Multivariate classification of systemic vascular resistance using photoplethysmography

Photoplethysmography waveform analysis for classification of vascular tone and arterial blood pressure: Study based on neural networks

BACKGROUND

To test whether a Shallow Neural Network (S-NN) can detect and classify vascular tone dependent changes in arterial blood pressure (ABP) by advanced photopletysmographic (PPG) waveform analysis.

METHODS

PPG and invasive ABP signals were recorded in 26 patients undergoing scheduled general surgery. We studied the occurrence of episodes of hypertension (systolic arterial pressure (SAP) >140 mmHg), normotension and hypotension (SAP < 90 mmHg). Vascular tone according to PPG was classified in two ways: 1) By visual inspection of changes in PPG waveform amplitude and dichrotic notch position; where Classes I-II represent vasoconstriction (notch placed >50% of PPG amplitude in small amplitude waves), Class III normal vascular tone (notch placed between 20-50% of PPG amplitude in normal waves) and Classes IV-V-VI vasodilation (notch <20% of PPG amplitude in large waves). 2) By an automated analysis, using S-NN trained and validated system that combines seven PPG derived parameters.

RESULTS

The visual assessment was precise in detecting hypotension (sensitivity 91%, specificity 86% and accuracy 88%) and hypertension (sensitivity 93%, specificity 88% and accuracy 90%). Normotension presented as a visual Class III (III-III) (median and 1st-3rd quartiles), hypotension as a Class V (IV-VI) and hypertension as a Class II (I-III); all p < .0001. The automated S-NN performed well in classifying ABP conditions. The percentage of data with correct classification by S-ANN was 83% for normotension, 94% for hypotension, and 90% for hypertension.

CONCLUSIONS

Changes in ABP were correctly classified automatically by S-NN analysis of the PPG waveform contour.

The Hybrid Excess and Decay (HED) model: an automated approach to characterising changes in the photoplethysmography pulse waveform

Photoplethysmography offers a widely used, convenient and non-invasive approach to monitoring basic indices of cardiovascular function, such as heart rate and blood oxygenation. Systematic analysis of the shape of the waveform generated by photoplethysmography might be useful to extract estimates of several physiological and psychological factors influencing the waveform. Here, we developed a robust and automated method for such a systematic analysis across individuals and across different physiological and psychological contexts. We describe a psychophysiologically-relevant model, the Hybrid Excess and Decay (HED) model, which characterises pulse wave morphology in terms of three underlying pressure waves and a decay function. We present the theoretical and practical basis for the model and demonstrate its performance when applied to a pharmacological dataset of 105 participants receiving intravenous administrations of the sympathomimetic drug isoproterenol (isoprenaline). We show that these parameters capture photoplethysmography data with a high degree of precision and, moreover, are sensitive to experimentally-induced changes in interoceptive arousal within individuals. We conclude by discussing the possible value in using the HED model as a complement to standard measures of photoplethysmography signals.

Photoplethysmogram Analysis and Applications: An Integrative Review

Beyond its use in a clinical environment, photoplethysmogram (PPG) is increasingly used for measuring the physiological state of an individual in daily life. This review aims to examine existing research on photoplethysmogram concerning its generation mechanisms, measurement principles, clinical applications, noise definition, pre-processing techniques, feature detection techniques, and post-processing techniques for photoplethysmogram processing, especially from an engineering point of view. We performed an extensive search with the PubMed, Google Scholar, Institute of Electrical and Electronics Engineers (IEEE), ScienceDirect, and Web of Science databases. Exclusion conditions did not include the year of publication, but articles not published in English were excluded. Based on 118 articles, we identified four main topics of enabling PPG: (A) PPG waveform, (B) PPG features and clinical applications including basic features based on the original PPG waveform, combined features of PPG, and derivative features of PPG, (C) PPG noise including motion artifact baseline wandering and hypoperfusion, and (D) PPG signal processing including PPG preprocessing, PPG peak detection, and signal quality index. The application field of photoplethysmogram has been extending from the clinical to the mobile environment. Although there is no standardized pre-processing pipeline for PPG signal processing, as PPG data are acquired and accumulated in various ways, the recently proposed machine learning-based method is expected to offer a promising solution.

