Showing papers in "Physiological Measurement in 2019"

Recent development of respiratory rate measurement technologies

Abstract: Respiratory rate (RR) is an important physiological parameter whose abnormality has been regarded as an important indicator of serious illness. In order to make RR monitoring simple to perform, reliable and accurate, many different methods have been proposed for such automatic monitoring. According to the theory of respiratory rate extraction, methods are categorized into three modalities: extracting RR from other physiological signals, RR measurement based on respiratory movements, and RR measurement based on airflow. The merits and limitations of each method are highlighted and discussed. In addition, current works are summarized to suggest key directions for the development of future RR monitoring methodologies.

An open access database for the evaluation of respiratory sound classification algorithms.

Abstract: Objective Over the last few decades, there has been significant interest in the automatic analysis of respiratory sounds. However, currently there are no publicly available large databases with which new algorithms can be evaluated and compared. Further developments in the field are dependent on the creation of such databases. Approach This paper describes a public respiratory sound database, which was compiled for an international competition, the first scientific challenge of the IFMBE's International Conference on Biomedical and Health Informatics. The database includes 920 recordings acquired from 126 participants and two sets of annotations. One set contains 6898 annotated respiratory cycles, some including crackles, wheezes, or a combination of both, and some with no adventitious respiratory sounds. In the other set, precise locations of 10 775 events of crackles and wheezes were annotated. Main results The best system that participated in the challenge achieved an average score of 52.5% with the respiratory cycle annotations and an average score of 91.2% with the event annotations. Significance The creation and public release of this database will be useful to the research community and could bring attention to the respiratory sound classification problem.

Detecting and interpreting myocardial infarction using fully convolutional neural networks.

Abstract: OBJECTIVE We aim to provide an algorithm for the detection of myocardial infarction that operates directly on ECG data without any preprocessing and to investigate its decision criteria. APPROACH We train an ensemble of fully convolutional neural networks on the PTB ECG dataset and apply state-of-the-art attribution methods. MAIN RESULTS Our classifier reaches 93.3% sensitivity and 89.7% specificity evaluated using 10-fold cross-validation with sampling based on patients. The presented method outperforms state-of-the-art approaches and reaches the performance level of human cardiologists for detection of myocardial infarction. We are able to discriminate channel-specific regions that contribute most significantly to the neural network's decision. Interestingly, the network's decision is influenced by signs also recognized by human cardiologists as indicative of myocardial infarction. SIGNIFICANCE Our results demonstrate the high prospects of algorithmic ECG analysis for future clinical applications considering both its quantitative performance as well as the possibility of assessing decision criteria on a per-example basis, which enhances the comprehensibility of the approach.

Validity of Mobility Lab (version 2) for gait assessment in young adults, older adults and Parkinson's disease

Abstract: Objective: Gait provides a sensitive measurement for signs of aging and neurodegenerative conditions. Measurement of gait is transitioning from the laboratory environment to the clinic with the use of inertial measurement units, providing a simple and cost-effective assessment tool. However, such assessments first need validation against reference systems. The aim of this study was to validate the APDM Mobility Lab (ML) system (version 2) against a pressure sensor walkway in younger adults (n = 18), older adults (n = 18) and people with mild Parkinson's disease (n = 21) in the laboratory. Approach: Participants completed a two-minute walk over a pressure sensor walkway whilst wearing three sensors (strapped to the lumbar spine and both feet). Comparison of output from the systems was then performed. Main results: Overall, we identified that ML provided good to excellent agreement (ICC > 0.75) for gait velocity, stride length, stride length SD, cadence, stride time and stride time SD. Measures of double support time, single support time and swing time had moderate to poor agreement (ICC 0.213–0.725), particularly for younger adults and PD. Significance: Overall, Mobility Lab provides a valid system for gait data collection for clinical and research application.

Heart rate variability monitoring for emotion and disorders of emotion

Abstract: BACKGROUND Emotion is composed of cognitive processing, physiological response and behavioral reaction. Heart rate variability (HRV) refers to the fluctuations between consecutive heartbeat cycles, and is considered as a non-invasive method for evaluating cardiac autonomic function. HRV analysis plays an important role in emotional study and detection. OBJECTIVE In this paper, the physiological foundation of HRV is briefly described, and then the relevant literature relating to HRV-based emotion studies for the performance of HRV in different emotions, emotion recognition, the evaluation of emotional disorders, HRV biofeedback, as well as HRV-based emotion analysis and management enhanced by wearable devices, are reviewed. SIGNIFICANCE It is suggested that HRV is an effective tool for the measurement and regulation of emotional response, with a broad application prospect.

Combining deep neural networks and engineered features for cardiac arrhythmia detection from ECG recordings

Abstract: Objective We aim to combine deep neural networks and engineered features (hand-crafted features based on medical domain knowledge) for cardiac arrhythmia detection from short single-lead ECG recordings. Approach We propose a two-stage method named ENCASE for cardiac arrhythmia detection. The first stage is feature extraction and the second stage is classifier building. In the feature extraction stage, we extract both deep features and engineered features. Deep features are obtained by modifying deep neural networks into a deep feature extractor. Engineered features are extracted by summarizing existing approaches into four feature groups. Then, we propose a feature aggregation approach to combine these features. In the classifier building stage, we build multiple gradient boosting decision trees and combine them to get the final detector. Main results Experiments are performed on the PhysioNet/Computing in Cardiology Challenge 2017 dataset (Clifford et al 2017 Computing in Cardiology vol 44). Using F 1 scores reported on the hidden test set as measurements, ENCASE got 0.9117 on Normal (F 1N ), 0.8128 on Atrial Fibrillation (AF) (F 1A ), 0.7505 on Others (F 1O ), and 0.5671 on Noise (F 1P ). It placed 5th in the Challenge and 8th in the follow-up challenge (ranked by considering the average of Normal, AF, and Others (F 1NAO = 0.825)). When rounding to two decimal places, we were part a three-way tie for 1st place and were part a seven-way tie for 2nd place in the follow-up challenge. Further experiments show that combined features perform better than individual features, and deep features show more importance scores than other features. Significance ENCASE can benefit from both feature engineering-based methods and recent deep neural networks. It is flexible and can easily assimilate the ability of new cardiac arrhythmia detection methods.

