Objective biomarkers for differential diagnosis in psychiatry are still scarce. Voice atypicalities characterize two prominent, often co-occurring psychiatric disorders: schizophrenia-spectrum disorders (SSD) and major depressive disorders (MDD). Given that voice recordings can be easily obtained, advanced speech analysis might facilitate the development of diagnostic biomarkers for SSD and MDD. Speech was recorded from a transdiagnostic sample comprising 47 SSD patients, 62 MDD patients, and 41 healthy controls (HC), during three different tasks: a semi-structured interview, a reading task and an empathy task. We evaluated the discriminative power of standardized speech parameters and compared the performance of the three tasks. The extended Geneva Acoustic Minimalistic Parameter Set (eGeMAPS) was extracted using openSMILE and fed into random forest (RF) algorithms with 10-fold crossvalidation. Model performances were evaluated using accuracy, F1-score, precision, and recall. Importance of specific predictors was assessed using Gini importance. In this three-class problem, a simple 1-minute video task reached best results with 57% accuracy. The acoustic parameters revealed distinct vocal profiles associated with each disorder. Considering the chance probability of 33%, our results show that automated speech analysis could predict diagnostic classes with good to high accuracy.

Acknowledging the vital importance of fatigue management for improving rehabilitation results, customizing treatment, safeguarding patient well-being, and enhancing the quality of life of hemiplegic patients, this study presents the development of a tailored fatigue model and a corresponding human-in-the-loop (HiL) control system for exoskeleton-assisted walking. For this, the selected three-compartment controller fatigue model including a resting recovery parameter was adapted to a dynamic walking task scenario, incorporating a torque–velocity–angle dependency to quantify muscle activity. The model parameters were experimentally verified in a study with six healthy subjects, demonstrating accurate prediction of maximum voluntary contraction (MVC) decline with an average mean absolute error of 4.9%MVC. Subsequently, an HiL control mechanism was developed, utilizing ratings of perceived fatigue and state of fatigue values as reference metrics. The presented control approach effectively regulates fatigue levels within a 0%MVC–6%MVC steady-state error range during simulations. Experimental validation confirmed this performance, however, with partly higher steady-state errors mainly due to the restrictions of the exoskeleton's assistance. This preliminary study provides a promising foundation for future research, demonstrating the potential to manage fatigue effectively in exoskeleton users, offering an improved, personalized experience.

Atrial fibrillation is a prevalent cardiac arrhythmia, significantly increasing the risk of stroke, heart failure, and mortality. Early detection, especially during asymptomatic and paroxysmal stages, is essential for effective intervention. This study explores the application of deep neural networks in simplified ECG screening to enhance population-wide detection of atrial fibrillation. A handheld device, MyDiagnostick, was employed for large-scale ECG data acquisition within a pharmacy-based clinical trial on 7295 subjects aged 65 years and older. Automated diagnosis yielded 6.08% of AF prevalence in the given dataset. The data were then analyzed using a validated deep neural network model for the detection of cardiac arrhythmia in 12-lead ECG data for feature extraction and detection of atrial fibrillation. In addition, we investigate the capabilities of explainable artificial intelligence to provide diagnostic support for cardiologists and assess the feasibility of implementing deep neural networks in wearable devices for continuous monitoring. The study also emphasizes the importance of interpretability in artificial intelligence models for medical applications, leveraging explainable artificial intelligence to highlight ECG segments indicative of atrial fibrillation. Our findings demonstrate the efficacy of deep neural networks in atrial fibrillation detection with an F1-score of 86% vs. 81% of the automated ECG stick analysis and the potential for their integration into wearable technology by successfully reducing the number of weights by 99% without significant loss of accuracy, providing a robust tool for early diagnosis and continuous monitoring of atrial fibrillation.

Regional perfusion monitoring, often performed by pulmonary perfusion computed tomography, is vital in intensive care units. Electrical impedance tomography, repeatable and of non-invasive nature, could provide an attractive alternative. This study compares non-enhanced and contrast-enhanced electrical impedance tomography to computed tomography under induced central to peripheral lung perfusion impairments and cardiac output modulation in 11 animals. A new algorithmic framework using multi-compartment modeling and tracer kinetics was developed to improve perfusion estimation. A multi-resolution mixed models analysis shows electrical impedance tomography agrees poorly with computed tomography in static monitoring, with limits of agreement exceeding relative errors of 100%. For trend tracking,contrast-enhancement with 5.85% NaCl yielded concordance rates above 80%, and over 90% for peripheral impairments, emerging as a robust tracker of coarse to fine perfusion changes. Non-enhanced electrical impedance tomography peaked around 60% under central impairment and cardiac modulation, proving to be less reliable.

