Search

Arterial stiffness (AS) serves as a crucial indicator of arterial elasticity and function, typically requiring expensive equipment for detection. Given the strong correlation between AS and various photoplethysmography (PPG) features, PPG emerges as a convenient method for assessing AS. However, the limitations of independent PPG features hinder detection accuracy. This study introduces a feature selection method leveraging the interactive relationships between features to enhance the accuracy of predicting AS from a single-channel PPG signal. Initially, an adaptive signal interception method was employed to capture high-quality signal fragments from PPG sequences. 58 PPG features, deemed to have potential contributions to AS estimation, were extracted and analyzed. Subsequently, the interaction factor (IF) was introduced to redefine the interaction and redundancy between features. A feature selection algorithm (IFFS) based on the IF was then proposed, resulting in a combination of interactive features. Finally, the Xgboost model is utilized to estimate AS from the selected features set. The proposed approach is evaluated on datasets of 268 male and 124 female subjects, respectively. The results of AS estimation indicate that IFFS yields interacting features from numerous sources, rejects redundant ones, and enhances the association. The interaction features combined with the Xgboost model resulted in an MAE of 122.42 and 142.12 cm/sec, an SDE of 88.16 and 102.56 cm/sec, and a PCC of 0.88 and 0.85 for the male and female groups, respectively. The findings of this study suggest that the stated method improves the accuracy of predicting AS from single-channel PPG, which can be used as a non-invasive and cost-effective screening tool for atherosclerosis.

Video-based Photoplethysmography (VPPG) offers the capability to measure heart rate (HR) from facial videos. However, the reliability of the HR values extracted through this method remains uncertain, especially when videos are affected by various disturbances. Confronted by this challenge, we introduce an innovative framework for VPPG-based HR measurements, with a focus on capturing diverse sources of uncertainty in the predicted HR values. In this context, a neural network named HRUNet is structured for HR extraction from input facial videos. Departing from the conventional training approach of learning specific weight (and bias) values, we leverage the Bayesian posterior estimation to derive weight distributions within HRUNet. These distributions allow for sampling to encode uncertainty stemming from HRUNet's limited performance. On this basis, we redefine HRUNet's output as a distribution of potential HR values, as opposed to the traditional emphasis on the single most probable HR value. The underlying goal is to discover the uncertainty arising from inherent noise in the input video. HRUNet is evaluated across 1,098 videos from seven datasets, spanning three scenarios: undisturbed, motion-disturbed, and light-disturbed. The ensuing test outcomes demonstrate that uncertainty in the HR measurements increases significantly in the scenarios marked by disturbances, compared to that in the undisturbed scenario. Moreover, HRUNet outperforms state-of-the-art methods in HR accuracy when excluding HR values with 0.4 uncertainty. This underscores that uncertainty emerges as an informative indicator of potentially erroneous HR measurements. With enhanced reliability affirmed, the VPPG technique holds the promise for applications in safety-critical domains.

Remote photoplethysmography (rPPG) enables non-contact measurement of heart rate (HR). However, the stability of rPPG extraction is a bottleneck limiting its application. To address this issue, a signal quality ranking and fusion (SQRF) approach based on HR continuity in the time–frequency domain is introduced. Firstly, the facial region is divided into multiple regions of interest (ROIs), and the raw blood volume pulse (BVP) signal is extracted from each ROI separately using a conventional rPPG method such as the plane orthogonal to skin (POS) method. Then, wavelet synchrosqueezed transform (WSST) is employed to convert the raw pulse signals into spectrograms, which are further ranked according to the HR instantaneous continuity. The selected spectrograms with high-quality HR continuity are then fused using a weighted average to predict the final HR. The proposed SQRF algorithm is verified on three public datasets DDPM, UBFC-Phys and PURE with real scenarios. The obtained mean absolute error (MAE) was reduced by 58.7%, 47.5%, and 16.0% respectively, compared to the original single-ROI method. The results prove that SORF with spectrogram-based HR continuity can consistently boost the stability of POS. • Blood volume pulse quality assessment based on the continuity of heart rate. • Fusion of selected spectrograms across multiple patches to achieve stable heart rate measurement. • Boosting the accuracy and stability of existing remote photoplethysmography methods. [ABSTRACT FROM AUTHOR]

Objective. Daily blood pressure (BP) monitoring is essential since BP levels can reflect the functions of heart pumping and vasoconstriction. Although various neural network-based BP estimate approaches have been proposed, they have certain practical shortcomings, such as low estimation accuracy and poor model generalization. Based on the strategy of pre-training and partial fine-tuning, this work proposes a non-invasive method for BP estimation using the photoplethysmography (PPG) signal. Approach. To learn the PPG-BP relationship, the deep convolutional bidirectional recurrent neural network (DC-Bi-RNN) was pre-trained with data from the public medical information mark for intensive care (MIMIC III) database. A tiny quantity of data from the target subject was used to fine-tune the specific layers of the pre-trained model to learn more individual-specific information to achieve highly accurate BP estimation. Main results. The mean absolute error and the Pearson correlation coefficient (r) of the proposed algorithm are 3.21 mmHg and 0.919 for systolic BP, and 1.80 mmHg and 0.898 for diastolic BP (DBP). The experimental results show that our method outperforms other methods and meets the requirements of the Association for the Advancement of Medical Instrumentation standard, and received an A grade according to the British Hypertension Society standard. Significance. The proposed method applies the strategy of pre-training and partial fine-tuning to BP estimation and verifies its effectiveness in improving the accuracy of non-invasive BP estimation. [ABSTRACT FROM AUTHOR]