Estimating blood pressure trends and the nocturnal dip from photoplethysmography.

Abstract: Objective Evaluate a method for the estimation of the nocturnal systolic blood pressure (SBP) dip from 24 h blood pressure trends using a wrist-worn photoplethysmography (PPG) sensor and a deep neural network in free-living individuals, comparing the deep neural network to traditional machine learning and non-machine learning baselines. Approach A wrist-worn PPG sensor was worn by 106 healthy individuals for 226 d during which 5111 reference values for blood pressure (BP) were obtained with a 24 h ambulatory BP monitor and matched with the PPG sensor data. Features based on heart rate variability and pulse morphology were extracted from the PPG waveforms. Long- and short term memory (LSTM) networks, dense networks, random forests and linear regression models were trained and evaluated in their capability of tracking trends in BP, as well as the estimation of the SBP dip. Main results Best performance for estimating the SBP dip were obtained with a deep LSTM neural network with a root mean squared error (RMSE) of 3.12 [Formula: see text] 2.20 [Formula: see text] mmHg and a correlation of 0.69 [Formula: see text]. This dip was derived from trend estimates of BP which had an RMSE of 8.22 [Formula: see text] 1.49 mmHg for systolic and 6.55 [Formula: see text] 1.39 mmHg for diastolic BP (DBP). While other models had similar performance for the tracking of relative BP, they did not perform as well as the LSTM for the SBP dip. Significance The work provides first evidence for the unobtrusive estimation of the nocturnal SBP dip, a highly prognostic clinical parameter. It is also the first to evaluate unobtrusive BP measurement in a large data set of unconstrained 24 h measurements in free-living individuals and provides evidence for the utility of LSTM models in this domain.

Beltrami-net: domain independent deep D-bar learning for absolute imaging with electrical impedance tomography (a-EIT)

Abstract: Objective To develop, and demonstrate the feasibility of, a novel image reconstruction method for absolute electrical impedance tomography (a-EIT) that pairs deep learning techniques with real-time robust D-bar methods and examine the influence of prior information on the reconstruction. Approach A D-bar method is paired with a trained convolutional neural network (CNN) as a post-processing step. Training data is simulated for the network using no knowledge of the boundary shape by using an associated nonphysical Beltrami equation rather than simulating the traditional current and voltage data specific to a given domain. This allows the training data to be boundary shape independent. The method is tested on experimental data from two EIT systems (ACT4 and KIT4) with separate training sets of varying prior information. Main results Post-processing the D-bar images with a CNN produces significant improvements in image quality measured by structural SIMilarity indices (SSIMs) as well as relative [Formula: see text] and [Formula: see text] image errors. Significance This work demonstrates that more general networks can be trained without being specific about boundary shape, a key challenge in EIT image reconstruction. The work is promising for future studies involving databases of anatomical atlases.

Availability and performance of image-based, non-contact methods of monitoring heart rate, blood pressure, respiratory rate, and oxygen saturation: a systematic review.

Abstract: Objective Over the last 15 years, developments in camera technology have coincided with increased availability and affordability. This has led to an increasing interest in using these technologies in healthcare settings. Image-based monitoring methods potentially allow multiple vital signs to be measured concurrently using a non-contact sensor. We have undertaken a systematic review of the current availability and performance of these monitoring methods. Approach A multiple database search was conducted using MEDLINE, Embase, CINAHL, Cochrane Library, OpenGrey, IEEE Xplore Library and ACM Digital Library to July 2018. We included studies comparing image-based heart rate, respiratory rate, oxygen saturation and blood pressure monitoring methods against one or more validated reference device(s). Each included study was assessed using the modified GRRAS criteria for reporting bias. Main results Of 30 279 identified studies, 161 were included in the final analysis. Twenty studies (20/161, 12%) were carried out on patients in clinical settings, while the remainder were conducted in academic settings using healthy volunteer populations. The 18-40 age group was best represented across the identified studies. One hundred and twenty studies (120/161, 75%) estimated heart rate, followed by 62 studies (62/161, 39%) estimating respiratory rate. Fewer studies focused on oxygen saturation (11/161, 7%) or blood pressure (6/161, 4%) estimation. Fifty-one heart rate studies (51/120, 43%) and 24 respiratory rate studies (24/62, 39%) used Bland-Altman analysis to report their results. Of the heart rate studies, 28 studies (28/51, 55%) showed agreement within industry standards of [Formula: see text]5 beats per minute. Only two studies achieved this within clinical settings. Of the respiratory rate studies, 13 studies (13/24, 54%) showed agreement within industry standards of [Formula: see text]3 breaths per minute, but only one study achieved this in a clinical setting. Statistical analysis was heterogeneous across studies with frequent inappropriate use of correlation. The majority of studies (99/161, 61%) monitored subjects for under 5 min. Three studies (3/161, 2%) monitored subjects for over 60 min, all of which were conducted in hospital settings. Significance Heart rate and respiratory rate monitoring using video images is currently possible and performs within clinically acceptable limits under experimental conditions. Camera-derived estimates were less accurate in the proportion of studies conducted in clinical settings. We would encourage thorough reporting of the population studied, details of clinically relevant aspects of methodology, and the use of appropriate statistical methods in future studies. Systematic review registration: PROSPERO CRD42016029167 Protocol: https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/s13643-017-0615-3.