Monitoring cardiac and respiratory activity, along with the overall physiological state of patients, is crucial in emergency rooms, especially when patients are waiting for further treatment without constant supervision. Unobtrusive sensing technology supports effective patient care by enabling seamless deployment of monitoring systems. This contribution introduces a multi-redundant sensor system for unobtrusive heart activity measurement integrated into an emergency bed. The system includes capacitively coupled electrocardiographic (cECG) electrodes and ballistocardiographic (BCG) sensors in the mattress, and a foot strap with an integrated reflective photoplethysmographic (rPPG) sensor. These sensors provide robust cardiac activity data and enable the extraction of blood pressure-related parameters such as pre-ejection period (PEP), pulse transit time (PTT), and pulse arrival time (PAT). A proof-of-concept study with four healthy volunteers validated the system, showing an RMSE of 0.058 s for RR intervals and 93.6% cardiac activity coverage.

At Present, cameras are increasingly used to measure physiological signals from human face and body for contactless health monitoring, thereby eliminating mechanical contact with the skin that are common in wearable sensors. This is an emerging research direction developing rapidly in the last decade and which is now gradually maturing into products for patient monitoring. Advancements in biomedical optics, physiological measurement, computer vision and artificial intelligence (AI) enabled various camera-based measurements, including vital signs like heart rate (HR), respiration rate (RR), oxygen saturation (SpO2), blood pressure (BP), and physiological markers that have diagnostic capabilities, such as the detection of arrhythmia, atrial fibrillation, apnea, hypertension, etc. Image and video analysis also permit the measurement of human semantics, context and behaviours that provide new insights into health informatics (e.g., facial analysis and body actigraphy for the assessment of patient delirium), which is a unique advantage of camera sensors as compared to biomedical sensors, like e.g., photoplethysmography (PPG) and electrocardiogram (ECG). Camera-based health monitoring will bring a rich set of compelling healthcare applications that directly improve upon contact-based monitoring solutions in various scenarios like clinical units including e.g., the intensive care unit (ICU), the neonatal ICU (NICU) or sleep centers, and assisted-living homes (e.g., elderly homes or confinement centers), improving patient care experience and people's quality of life.

The unobtrusive monitoring of vital parameters has recently gained significant importance in home care and the automotive industry to supervise the driver's awareness. In wearable devices, contact-based photoplethysmography (PPG) enables the estimation of heart rate, respiratory rate, blood pressure, and peripheral arterial oxygen saturation through the optical acquisition of blood volume variations in the skin. This work analyzes the feasibility of unobtrusive acquisition of the PPG signal through clothing using Monte Carlo simulations. For that, we suggest a male and a female model of the human back skin covered with six interchangeable clothing items with different tissue, color, and thickness. The arterial pulsation is simulated through blood volume variation in a modeled vessel during systole, diastole, and end-diastole. The absorption coefficients of each clothing were measured with an optical power meter and modelled accordingly. Nine increasing distances of 0 to 2.95 cm between a red (760 nm) or infra-red (900 nm) Light Emitting Diode (LED) and a photodiode (PD) are investigated. The quality of the PPG signal is evaluated by considering the detected light power, the order of the detected light intensities at systole, diastole and end-diastole, and the perfusion index. The simulations revealed that the ability to detect the PPG pulse through clothing depends on the thickness and type of fabric used, the LED-PD distance of the sensor, and its wavelength. The optimal LED-PD distance was found to be 1.2 to 1.8 cm. The infrared LED generally generates reproducible and better quality PPG signals for cotton and polyamide T-shirts, but not for a thick gray cotton hoodie and a thin acrylic red sweater. Besides, significant differences are reported between the male and the female models. Our results evidence the limitations of PPG monitoring through clothes and suppose a step forward in unobtrusive sensors' design for hospitals and smart cars.

Thoracic electrical impedance tomography provides a non-invasive signal that combines changes from both cardiac and respiratory events. Heart and respiratory rates are important physiological parameters to diagnose and monitor cardiorespiratory conditions, and necessary for further more sophisticated analyses. We propose a real-time, model-based approach in the time domain to estimate both heart and respiratory rates simultaneously from the electrical impedance tomography signal. We tested our method using simulated non-stationary signals with varying relative cardiac and respiratory contributions. Our method could accurately estimated heart and respiratory rates with root mean square errors as low as 0.54 and 0.86 cycles per minute, and Bland-Altman biases as low as 0.04 ± 1.04 and -0.07 ± 1.53 cycles per minute, respectively. Additionally, the algorithm ran in real-time with or without parallelization, reaching computation times of 10 ± 1.1 ms.

