarean labor were explored for hypotension prediction. Lateral and supine decubitus, and sitting position were considered while electrocardiographic and pulse photoplethysmographic signals were recorded. Features based on heart rate variability, pulse rate variability (PRV) and pulse transit time (PTT) analysis were used in a logistic regression classifier. The results showed that PRV irregularity changes, assessed by approximate entropy, from supine to lateral decubitus, and standard deviation of PTT in supine decubitus were found as the combination of features that achieved the best classification results sensitivity of 76%, specificity of 70% and accuracy of 72%, being normotensive the positive class. Peripheral regulation and blood pressure changes, measured by PRV and PTT analysis, could help to predict hypotension events reducing prophylactic side-effects in the low-risk population.

In this study, a new method for detecting emotions using Blood Volume Pulse (BVP) signals and machine learning was presented. The BVP of 30 subjects from the publicly available CASE dataset was pre-processed, and 39 features were extracted from various emotional states, such as amusing, boring, relaxing, and scary. The features were categorized into time, frequency, and time-frequency domains and used to build an emotion detection model with XGBoost. The model achieved the highest classification accuracy of 71.88% using the top 10 features. The most significant features of the model were computed from time (5 features), time-frequency (4 features), and frequency (1 feature) domains. The skewness calculated from the time-frequency representation of the BVP was ranked highest and played a crucial role in the classification. Our study suggests the potential of using BVP recorded from wearable devices to detect emotions in healthcare applications.

ObjectivesRemote photoplethysmography (rPPG) is a promising non-contact measurement technique for assessing numerous physiological parameters: pulse rate, pulse rate variability (PRV), respiratory rate, pulse wave velocity, blood saturation, blood pressure, etc. To justify its use in ultra-short-term (UST) PRV analysis, which is of great benefit for several healthcare applications, the agreement between rPPG- and PPG-derived UST-PRV metrics was studied.ApproachThree time-domain metrics—standard deviation of normal-to-normal (NN) intervals (SDNN), root mean square of successive NN interval differences (RMSSD), and the percentage of adjacent NN intervals that differ from each other by more than 50 ms (pNN50)—were extracted from 56 video recordings in a publicly available data set. The selected metrics were calculated on the basis of three groups of 10 s recordings and their average, two groups of 30 s recordings and their average, and a group of 60 s recordings taken from the full-length recordings and then compared with metrics derived from the corresponding reference (PPG) pulse waveform signals by using correlation and effect size parameters, and Bland–Altman plots.Main resultsThe results show there is stronger agreement as the recording length increases for SDNN and RMSSD, yet there is no significant change for pNN50. The agreement parameters reach r = 0.841 (p < 0.001), r = 0.529 (p < 0.001), and r = 0.657 (p < 0.001), estimated median bias −1.52, −2.28 ms and −1.95% and a small effect size for SDNN, RMSSD, and pNN50 derived from the 60 s recordings, respectively.SignificanceRemote photoplethysmography-derived UST-PRV metrics manage to capture UST-PRV metrics derived from reference (PPG) recordings well. This feature is highly desirable in numerous applications for the assessment of one’s health and well-being. In future research, the validity of rPPG-derived UST-PRV metrics compared to the gold standard electrocardiography recordings is to be assessed.

Heart rate variability (HRV) is a noninvasive marker for the cardiovascular system and the substantial influence of this system on the autonomous nervous system (ANS). Contact and remote photoplethysmography (PPG) proved to be a reliable means of pulse rate variability (PRV) measurement as in wearable devices or imaging PPG using video cameras with evolved technical specifications. However, PRV effectiveness remains unclear during controlled breathing rate. Specifically, respiratory rate, among other physiological factors, applies profound effects on ANS balance. In this study, we demonstrated that PRV can be a reliable marker, as HRV, during various controlled breathing rates. We evaluated the proposed PRV monitoring study against reference HRV obtained from an ECG measurement device. PRV was measured from video streams captured using a smartphone and a webcam for the upper abdominal section of healthy volunteers while in standing position and performing six different respiratory maneuvers. We evaluated frequency and nonlinear HRV indices. Agreement between all indices was tested for camera-based devices and ground truth for all performed tasks. The agreement presented a consistent mean close to 95% limit of agreement (LoA). Also, we employed spectral analysis and measured the spectral divergences, normalized root mean squared error (NRMSE), and correlation. Kullback-Leibler Divergence median and percentiles results indicate congruent spectrums, where the maximum median was 0.18. A significant spectral correlation with minimum value of 0.98, and relatively low spectrum NRMSE of 0.24 were observed. Moreover, the results led to HRV physiological related phenomena which agreed with physiologically interpretations found in the literature.