Optimal fiducial points for pulse rate variability analysis from forehead and finger photoplethysmographic signals.

Abstract: Objective The aim of this work is to evaluate and compare five fiducial points for the temporal location of each pulse wave from forehead and finger photoplethysmographic (PPG) pulse wave signals to perform pulse rate variability (PRV) analysis as a surrogate for heart rate variability (HRV) analysis. Approach Forehead and finger PPG signals were recorded during a tilt-table test simultaneously with the electrocardiogram (ECG). Artefacts were detected and removed and five fiducial points were computed: apex, middle-amplitude and foot points of the PPG signal, apex point of the first derivative signal and the intersection point of the tangent to the PPG waveform at the apex of the derivative PPG signal and the tangent to the foot of the PPG pulse, defined as the intersecting tangents method. Pulse period (PP) time interval series were obtained from both PPG signals and compared with the RR intervals obtained from the ECG. HRV and PRV signals were estimated and classical time and frequency domain indices were computed. Main results The middle-amplitude point of the PPG signal (n M ), the apex point of the first derivative ([Formula: see text]), and the tangent intersection point (n T ) are the most suitable fiducial points for PRV analysis, resulting in the lowest relative errors estimated between PRV and HRV indices and higher correlation coefficients and reliability indices. Statistically significant differences according to the Wilcoxon test between PRV and HRV signals were found for the apex and foot fiducial points of the PPG, as well as the lowest agreement between RR and PP series according to Bland-Altman analysis. Hence, these signals have been considered less accurate for variability analysis. In addition, the relative errors are significantly lower for n M and [Formula: see text] using Friedman statistics with a Bonferroni multiple-comparison test, and we propose that n M is the most accurate fiducial point. Based on our results, forehead PPG seems to provide more reliable information for a PRV assessment than finger PPG. Significance The accuracy of the pulse wave detection depends on the morphology of the PPG. There is therefore a need to widely define the most accurate fiducial point for performing a PRV analysis under non-stationary conditions based on different PPG sensor locations and signal acquisition techniques.

Automatic evaluation of fetal head biometry from ultrasound images using machine learning

Abstract: Objective: Ultrasound-based fetal biometric measurements, such as head circumference (HC) and biparietal diameter (BPD), are frequently used to evaluate gestational age and diagnose fetal central nervous system pathology. Because manual measurements are operator-dependent and time-consuming, much research is being actively conducted on automated methods. However, the existing automated methods are still not satisfactory in terms of accuracy and reliability, owing to difficulties dealing with various artefacts in ultrasound images. Approach: Using the proposed method, a labeled dataset containing 102 ultrasound images was used for training, and validation was performed with 70 ultrasound images. Main results: A success rate of 91.43% and 100% for HC and BPD estimations, respectively, and an accuracy of 87.14% for the plane acceptance check. Significance: This paper focuses on fetal head biometry and proposes a deep-learning-based method for estimating HC and BPD with a high degree of accuracy and reliability.

The influence of resting heart rate on pulse wave velocity measurement is mediated by blood pressure and depends on aortic stiffness levels: insights from the Corinthia study.

Abstract: Objective Increased resting heart rate as well as increased arterial stiffness are both independent predictors of cardiovascular events and mortality. Results of previous studies have failed to converge concerning the association between heart rate and arterial stiffness, regardless of other potential confounders, such as age, gender and particularly blood pressure (BP). We aimed to investigate: (a) the degree of association (if any) between resting heart rate and carotid-to-femoral pulse wave velocity (PWV), the gold standard index of arterial stiffness, (b) if the relationship between heart rate and PWV is mediated by BP levels and (c) whether their association is affected by the levels of aortic stiffening. Approach Demographic, hemodynamic, laboratory and clinical data of 1566 subjects from the cross-sectional observational 'Corinthia' study were analyzed using univariate and multivariate regression models. Mediation analysis was performed to test whether mean arterial pressure (MAP) is a significant mediator in the heart rate-PWV relationship. The total population was divided in two groups of low and high arterial stiffness according to the median PWV value (8.6 m s-1). Main results We found that (i) there is a significant association between heart rate and PWV, regardless of other confounding factors. An increase in heart rate by 20 b.p.m. can increase PWV by 0.5 m s-1. However, this association was significant only for subjects with increased aortic stiffness (PWV > 8.6 m s-1) and not for those with PWV ⩽ 8.6 m s-1. Further, (ii) heart rate-PWV association was partially mediated by MAP. Significance Increased resting heart rate is related to increased aortic stiffness, only in subjects with stiffer aortas, regardless of BP and other risk factors and subjects' characteristics. The synergistic prognostic effect of increased arterial stiffness and elevated heart rate on target organ damage, cardiovascular events and mortality should be explored in future studies.

Detection of atrial fibrillation using a wrist-worn device.