The classification of Electrocardiograms (ECG) sig-nals is vital for diagnosing and monitoring cardiovascular diseases. Recently, there has been increasing interest in leveraging efficient machine learning models for ECG classification. Op-timizing computational resources in this domain is crucial for real-time analysis and reducing power consumption. This paper explores two approaches to enhance the efficiency of ECG classification on edge devices. Firstly, a smart lead-switching mechanism intelligently switches between a high-power model and a low- power model. Secondly, Apache Tensor Virtual Machine (TVM) is employed to optimize the implementation of the developed models on the Jetson Nano single-board computer. Experimental results demonstrate noticeable achievements, including up to 96% reduction in FLOPS and 95% reduction in inference time. Our results hold significant promise for improving ECG classification on edge devices, enabling early detection of abnormalities, and enhancing patient care.

The determination of step length, an important gait parameter, has been a challenging task. Although unobtrusive sensors (inertial measurement units) have been developed recently, they cannot facilitate the automatic estimation of step length. In this article, we use a model-based technique to determine the step length using the Unscented Kalman Filter with angular velocity from a gyroscope inside the thigh pocket. We then propose a novel covariance estimation algorithm based on a screening technique that performs a search for the optimal Process Noise Covariance matrix. Upon implementing the Unscented Kalman Filter, the step length is found using the horizontal position of the foot relative to the hip using a patient-independent robust peak detection algorithm. This research article paves the way for algorithms that are computationally much faster than black box methods, with more scope for the development of better algorithms for covariance estimation using the one proposed in this article as a foundation.

Objective: For driver state estimation, physiological features might be promising input parameters. As cable-bound sensing of these parameters is impractical for ubiquitous monitoring, the measurement certainly has to be based on unobtrusive and contact-free technologies. In this work, unobtrusive methods for heart rate (HR) and respiration rate (RR) monitoring, including a hybrid imaging approach, are evaluated under simulated and real driving conditions. Methods: The feasability of unobtrusive methods was tested by comparing measurements from unobtrusive sensors to reference sensors. Under laboratory conditions, magnetic induction and photoplethysmography, both integrated into the seat belt, and hybrid imaging, combining visual and thermal imaging, were evaluated for RR sensing. In real driving, creating an urban and a rural scenario, sensing of RR by hybrid imaging and sensing of HR by a seat-integrated capacitive ECG were evaluated. Results: Under laboratory conditions, a reliable RR detection was possibly using all three sensor technologies. In real-world driving, a reliable HR and RR detection was possible during the rural scenario. In the urban scenario, only the RR detection was feasible. Due to motion artifacts, the capacitive ECG was disturbed and the HR detection impaired. Conclusion: The evaluated unobtrusive measurement systems can monitor physiological parameters during e.g. long-time driving on highways, but may not yet be feasible for monitoring during agile inner-city driving situations, due to motion artifacts. Therefore, future work should focus on artifact reduction. Significance: Physiological features might be used as input parameters for driver state estimation systems. This work presents unobtrusive sensing methods for these parameters.

A driving study was performed to induce stress with 24 participants performing different inducement tasks (n-back task, Sing-a-Song Stress Test and noise exposure). Both performance-based measures (Tactile Detection Response Task) as well as subjective measures were recorded to assess the driver state. Subjective ratings indicate that stress was most successfully induced on a group level with the n-back task with high inter-individual variation. The average response times doubled during the n-back task for a simultaneously performed tactile detection response task compared to baseline response times. Sympathetic nervous activation resulting in the increase of heart rate, respiration rate and decrease in heart rate variability (RMSSD) was found as a physiological reaction on stress-inducing secondary tasks. The most prominent physiological responses were found during the modified Sing-a-Song Stress Test. Subjective ratings on the perceived stress level and physiological response rarely correlated. This study provided reference data for driver state algorithm development in the EU-funded project ADAS&ME.

In the wake of Big Data, traditional Machine Learning techniques are now often integrated in the clinical workflow. Despite more capable, Deep Learning methods are not equally accepted given their unsatiated need for great amounts of training data and transversal use of the same architectures in fundamentally different areas with weakly-substantiated adaptations. To address the former, a cardiorespiratory signal synthesizer was designed by conditional sampling from a multimodally trained stochastic system of Gaussian copulas integrated in a Markov chain. With respect to the latter, a multi-branch convolutional neural network architecture was conceived to learn the best cardiac sensor-fusion strategy at every abstraction layer. The network was tailored to the tasks of cycle detection and classification for different cardiac modality combinations by a synthesizer-based data augmentation training framework and Bayesian hyperparameter optimization. The synthesizer yielded highly realistic signals in the time, frequency and phase domains for both healthy and pathological heart cycles as well as artifacts of different modalities. Benchmarking suggested that the network is able to surpass previous architectures and data augmentation provided a performance boost in realistic data availability scenarios. These included insufficient training data volume, as low as 150 cycles long, artifact contamination and absence of a classification data type in training.