This study addresses the question whether galvanic skin response (GSR) and blood volume pulse (BVP) of untrained and unaided observers can be used to identify real and posed smiles from different sets of smile videos or smile images. Observers were shown smile face videos/images, either singly or paired, with the intention to recognise each viewed as real or posed smiles. We created four experimental situations, namely single images (SI), single videos (SV), paired images (PI), and paired videos (PV). The GSR and BVP signals were recorded and processed. Our machine learning classifiers reached the highest accuracy of 93.3%, 87.6%, 92.0%, 91.7% for PV, PI, SV, and SI respectively. Finally, PV and SI were found to be the easiest and hardest way to identify real and posed smiles respectively. Overall, we demonstrated that observers’ subconscious physiological signals (GSR and BVP) are able to identify real and posed smiles at a good accuracy.KeywordsPhysiological signalsAffective computingClassificationMachine learningSmile

Remote photoplethysmography (rPPG) is a kind of video-based heart rate (HR) estimation technique which has widely potential applications in health monitoring and human- computer interaction. However, the accuracy of conventional rPPG methods is easily disturbed by motion and illumination artifacts. Recently, some deep learning based rPPG methods have attracted many attentions due to its good performance and robustness to noise. This paper proposes a new rPPG scheme using a convolutional neural network (CNN) to map the pulse accumulated image to corresponding true heart rate value, where the spatial-temporal input images are constructed with raw pulses from conventional rPPG methods. In order to check the feasibility and ideal performance of this method, synthetic rPPG pulses are built using real electrocardiograph (ECG) or blood volume pulse (BVP) signals through a modified Akima cubic Hermite interpolation. We test the proposed method in three cases, subject dependent, subject independent, and also a cross-dataset one. The experimental results show that our method performs well in heart rate value estimation with synthetic rPPG pulses even for the cross-dataset case (mean absolute error HR mae = 4.36 BPM, root mean square error HR rmse = 6.26 BPM, mean error rate percentage HR mer = 5.46%). This pilot study verifies the feasibility of the proposed method and provides a solid foundation for the follow-up research with real rPPG pulses.

Non-contact heart rate and heart rate variability measurements have applications in healthcare and affective computing. Recently, a system utilizing a five-band camera (RGBCO: red, green, blue, cyan, orange) was proposed, and shown to improve both remote measurement of heart rate and heart rate variability over an RGB camera. In this paper, we propose an improved method for video-based estimation of heart rate variability. We introduce three advancements over previous work utilizing five-band cameras: (1) an adaptive non-rectangular region of interest identified using automatically detected facial feature points, (2) improved peak detection within the blood volume pulse (BVP) signal, and (3) improved HRV calculation using the Welch periodogram. We apply our method to a test dataset of subjects at rest and under cognitive stress and show qualitative improvements in the stability of HRV spectrogram estimation. Although we evaluate our method using a five-band camera, the method could be applied to video recorded with any camera.

Remote photoplethysmography (rPPG) has shown great potential for heart rate variability (HRV) analysis due to its non-contact and convenient nature. However, HRV measurement relies on long-term and high-quality blood volume pulse (BVP) signals, which pose substantial challenges for existing rPPG techniques. To address this, we propose HRVFusion, a conditional diffusion model-based framework for HRV estimation from rPPG signals. In the forward diffusion process, raw chrominance (CHROM) signals extracted from a video are incorporated to simulate the complex noise disturbances typically observed in rPPG signals. During the reverse process, to enhance the model's sensitivity to HRVs, the high-precision BVP reconstruction is jointly guided using both the chrominance signal and its time-frequency ridge extracted via the wavelet synchrosqueezed transform (WSST). Furthermore, a convolution and mamba hybrid network is introduced to handle inputs of arbitrary length, meeting the requirements for long-term HRV analysis. A public dataset for long-term video-based HRV extraction, consisting of 15-minute recordings from 18 participants, is presented for the first time. Experimental results demonstrate that the proposed method achieves superior performance across multiple HRV metrics, with a mean absolute error (MAE) of 5-8 ms for SDNN, RMSSD below 18 ms, and Pearson correlation coefficients above 0.85 for frequency-domain indices (LF, HF, LF/HF), outperforming existing approaches. This study provides an accurate and robust non-contact solution for long-term video-based HRV measurement.