Abstract: Objective This study proposes an algorithm for the detection of atrial fibrillation (AF), designed to operate on extended photoplethysmographic (PPG) signals recorded using a wrist-worn device of own design. Approach Robustness against false alarms is achieved by means of signal quality assessment and different techniques for suppression of ectopic beats, bigeminy, and respiratory sinus arrhythmia. The decision logic is based on our previously proposed RR interval-based AF detector, but modified to account for differences between interbeat intervals in the ECG and the PPG. The detector is evaluated on simulated PPG signals as well as on clinical PPG signals recorded during cardiac rehabilitation after myocardial infarction. Main results Analysis of the clinical signals showed that 1.5 false alarms were on average produced per day with a sensitivity of 72.0% and a specificity of 99.7% when 89.2% of the database was available for analysis, whereas as many as 15 when the RR interval-based AF detector, boosted by accelerometer information for signal quality assessment, was used. However, a sensitivity of 97.2% and a specificity of 99.6% were achieved when increasing the demands on signal quality so that 50% was available for analysis. Significance The proposed detector offers promising performance and is particularly well-suited for implementation in low-power wearable devices, e.g. wrist-worn devices, with significance in mass screening applications.

Cardio-respiratory signal extraction from video camera data for continuous non-contact vital sign monitoring using deep learning.

Abstract: Non-contact vital sign monitoring enables the estimation of vital signs, such as heart rate, respiratory rate and oxygen saturation (SpO2), by measuring subtle color changes on the skin surface using a video camera. For patients in a hospital ward, the main challenges in the development of continuous and robust non-contact monitoring techniques are the identification of time periods and the segmentation of skin regions of interest (ROIs) from which vital signs can be estimated. This paper presents two convolutional neural network (CNN) models. The first network was designed for detecting the presence of a patient and segmenting the patient's skin area. The second network combined the output from the first network with optical flow for identifying time periods of clinical intervention so that these periods can be excluded from the estimation of vital signs. Both networks were trained using video recordings from a clinical study involving 15 pre-term infants conducted in the high dependency area of the Neonatal Intensive Care Unit (NICU) of the John Radcliffe Hospital in Oxford, UK. The proposed methods achieved an accuracy of 98.8\% for patient detection, a mean intersection-over-union (IOU) score of 88.6\% for skin segmentation and an accuracy of 94.5\% for clinical intervention detection using two-fold cross validation. Our deep learning models produced accurate results and were robust to different skin tones, changes in light conditions, pose variations and different clinical interventions by medical staff and family visitors. Finally, we show that cardio-respiratory signals can be continuously derived from the patient's skin during which the patient is present and no clinical intervention is undertaken.

Video and audio processing in paediatrics: a review.

Abstract: Objective - Video and sound acquisition and processing technologies have seen great improvements in recent decades, with many applications in the biomedical area. The aim of this paper is to review the overall state of the art of advances within these topics in paediatrics and to evaluate their potential application for monitoring in the neonatal intensive care unit (NICU). Approach - For this purpose, more than 150 papers dealing with video and audio processing were reviewed. For both topics, clinical applications are described according to the considered cohorts-full-term newborns, infants and toddlers or preterm newborns. Then, processing methods are presented, in terms of data acquisition, feature extraction and characterization. Main results - The paper first focuses on the exploitation of video recordings; these began to be automatically processed in the 2000s and we show that they have mainly been used to characterize infant motion. Other applications, including respiration and heart rate estimation and facial analysis, are also presented. Audio processing is then reviewed, with a focus on the analysis of crying. The first studies in this field focused on induced-pain cries and the newest ones deal with spontaneous cries; the analyses are mainly based on frequency features. Then, some papers dealing with non-cry signals are also discussed. Significance - Finally, we show that even if recent improvements in digital video and signal processing allow for increased automation of processing, the context of the NICU makes a fully automated analysis of long recordings problematic. A few proposals for overcoming some of the limitations are given.

Deep learning approaches for plethysmography signal quality assessment in the presence of atrial fibrillation.

Abstract: Objective Photoplethysmography (PPG) monitoring has been implemented in many portable and wearable devices we use daily for health and fitness tracking. Its simplicity and cost-effectiveness has enabled a variety of biomedical applications, such as continuous long-term monitoring of heart arrhythmias, fitness, and sleep tracking, and hydration monitoring. One major issue that can hinder PPG-based applications is movement artifacts, which can lead to false interpretations. In many implementations, noisy PPG signals are discarded. Misinterpreted or discarded PPG signals pose a problem in applications where the goal is to increase the yield of detecting physiological events, such as in the case of paroxysmal atrial fibrillation (AF)-a common episodic heart arrhythmia and a leading risk factor for stroke. In this work, we compared a traditional machine learning and deep learning approaches for PPG quality assessment in the presence of AF, in order to find the most robust method for PPG quality assessment. Approach The training data set was composed of 78 278 30 s long PPG recordings from 3764 patients using bedside patient monitors. Two different representations of PPG signals were employed-a time-series based (1D) one and an image-based (2D) one. Trained models were tested on an independent set of 2683 30 s PPG signals from 13 stroke patients. Main results ResNet18 showed a higher performance (0.985 accuracy, 0.979 specificity, and 0.988 sensitivity) than SVM and other deep learning approaches. 2D-based models were generally more accurate than 1D-based models. Significance 2D representation of PPG signal enhances the accuracy of PPG signal quality assessment.

Detection of atrial fibrillation and other abnormal rhythms from ECG using a multi-layer classifier architecture.