Pulse taking by a piezoelectric film sensor via mode energy ratio analysis helps identify pregnancy status

In order to solve the problem of non-invasive diagnosis and monitoring of women during pregnancy, a piezoelectric film pulse sensing system combined with the mode energy ratio (MER) analysis is utilized to detect human pulses to reveal pregnant conditions. Inspired by traditional Chinese medicine (TCM), pulse diagnosis has a history of more than 2,500 years. The life energy of the human body helps the diagnosis of the disease through the circulation of blood vessels connected to the organs. A PVDF piezoelectric film sensor is used to emulate the pulse taking process in TCM to record the pulse signals. And the algorithm of MER is proposed based on empirical mode decomposition (EMD). Through the MER analysis of 83 female volunteers with different pregnancy statuses, the identification and warning of pregnancy status and physical health indicators are realized.

Quantitative Comparison of Photoplethysmographic Waveform Characteristics: Effect of Measurement Site

Introduction: Photoplethysmography (PPG) has been widely used to assess cardiovascular function. However, few studies have comprehensively investigated the effect of measurement site on PPG waveform characteristics. This study aimed to provide a quantitative comparison on this. Methods: Thirty six healthy subjects participated in this study. For each subject, PPG signals were sequentially recorded for 1 min from six different body sites (finger, wrist under (anatomically volar), wrist upper (dorsal), arm, earlobe, and forehead) under both normal and deep breathing patterns. For each body site under a certain breathing pattern, the mean amplitude was firstly derived from recorded PPG waveform which was then normalized to derive several waveform characteristics including the pulse peak time (Tp), dicrotic notch time (Tn), and the reflection index (RI). The effects of breathing pattern and measurement site on the waveform characteristics were finally investigated by the analysis of variance (ANOVA) with post hoc multiple comparisons. Results: Under both breathing patterns, the PPG measurements from the finger achieved the highest percentage of analyzable waveforms for extracting waveform characteristics. There were significant effects of breathing pattern on Tn and RI (larger Tn and smaller RI with deep breathing on average, both p < 0.03). The effects of measurement site on mean amplitude, Tp, Tn, and RI were significant (all p < 0.001). The key results were that, under both breathing patterns, the mean amplitude from finger PPG was significantly larger and its Tp and RI were significantly smaller than those from the other five sites (all p < 0.001, except p = 0.04 for the Tp of "wrist under"), and Tn was only significantly larger than that from the earlobe (both p < 0.05). Conclusion: This study has quantitatively confirmed the effect of PPG measurement site on PPG waveform characteristics (including mean amplitude, Tp, Tn, and RI), providing scientific evidence for a better understanding of the PPG waveform variations between different body sites.

Photoplethysmographic characterization of vascular tone mediated changes in arterial pressure: an observational study

To determine whether a classification based on the contour of the photoplethysmography signal (PPGc) can detect changes in systolic arterial blood pressure (SAP) and vascular tone. Episodes of normotension (SAP 90-140 mmHg), hypertension (SAP > 140 mmHg) and hypotension (SAP < 90 mmHg) were analyzed in 15 cardiac surgery patients. SAP and two surrogates of the vascular tone, systemic vascular resistance (SVR) and vascular compliance (Cvasc = stroke volume/pulse pressure) were compared with PPGc. Changes in PPG amplitude (foot-to-peak distance) and dicrotic notch position were used to define 6 classes taking class III as a normal vascular tone with a notch placed between 20 and 50% of the PPG amplitude. Class I-to-II represented vasoconstriction with notch placed > 50% in a small PPG, while class IV-to-VI described vasodilation with a notch placed < 20% in a tall PPG wave. 190 datasets were analyzed including 61 episodes of hypertension [SAP = 159 (151-170) mmHg (median 1st-3rd quartiles)], 84 of normotension, SAP = 124 (113-131) mmHg and 45 of hypotension SAP = 85(80-87) mmHg. SAP were well correlated with SVR (r = 0.78, p < 0.0001) and Cvasc (r = 0.84, p < 0.0001). The PPG-based classification correlated well with SAP (r = - 0.90, p < 0.0001), SVR (r = - 0.72, p < 0.0001) and Cvasc (r = 0.82, p < 0.0001). The PPGc misclassified 7 out of the 190 episodes, presenting good accuracy (98.4% and 97.8%), sensitivity (100% and 94.9%) and specificity (97.9% and 99.2%) for detecting episodes of hypotension and hypertension, respectively. Changes in arterial pressure and vascular tone were closely related to the proposed classification based on PPG waveform.Clinical Trial Registration NTC02854852.