Effective blood pressure (BP) management is crucial for early cardiovascular disease treatment. Photoplethysmography (PPG) waveform analysis presents a promising avenue for non-invasive BP estimation. However, existing methods often overlook the impact of PPG feature changes on BP variations, and the effectiveness and physiological interpretability of the features they employ remain largely unclear. Hence, this work introduces a single-channel BP estimation algorithm founded on the PPG differential features (DFs). A three-month tracking of PPG and BP data was conducted on 90 volunteers, and 68 blood pressure-related features were extracted from their PPG. The features in the first measurement data were used as baseline features. The differential features were obtained by subtracting the baseline features from the features in the subsequent measurement data through differential feature processing, and 11 DFs for BP estimation were retained using feature selection. Subsequently, four machine learning models were applied for the estimation of systolic (SBP) and diastolic blood pressure (DBP) changes, and finally introduced baseline blood pressure values to achieve BP estimation. Investigating the correlation between PPG DFs and BP changes, we observed a more significant correlation with BP changes for PPG DFs compared to PPG features. Among the machine learning models, the Random Forest model exhibited the most accurate estimation results in blood pressure, achieving standard deviation of the error(STD) of 7.15 mmHg for SBP and 5.30 mmHg for DBP, along with Pearson correlation coefficients (PCC) of 0.90 and 0.86, respectively. This work demonstrates that incorporating PPG feature changes with BP changes achieves higher accuracy in BP estimation, confirming the effectiveness of PPG DFs in BP estimation and providing valuable insights for long-term non-invasive blood pressure monitoring. • Estimation of blood pressure (BP) from a single-channel PPG signal. • Combining changes in features with changes in BP, the PPG differential feature is presented. • A collaborative filter-wrapper feature selection method is used to select the optimal subset of differential features. • Improving PPG signal quality using adaptive interception methods. • Introduction of individual baseline BP values improves the accuracy of BP estimation. [ABSTRACT FROM AUTHOR]

Objective: Currently, BP measurement devices are mainly cuff-based which are not portable or convenient for users. To simplify the measurement of BP, this article proposed a new framework for noninvasive BP estimation using single-channel PPG signals. Methods: Various PPG features that may be related to BP were extracted and a filter-wrapper collaborated feature selection method was used for rejecting irrelevant and redundant features. The features that maximize the correlation with BP were finally selected as the BP-oriented IFS, and a new LASSO-LSTM model was designed to estimate BP from the IFS. Results: Experiments were conducted on a public dataset and a self-collected clinical dataset, respectively. Results demonstrated that the proposed method is superior to previously reported methods in the literature, giving a mean absolute error of 4.95 mmHg for SBP and 3.15 mmHg for DBP which complies with the standard of the AAMI. Conclusion: The proposed filter-wrapper collaborated feature selection method could effectively reject weak correlation and redundant features, and the designed LASSO-LSTM model is capable of learning complicated nonlinear relations between the selected IFS and BP. The proposed method shows improved accuracy of noninvasive BP estimation. [ABSTRACT FROM AUTHOR]

Recent studies have shown that changes in human facial skin color due to varying blood flows can be recorded on videos by consumer-level cameras and, interestingly, can be used to measure pulse rates in a noncontact manner. However, it remains unclear whether the video-based pulse rate detection is feasible when videos are recorded in unstable lighting conditions. To address unstable illumination, a novel algorithm named adaptive spatiotemporal homomorphic filtering (ASTHF) is proposed in this article, with a focus on the separation of pulse signals and illumination-induced noise. This method first employs the homomorphic filtering framework to make illumination-induced noise in a video linearly separable. Subsequently, it incorporates a spatiotemporal filter into the framework, aiming to remove illumination-induced noise from both the time domain and the space domain of the video. To adapt to diverse illumination in the real world, the filter parameters are determined adaptively according to the sparsity in frequency spectrum of pulse signals. ASTHF is tested on two data sets, including 406 video samples. The test results show that ASTHF has stronger robustness to unstable illumination than state-of-the-art methods, improving the Pearson’s correlation coefficient by 5% in a challenging scenario compared with the best contrasting method. The proposed method demonstrates promise to achieve noncontact pulse rate detection from facial videos involving unsteady illumination. [ABSTRACT FROM AUTHOR]