Abstract: Objective Atrial fibrillation (AF) and other types of abnormal heart rhythm are related to multiple fatal cardiovascular diseases that affect the quality of human life. Hence the development of an automated robust method that can reliably detect AF, in addition to other non-sinus and sinus rhythms, would be a valuable addition to medicine. The present study focuses on developing an algorithm for the classification of short, single-lead electrocardiogram (ECG) recordings into normal, AF, other abnormal rhythms and noisy classes. Approach The proposed classification framework presents a two-layer, three-node architecture comprising binary classifiers. PQRST markers are detected on each ECG recording, followed by noise removal using a spectrogram power based novel adaptive thresholding scheme. Next, a feature pool comprising time, frequency, morphological and statistical domain ECG features is extracted for the classification task. At each node of the classification framework, suitable feature subsets, identified through feature ranking and dimension reduction, are selected for use. Adaptive boosting is selected as the classifier for the present case. The training data comprises 8528 ECG recordings provided under the PhysioNet 2017 Challenge. F1 scores averaged across the three non-noisy classes are taken as the performance metric. Main result The final five-fold cross-validation score achieved by the proposed framework on the training data has high accuracy with low variance (0.8254 [Formula: see text] 0.0043). Significance Further, the proposed algorithm has achieved joint first place in the PhysioNet/Computing in Cardiology Challenge 2017 with a score of 0.83 computed on a hidden test dataset.

Assessment of microcirculation in the diabetic foot with laser speckle contrast imaging

Abstract: OBJECTIVE: A major challenge for treating diabetic foot ulcers is estimating the severity of ischemia, as the currently used non-invasive diagnostic techniques provide relatively poor prognostic values. Laser speckle contrast imaging (LSCI) is a promising non-invasive technique to assess microcirculation. Our aim was to investigate the stability and reproducibility of LSCI for the assessment of microcirculation in the diabetic foot, the relation of LSCI results to currently used non-invasive blood pressure measurements, and the ability of LSCI to discriminate between the degrees of ischemia. APPROACH: Thirty-three participants with diabetic foot ulcers were included in this prospective, single centre, observational cohort study that was conducted in the Netherlands. They were classified as non-ischemic, ischemic or critical-ischemic based on criteria formulated in the international guidelines. Two clinicians performed LSCI scans of the foot, consisting of baseline measurements, followed by two stress tests (post-occlusion peak and elevation test). With three measurement conditions and five regions of interest of the foot per patient, a total of 15 measurements were available for analyses. MAIN RESULTS: The intra-observer agreement of LSCI was high (interclass correlation coefficient (ICC) = 0.711-0.950; p < 0.001) for all 15 measurements. The inter-observer agreement was high (ICC = 0.728-0.861; p ⩽ 0.001) for 10 measurements and moderate (ICC = 0.476-0.570; p ⩽ 0.005) for the remaining five measurements. The inter-assessor agreement was high and significant (ICC = 0.857-0.996; p ⩽ 0.001) for all measurements. Correlation between LSCI and non-invasive blood pressure measurements was low (ICC = -0.272-0.582). During both stress tests, microcirculation was significantly lower in critical-ischemic feet compared to non-ischemic feet (67.5 perfusion units (PU) versus 96.3 PU and 41.0 PU versus 63.9 PU; p < 0.05). SIGNIFICANCE: LSCI is a stable and reproducible technique for assessment of microcirculation in people with diabetic foot ulcers and shows significant differences between non-ischemic, ischemic and critical-ischemic patient populations.

Brain-heart interactions considering complex physiological data: processing schemes for time-variant, frequency-dependent, topographical and statistical examination of directed interactions by convergent cross mapping

Abstract: Background A multitude of complex methods is available to quantify interactions in highly complex physiological systems. Brain-heart interactions play an important role in identifying couplings between the central nervous system and the autonomic nervous system during defined physiological states or specific diseases. The crucial point of those interaction analyses is adequate pre-processing taking into account nonlinearity of data, intuitive graphical representation and suitable statistical evaluation of the achieved results. Objective The aim of this study is to provide generalized processing schemes for such investigations taking into account pre-processing, graphical representation and statistical analysis. Approach Two defined data sets were used to develop these processing schemes. Brain-heart interactions in children with temporal lobe epilepsy during the pre-ictal, ictal and post-ictal periods as well as in patients with paranoid schizophrenia and healthy control subjects during the resting state period were investigated by nonlinear convergent cross mapping (CCM). Surrogate data, bootstrapping and linear mixed-effects model approaches were utilized for statistical analyses. Main results CCM was able to reveal specific and statistically significant time- and frequency-dependent patterns of brain-heart interactions for children with temporal lobe epilepsy and provide a statistically significant pattern of topographic- and frequency-dependent brain-heart interactions for schizophrenic patients, as well as to show the differences from healthy control subjects. Suitable statistical models were found to quantify group differences. Significance Generalized processing schemes and crucial points of pre-processing, adapted interaction analysis and performed statistical analysis are provided. The general concept of analyses is transferable also to other methods of interactions analysis and data representing even more complex physiological systems.

Ventricular ectopic beat detection using a wavelet transform and a convolutional neural network