Multi-Wavelength Photoplethysmography Enabling Continuous Blood Pressure Measurement With Compact Wearable Electronics

OBJECTIVE

To fight the "silent killer" hypertension, continuous blood pressure (BP) monitoring has been one of the most desired functions in wearable electronics. However, current BP measuring principles and protocols either involve a vessel occlusion process with a cuff or require multiple sensing nodes on the body, which makes it difficult to implement them in compact wearable electronics like smartwatches and wristbands with long-term wearability.

METHODS

In this work, we proposed a highly compact multi-wavelength photoplethysmography (MWPPG) module and a depth-resolved MWPPG approach for continuous monitoring of BP and systemic vascular resistance (SVR). By associating the wavelength-dependent light penetration depth in the skin with skin vasculatures, our method exploited the pulse transit time (PTT) on skin arterioles for tracking SVR (n = 20). Then, we developed an arteriolar PTT-based method for beat-to-beat BP measurement. The BP estimation accuracy of the proposed arteriolar PTT method was validated against Finometer (n = 20) and the arterial line (n = 4).

RESULTS

The correlation between arteriolar PTT and SVR was theoretically deduced and experimentally validated on 20 human subjects performing various maneuvers. The proposed arteriolar PTT-based method outperformed the traditional arterial PTT-based method with better BP estimation accuracy and simpler measurement setup, i.e., with a single sensing node.

CONCLUSION

The proposed depth-resolved MWPPG method can provide accurate measurements of SVR and BP, which are traditionally difficult to measure in a noninvasive or continuous fashion.

SIGNIFICANCE

This MWPPG work provides the wearable healthcare electronics of compact size with a low-cost and physiology-based solution for continuous measurement of BP and SVR.

Investigation of peripheral photoplethysmographic morphology changes induced during a hand-elevation study

A hand-elevation study was carried out in the laboratory in order to alter peripheral blood flow with the aim of increasing understanding of factors affecting the morphology of peripheral photoplethysmographic signals. Photoplethysmographic (PPG) signals were recorded from twenty healthy volunteer subjects during a hand-elevation study in which the right hand was raised and lowered relative to heart level, while the left hand remained static. Red and infrared (IR) PPG signals were obtained from the right and left index fingers using a custom-made PPG processing system. PPG features were identified using a feature-detection algorithm based on the first derivative of the PPG signal. The systolic PPG amplitude, the reflection index, crest time, pulse width at half height, and delta T were calculated from 20 s IR PPG signals from three positions of the right hand with respect to heart level (-50, 0, +50 cm) in 19 volunteers. PPG features were found to change with hand elevation. On lowering the hand to 50 cm below heart level, ac systolic PPG amplitudes from the finger decreased by 68.32 %, while raising the arm increased the systolic amplitude by 69.99 %. These changes in amplitude were attributed to changes in hydrostatic pressure and the veno-arterial reflex. Other morphological variables, such as crest time, were found to be statistically significantly different across hand positions, indicating increased vascular resistance on arm elevation than on dependency. It was hypothesized that these morphological PPG changes were influenced by changes in downstream venous resistance, rather than arterial, or arteriolar, resistance. Changes in hand position relative to heart level can significantly affect the morphology of the peripheral ac PPG waveform. These alterations are due to a combination of physical effects and physiological responses to changes in hand position, which alter vascular resistance. Care should be taken when interpreting morphological data derived from PPG signals and methods should be standardized to take these effects into account.

The effect of vascular changes on the photoplethysmographic signal at different hand elevations

In order to further understand the contribution of venous and arterial effects to the photoplethysmographic (PPG) signal, recordings were made from 20 healthy volunteer subjects during an exercise in which the right hand was raised and lowered with reference to heart level. Red (R) and infrared (IR) PPG signals were obtained from the right index finger using a custom-made PPG processing system. Laser Doppler flowmetry (LDF) signals were also recorded from an adjacent fingertip. The signals were compared with simultaneous PPG signals obtained from the left index finger. On lowering the hand to 50 cm below heart level, both ac and dc PPG amplitudes from the finger decreased (e.g. 18.70 and 63.15% decrease in infrared dc and ac signals respectively). The decrease in dc amplitude most likely corresponded to increased venous volume, while the decrease in ac PPG amplitude was due to regulatory adjustments on the arterial side in response to venous distension. Conversely, ac and dc PPG amplitudes increased on raising the arm above heart level. Morphological changes in the ac PPG signal are thought to be due to vascular resistance changes, predominately venous, as the hand position is changed.