Objective. There is growing interest in the use of physiological signals beyond electrocardiography (ECG), particularly photoplethysmography-based noninvasive arterial blood pressure (nABP), to assess autonomic nervous system (ANS) activity with minimal recording durations. This study compared heart rate variability (HRV) and pulse rate variability (PRV) derived from ECG and nABP, respectively. We investigated how signal shortening and calculation window size affect time-domain, frequency-domain, and nonlinear ANS metrics.Approach. Photoplethysmography was used to measure nABP, whereas ECG was recorded with a 3-lead device in healthy individuals (18-31 years). The HRV and PRV were analyzed using time- and frequency-domain metrics, and nonlinear indices, including entropy metrics and Poincaré plots (SD1, SD2). Agreement between signal lengths of 3 min and 5 min was assessed in 86 nABP and 70 ECG participants using intraclass correlation coefficients (ICCs). To evaluate the effect of window size, 15 min recordings from 16 participants were segmented into windows of 3 min, 5 min, and 15 min. HRV-PRV agreement was evaluated using Bland-Altman analysis.Main results. The time-domain metrics demonstrated excellent reproducibility when the signal length (ICCs ⩾ 0.96) and window size (ICCs ⩾ 0.98) were shortened, but moderate agreement between HRV and PRV. Entropy metrics were most affected by signal shortening (e.g. HRV multiscale entropy ICC (95%CI]): 0.67 (0.47-0.80); PRV approximate entropy: 0.45 (0.15-0.64)). Shorter window sizes affected selected ANS metrics, including reduced SD2 (p= 0.003 for HRV,p< 0.001 for PRV) and increased frequency-domain values (p< 0.001 for HRV and PRV).Significance. Time-domain metrics are more robust to reductions in signal length and calculation window size but demonstrate lower interchangeability between HRV and PRV. Both signal length and window size influence selected ANS metrics and should be considered, particularly when employing entropy-based indices in wearable, remote, and short-duration physiological monitoring.

Peripheral hemodynamics, measured via the blood volume pulse and vasomotion, provide a valuable way of monitoring physiological state. Camera imaging-based systems can be used to measure these peripheral signals without contact with the body, at distances of multiple meters. While researchers have paid attention to non-contact imaging photoplethysmography, the study of peripheral hemodynamics and the effect of autonomic nervous system activity on these signals has received less attention. Using a method, based on a tissue-like model of the skin, we extract melanin text {C}_{m} and hemoglobin text {C}_{HbO} concentrations from videos of the hand and face and show that significant decreases in peripheral pulse signal power (by 36% ± 29%) and vasomotion signal power (by 50% ± 26%) occur during periods of cognitive and psychological stress. Via three experiments we show that similar results are achieved across different stimuli and regions of skin (face and hand). While changes in peripheral pulse and vasomotion power were significant the changes in pulse rate variability were less consistent across subjects and tasks.

Extreme bradycardia (EB), extreme tachycardia (ET), ventricular tachycardia (VT), and ventricular flutter (VF) are the four types of life-threatening arrhythmias, which are symptoms of cardiovascular diseases. Therefore, in this study, a method of life-threatening arrhythmia recognition is proposed based on pulse rate variability (PRV). First, noise and interference are wiped out from the arterial blood pressure (ABP), and the PRV signal is extracted. Then, 19 features are extracted from the PRV signal, and 15 features with highly important and significant variation were selected by random forest (RF). Finally, the back-propagation neural network (BPNN), extreme learning machine (ELM), and decision tree (DT) are used to build, train, and test classifiers to detect life-threatening arrhythmias. The experimental data are obtained from the MIMIC/Fantasia and the 2015 Physiology Net/CinC Challenge databases. The experimental results show that the DT classifier has the best average performance with accuracy and kappa coefficient (kappa) of 98.76 ± 0.08% and 97.59 ± 0.15%, which are higher than those of the BPNN (accuracy = 94.85 ± 1.33% and kappa = 89.95 ± 2.62%) and ELM (accuracy = 95.05 ± 0.14% and kappa = 90.28 ± 0.28%) classifiers. The proposed method shows better performance in identifying four life-threatening arrhythmias compared to existing methods and has potential to be used for home monitoring of patients with life-threatening arrhythmias.