Abstract: Objective Ventricular contractions in healthy individuals normally follow the contractions of atria to facilitate more efficient pump action and cardiac output. With a ventricular ectopic beat (VEB), volume within the ventricles are pumped to the body's vessels before receiving blood from atria, thus causing inefficient blood circulation. VEBs tend to cause perturbations in the instantaneous heart rate time series, making the analysis of heart rate variability inappropriate around such events, or requiring special treatment (such as signal averaging). Moreover, VEB frequency can be indicative of life-threatening problems. However, VEBs can often mimic artifacts both in morphology and timing. Identification of VEBs is therefore an important unsolved problem. The aim of this study is to introduce a method of wavelet transform in combination with deep learning network for the classification of VEBs. Approach We proposed a method to automatically discriminate VEB beats from other beats and artifacts with the use of wavelet transform of the electrocardiogram (ECG) and a convolutional neural network (CNN). Three types of wavelets (Morlet wavelet, Paul wavelet and Gaussian derivative) were used to transform segments of single-channel (1D) ECG waveforms to two-dimensional (2D) time-frequency 'images'. The 2D time-frequency images were then passed into a CNN to optimize the convolutional filters and classification. Ten-fold cross validation was used to evaluate the approach on the MIT-BIH arrhythmia database (MIT-BIH). The American Heart Association (AHA) database was then used as an independent dataset to evaluate the trained network. Main results Ten-fold cross validation results on MIT-BIH showed that the proposed algorithm with Paul wavelet achieved an overall F1 score of 84.94% and accuracy of 97.96% on out of sample validation. Independent test on AHA resulted in an F1 score of 84.96% and accuracy of 97.36%. Significance The trained network possessed exceptional transferability across databases and generalization to unseen data.

High-pressure blood flow restriction with very low load resistance training results in peripheral vascular adaptations similar to heavy resistance training.

Abstract: OBJECTIVE To investigate vascular adaptations to eight weeks of resistance exercise, with and without different pressures of blood flow restriction (BFR), in the upper and lower body. APPROACH Forty individuals (men = 20, women = 20) completed eight weeks of resistance exercise at very low loads (15% of one-repetition maximum (1RM)), with two levels of BFR (40% arterial occlusion pressure (AOP), 80% AOP), without BFR, and 70% of 1RM. Vascular conductance and venous compliance were measured via plethysmography before and following training in the forearms and in the calves. MAIN RESULTS Values reported as means (95% confidence intervals). Pre to post changes in vascular conductance occurred only in the high-pressure conditions (upper body: +8.29 (3.01-13.57) ml · mmHg-1; lower body: +7.86 (3.37-12.35) ml · mmHg-1) and high-load conditions (upper body: +8.60 (3.45-13.74) ml · mmHg-1); lower body: +7.20 (2.71-11.69) ml · mmHg-1) only. In the upper body, the change was significantly greater in the high-pressure and high-load conditions compared to the change observed in the low-pressure condition (-0.41 (-5.56, 4.73) ml · mmHg-1). These changes were not greater than the change observed in the low-load condition without pressure (+1.81 (-3.47, 7.09) ml · mmHg-1). In the lower body, the change in the high-pressure and high-load conditions were significantly greater than the changes observed with low-load training with (-0.86 (-5.60, 3.87) ml · mmHg-1) and without (-1.22 (-5.71, 3.27) ml · mmHg-1) a low pressure. Venous compliance increased in all groups in the upper body (+0.003 (.000 08, 0.006) ml · 100 ml-1 · mmHg-1) only, with no changes in the lower body. SIGNIFICANCE High-pressure BFR causes adaptations in vascular function following eight weeks of training at mechanical loads not typically associated with such adaptations.

Eye-tracker algorithms to detect saccades during static and dynamic tasks: a structured review

Abstract: Objective: Eye-tracking devices have become widely used as clinical assessment tools in a variety of applied-scientific fields to measure saccadic eye movements. With the emergence of multiple static and dynamic devices, the concurrent need for algorithm development and validation is paramount. Approach: This review assesses the prevalence of current saccade detection algorithms, their associated validation methodologies and the suitability of their application. Medline, Embase, PsychInfo, Scopus, IEEEXplore and ACM Digital Library databases were searched. Two independent reviewers and an adjudicator screened articles describing the detection of saccades from raw infrared/video-based eye-tracker data. Main results: Thirteen articles were screened and met the inclusion criteria. Overall, the majority of reviewed saccadic detection algorithms used simple velocity-based classifications with static eye-tracking systems. Studies demonstrated validity but are limited by the static nature of testing. Heterogeneity in system design, proprietary and bespoke algorithmic methods used, processing strategies, and outcome reporting is evident. Significance: This paper suggests the use of a more standardised methodology to facilitate experimental validity and improve comparison of results across studies.

Phase entropy: a new complexity measure for heart rate variability.

Abstract: Objective: Information entropy is generally employed for analysing the complexity of physiological signals. However, most of the entropy definitions estimate the degree of compressibility and thus quantify the randomness. Physiological signals are very complex because of a nonlinear relationship and interaction between various systems and subsystems of the body. Therefore, analysis of randomness may not be sufficient to describe the complexity. For analysing the complexity of heart rate variability (HRV), a new entropy method, phase entropy (PhEn), has been proposed as a quantification of two dimensional phase space. Approach: The second order difference plot (SODP), a two dimensional phase space, provides a visual summary of the rate of variability. The distribution of scatter points in a SODP provides information about the dynamics of the underlying system. PhEn estimates the Shannon entropy of the weighted distribution in a coarse-grained SODP. Results: The performance of PhEn has been evaluated by using simulated signals, synthetic HRV signals, and real HRV signals. PhEn shows a better discriminating power and stability as compared with other entropy measures. It is computationally efficient. Moreover, it has the ability to assess temporal asymmetry of physiological signals. Significance: The PhEn quantifies the multiplicity and rate of variability associated with the physiological signals. It is sensitive to time-irreversibility. Therefore, it may serve as a promising tool for analysing physiological signals such as HRV.

Effect of sampling rate on acceleration and counts of hip- and wrist-worn ActiGraph accelerometers in children.