Applications of supervised learning to biological signals: ECG signal quality and systemic vascular resistance

Discovering information encoded in non-invasively recorded biosignals which belies an individual's well-being can help facilitate the development of low-cost unobtrusive medical device technologies, or enable the unsupervised performance of physiological assessments without excessive oversight from trained clinical personnel. Although the unobtrusive or unsupervised nature of such technologies often results in less accurate measures than their invasive or supervised counterparts, this disadvantage is typically outweighed by the ability to monitor larger populations than ever before. The expected consequential benefit will be an improvement in healthcare provision and health outcomes for all. The process of discovering indicators of health in unsupervised or unobtrusive biosignal recordings, or automatically ensuring the validity and quality of such signals, is best realized when following a proven systematic methodology. This paper provides a brief tutorial review of supervised learning, which is a sub-discipline of machine learning, and discusses its application in the development of algorithms to interpret biosignals acquired in unsupervised or semi-supervised environments, with the aim of estimating well-being. Some specific examples in the disparate application areas of telehealth electrocardiogram recording and calculating post-operative systemic vascular resistance are discussed in the context of this systematic approach for information discovery.

Machine learning techniques for arterial pressure waveform analysis

The Arterial Pressure Waveform (APW) can provide essential information about arterial wall integrity and arterial stiffness. Most of APW analysis frameworks individually process each hemodynamic parameter and do not evaluate inter-dependencies in the overall pulse morphology. The key contribution of this work is the use of machine learning algorithms to deal with vectorized features extracted from APW. With this purpose, we follow a five-step evaluation methodology: (1) a custom-designed, non-invasive, electromechanical device was used in the data collection from 50 subjects; (2) the acquired position and amplitude of onset, Systolic Peak (SP), Point of Inflection (Pi) and Dicrotic Wave (DW) were used for the computation of some morphological attributes; (3) pre-processing work on the datasets was performed in order to reduce the number of input features and increase the model accuracy by selecting the most relevant ones; (4) classification of the dataset was carried out using four different machine learning algorithms: Random Forest, BayesNet (probabilistic), J48 (decision tree) and RIPPER (rule-based induction); and (5) we evaluate the trained models, using the majority-voting system, comparatively to the respective calculated Augmentation Index (AIx). Classification algorithms have been proved to be efficient, in particular Random Forest has shown good accuracy (96.95%) and high area under the curve (AUC) of a Receiver Operating Characteristic (ROC) curve (0.961). Finally, during validation tests, a correlation between high risk labels, retrieved from the multi-parametric approach, and positive AIx values was verified. This approach gives allowance for designing new hemodynamic morphology vectors and techniques for multiple APW analysis, thus improving the arterial pulse understanding, especially when compared to traditional single-parameter analysis, where the failure in one parameter measurement component, such as Pi, can jeopardize the whole evaluation.

Estimation of cardiac output and systemic vascular resistance using a multivariate regression model with features selected from the finger photoplethysmogram and routine cardiovascular measurements

Background

Cardiac output (CO) and systemic vascular resistance (SVR) are two important parameters of the cardiovascular system. The ability to measure these parameters continuously and noninvasively may assist in diagnosing and monitoring patients with suspected cardiovascular diseases, or other critical illnesses. In this study, a method is proposed to estimate both the CO and SVR of a heterogeneous cohort of intensive care unit patients (N=48).

Methods

Spectral and morphological features were extracted from the finger photoplethysmogram, and added to heart rate and mean arterial pressure as input features to a multivariate regression model to estimate CO and SVR. A stepwise feature search algorithm was employed to select statistically significant features. Leave-one-out cross validation was used to assess the generalized model performance. The degree of agreement between the estimation method and the gold standard was assessed using Bland-Altman analysis.

Results

The Bland-Altman bias ±precision (1.96 times standard deviation) for CO was -0.01 ±2.70 L min^-1 when only photoplethysmogram (PPG) features were used, and for SVR was -0.87 ±412 dyn.s.cm^-5 when only one PPG variability feature was used.

Conclusions

These promising results indicate the feasibility of using the method described as a non-invasive preliminary diagnostic tool in supervised or unsupervised clinical settings.