Abstract: Sampling rate (Hz) of ActiGraph accelerometers may affect processing of acceleration to activity counts when using a hip-worn monitor, but research is needed to quantify if sampling rate affects actual acceleration (mgs), when using wrist-worn accelerometers and during non-locomotive activities. OBJECTIVE To assess the effect of ActiGraph sampling rate on total counts/15 s and mean acceleration and to compare differences due to sampling rate between accelerometer wear locations and across different types of activities. APPROACH Children (n = 29) wore a hip- and wrist-worn accelerometer (sampled at 100 Hz, downsampled in MATLAB to 30 Hz) during rest/transition periods, active video games, and a treadmill test to volitional exhaustion. Mean acceleration and counts/15 s were computed for each axis and as vector magnitude. MAIN RESULTS There were mostly no significant differences in mean acceleration. However, 100 Hz data resulted in significantly more total counts/15 s (mean bias 4-43 counts/15 s across axes) for both the hip- and wrist-worn monitor when compared to 30 Hz data. Absolute differences increased with activity intensity (hip: r = 0.46-0.63; wrist: r = 0.26-0.55) and were greater for hip- versus wrist-worn monitors. Percent agreement between 100 and 30 Hz data was high (97.4%-99.7%) when cut-points or machine learning algorithms were used to classify activity intensity. SIGNIFICANCE Our findings support that sampling rate affects the generation of counts but adds that differences increase with intensity and when using hip-worn monitors. We recommend researchers be consistent and vigilantly report the sampling rate used, but note that classifying data into activity intensities resulted in agreement despite differences in sampling rate.

Chest electrical impedance tomography measures in neonatology and paediatrics-a survey on clinical usefulness.

Abstract: Objective Chest electrical impedance tomography (EIT) is currently applied in neonatal and paediatric patients, mainly within clinical studies. The findings of these studies imply that lung monitoring using EIT provides valuable information on regional lung ventilation and aeration at the bedside that might improve the therapy and care of this fragile patient population. In view of this postulated future use of EIT in neonatology and paediatrics we have conducted an international survey to assess the perceived usefulness of several measures derived from EIT examinations. Approach A questionnaire validating the clinical usefulness of 14 previously described EIT measures was designed and sent to 36 clinicians with previous experience with EIT in neonatal and paediatric patients. A numerical rating scale was used to assess the usefulness of each measure. Main results Thirty-four clinicians from 12 countries responded to the invitation and 32 filled in the questionnaire. The mean clinical and EIT experience (±SD) of the respondents was 19.4 ± 9.1 years and 7.7 ± 5.8 years, respectively. The top-rated measures were the global inhomogeneity index, silent spaces, change in end-expiratory lung impedance and ventrodorsal centre of ventilation. The bottom-rated were the regional respiratory time constant, tidal volume normalised to ml, respiratory rate and heart rate on the last rank. Significance The survey revealed that EIT measures characterising the ventilation and aeration distribution and the degree of their heterogeneity were deemed particularly useful. Respiratory rate, heart rate and overall tidal volume were considered less useful probably because these parameters are already routinely assessed by other conventional methods.

Evaluation of the signal quality of wrist-based photoplethysmography

Abstract: Objective Wearable devices with embedded photoplethysmography (PPG) sensors enable continuous monitoring of cardiovascular activity, allowing for the detection cardiovascular problems, such as arrhythmias. However, the quality of wrist-based PPG is highly variable, and is subject to artifacts from motion and other interferences. The goal of this paper is to evaluate the signal quality obtained from wrist-based PPG when used in an ambulatory setting. Approach Ambulatory data were collected over a 24 h period for 10 elderly, and 16 non-elderly participants. Visual assessment is used as the gold standard for PPG signal quality, with inter-rater agreement evaluated using Fleiss' Kappa. With this gold standard, 5 classifiers were evaluated using a modified 13-fold cross-validation approach. Main results A Random Forest quality classification algorithm showed the best performance, with an accuracy of 74.5%, and was then used to evaluate 24 h long ambulatory wrist-based PPG measurements. Significance In general, data quality was high at night, and low during the day. Our results suggest wrist-based PPG may be best for continuous cardiovascular monitoring applications during the night, but less useful during the day unless methods can be identified to improve low quality signal segments.

Imputation of partial pressures of arterial oxygen using oximetry and its impact on sepsis diagnosis.

Abstract: Background: The ratio of the partial pressure of arterial oxygen to fraction of inspired oxygen is a key component of the sequential organ failure assessment score that operationally defines sepsis. But, it is calculated infrequently due to the need for the acquisition of an arterial blood gas. So, we sought to find an optimal imputation strategy for the estimation of sepsis-defining hypoxemic respiratory failure using oximetry instead of an arterial blood gas. Methods: We retrospectively studied a sample of non-intubated acute-care patients with oxygen saturation recorded ≤10 minutes before arterial blood sampling (N=492 from 2013-2017). We imputed ratios of the partial pressure of arterial oxygen to the fraction of inspired oxygen and sepsis criteria from existing imputation equations (Hill, Severinghaus-Ellis, Rice, and Pandharipande) and compared them with the ratios and sepsis criteria measured from arterial blood gases. We devised a modified model-based equation to eliminate the bias of the results. Results: Hypoxemia severity estimates from the Severinghaus-Ellis equation were more accurate than those from other existing equations, but showed significant proportional bias towards under-estimation of hypoxemia severity, especially at oxygen saturations > 96%. Our modified equation eliminated bias and surpassed others on all imputation quality metrics. Conclusions: Our modified imputation equation, PaO_2=(23400/(1/(SpO_2 )-0.99))^(1/3), is the first one that is free of bias at all oxygen saturations. It resulted in ratios of partial pressure of arterial oxygen to fraction of inspired oxygen and sepsis respiratory criteria closest to those obtained by arterial blood gas testing and is the optimal imputation strategy for non-intubated acute-care patients.

Effects of different classroom temperatures on cardiac autonomic control and cognitive performances in undergraduate students.

Abstract: OBJECTIVE Indoor microclimate may affect students' wellbeing, cardiac autonomic control and cognitive performance with potential impact on learning capabilities. To assess the effects of classroom temperature variations on the autonomic profile and students' cognitive capabilities. APPROACH Twenty students attending Humanitas University School, (14M, age 21 ± 3 years) underwent a single-lead ECG continuous recording by a portable device during a 2 h lecture when classroom temperature was set 'neutral' (20 °C-22 °C, Day 1) and when classroom temperature was set to 24 °C-26 °C (Day 2). ECGs were sent by telemetry to a server for off-line analysis. Spectral analysis of RR variability provided indices of cardiac sympathetic (LFnu), vagal (HF, HFnu) and cardiac sympatho-vagal modulation (LF/HF). Symbolic analysis of RR variability provided the percentage of sequences of three heart periods with no significant change in RR interval (0V%) and with two significant variations (2V%) reflecting cardiac sympathetic and vagal modulation, respectively. Students' cognitive performance (memory, verbal comprehension and reasoning) was assessed at the end of each lecture using the Cambridge Brain Sciences cognitive evaluation tool. MAIN RESULTS Classroom temperature and CO2 were assessed every 5 min. Classroom temperatures were 22.4 °C ± 0.1 °C (Day 1) and 26.2 °C ± 0.1 °C (Day 2). Student's thermal comfort was lower during Day 2 compared to Day 1. HR, LF/HF and 0V% were greater during Day 2 (79.5 ± 12.1 bpm, 6.9 ± 7.1 and 32.8% ± 10.3%) than during Day 1 (72.6 ± 10.8 bpm, 3.4 ± 3.7, 21.4% ± 9.2%). Conversely, 2V% was lower during Day 2 (23.1% ± 8.1%) than during Day 1 (32.3% ± 11.4%). Short-term memory, verbal ability and the overall cognitive C-score scores were lower during Day 2 (10.3 ± 0.3; 8.1 ± 1.2 and 10.9 ± 2.0) compared to Day 1 (11.7 ± 2.1; 10.7 ± 1.7 and 12.6 ± 1.8). SIGNIFICANCE During Day 2, a shift of the cardiac autonomic control towards a sympathetic predominance was observed compared to Day 1, in the presence of greater thermal discomfort. Furthermore, during Day 2 reduced cognitive performances were found.

Vagal contributions to fetal heart rate variability: an omics approach.

Abstract: Objective Fetal heart rate variability (fHRV) is an important indicator of health and disease, yet its physiological origins, neural contributions, in particular, are not well understood. We aimed to develop novel experimental and data analytical approaches to identify fHRV measures reflecting the vagus nerve contributions to fHRV. Approach In near-term ovine fetuses, a comprehensive set of 46 fHRV measures was computed from fetal pre-cordial electrocardiogram recorded during surgery and 72 h later without (n = 24) and with intra-surgical bilateral cervical vagotomy (n = 15). Main results The fetal heart rate did not change due to vagotomy. We identify fHRV measures specific to the vagal modulation of fHRV: multiscale time irreversibility asymmetry index (AsymI), detrended fluctuation analysis (DFA) α 1, Kullback-Leibler permutation entropy (KLPE) and scale-dependent Lyapunov exponent slope (SDLE α). Significance We provide a systematic delineation of vagal contributions to fHRV across signal-analytical domains which should be relevant for the emerging field of bioelectronic medicine and the deciphering of the 'vagus code'. Our findings also have clinical significance for in utero monitoring of fetal health during surgery. Key points •Fetal surgery causes a complex pattern of changes in heart rate variability measures with an overall reduction of complexity or variability. •At 72 h after surgery, many of the HRV measures recover and this recovery is delayed by an intrasurgical cervical bilateral vagotomy. •We identify HRV pattern representing complete vagal withdrawal that can be understood as part of 'HRV code', rather than any single HRV measure. •We identify HRV biomarkers of recovery from fetal surgery and discuss the effect of anticholinergic medication on this recovery.

Finger and forehead PPG signal comparison for respiratory rate estimation.

Abstract: Objective An evaluation of the location of the photoplethysmogram (PPG) sensor for respiratory rate estimation is performed. Approach Finger PPG, forehead PPG, and respiratory signal were simultaneously recorded from 35 subjects while breathing spontaneously, and during controlled respiration experiments at a constant rate from 0.1 Hz to 0.6 Hz, in 0.1 Hz steps. Four PPG-derived respiratory (PDR) signals were extracted from each one of the recorded PPG signals: pulse rate variability (PRV), pulse width variability, pulse amplitude variability and the respiratory-induced intensity variability (RIIV). Respiratory rate was estimated from each one of the four PDR signals for both PPG sensor locations. In addition, different combinations of PDR signals, power distribution of the respiratory frequency range and differences of the morphological parameters extracted from both PPG signals have been analysed. Main results Results show better performance in terms of successful estimation and relative error when: (i) PPG signal is recorded in the finger; (ii) the respiratory rate is less than 0.4 Hz; (iii) RIIV signal is not considered. Furthermore, lower spectral power around the respiratory rate in the PDR signals recorded from the forehead was observed. Significance These results suggest that respiratory rate estimation is better at lower rates (0.4 Hz and below) and that the finger is better than the forehead to estimate respiratory rate.