The papers in this special section focus on camera-based monitoring for pervasive healthcare informatics. Measuring physiological signals from the human face and body using video cameras is an emerging research topic that has grown rapidly in the last decade. Remote cameras (in both visible and infrared wavelengths) can be used to measure vital signs from a human body based on skin optics or body movements thereby avoiding mechanical contact with the skin. Camera-based health monitoring will bring a rich set of compelling healthcare applications that directly improve upon contact-based monitoring solutions and impact people’s care experience and quality of life in various scenarios, such as in hospital care units, sleep/senior centers, assisted-living homes, telemedicine and e-health, baby/elderly care at home, fitness and sports, driver monitoring in automotive applications, cardiac/ respiratory gating for MRI/CT, AR/VR based therapy and clinical training, e

OBJECTIVE: Geriatric patients, especially those with dementia or in a delirious state, do not accept conventional contact-based monitoring. Therefore, we propose to measure heart rate (HR) and heart rate variability (HRV) of geriatric patients in a noncontact and unobtrusive way using photoplethysmography imaging (PPGI). METHODS: PPGI video sequences were recorded from 10 geriatric patients and 10 healthy elderly people using a monochrome camera operating in the near-infrared spectrum and a colour camera operating in the visible spectrum. PPGI waveforms were extracted from both cameras using superpixel-based regions of interests (ROI). A classifier based on bagged trees was trained to automatically select artefact-free ROIs for HR estimation. HRV was calculated in the time-domain and frequency-domain. RESULTS: an RMSE of 1.03 bpm and a correlation of 0.8 with the reference was achieved using the NIR camera for HR estimation. Using the RGB camera, RMSE and correlation improved to 0.48 bpm and 0.95, respectively. Correlation for HRV in the frequency-domain (LF/HF-ratio) was 0.50 using the NIR camera and 0.70 using the RGB camera. CONCLUSION: We were able to demonstrate that PPGI is very suitable to measure HR and HRV in geriatric patients. We strongly believe that PPGI will become clinically relevant in monitoring of geriatric patients. SIGNIFICANCE: we are the first group to measure both HR and HRV in awake geriatric patients using PPGI. Moreover, we systematically evaluate the effects of the spectrum (near-infrared vs. visible), ROI, and additional motion artefact reduction algorithms on the accuracy of estimated HR and HRV.

We present a novel application of the variable stiffness actuator (VSA)-based assistance/rehabilitation robot-featured impedance control using a cascaded position torque control loop. The robot follows the adaptive impedance control paradigm, thereby achieving an adaptive assistance level according to human joint torque. The feedforward human joint torque command is used to cooperatively adjust the impedance controller and the stiffness trajectory of the VSA (this functional architecture is referred to as the cooperative control framework). In this way, the task performance during movement training can be improved regarding: 1) safety-for example, when the subject intends to contribute considerable effort, low-gain impedance control is activated with a low stiffness actuator to further decrease output impedance and 2) tracking performance-for example, for the subject with less effort, high-gain impedance control is used while pursuing high stiffness to enhance the torque bandwidth. Regarding the safety aspect, we demonstrate that the torque controller designed at low stiffness can be sensitive to the disturbance for low output impedance while maintaining tracking performance. A precondition for this is to treat the input disturbance separately. This is guaranteed by our previously proposed torque control of the VSA using the linear quadratic Gaussian technique. This approach is also employed here, but with additional discussion on the observer design to serve the proposed cooperative control approach. Here, the effectiveness of the proposed control system is experimentally verified using a VSA prototype and a one-degree-of-freedom lower limb exoskeleton worn by a human test person. Note to Practitioners-Control of “physical human-robot interaction” can be achieved by the mechanical parts of the variable stiffness actuator (VSA). However, the mechanical construction for stiffness variation may limit the capacity to achieve low output stiffness and fast stiffness variation in speed. These limitations may become more evident in the assistance/rehabilitation robot applications. To overcome these limitations, the impedance control scheme can be employed to achieve a programmable impedance range and impedance variation speed. This control scheme has been widely applied on the fixed-compliance joint but lacks a way to be implemented on the VSA joint because of its existing capacity to control the impedance with the mechanical construction. This article presents a novel application of the impedance-controlled VSA used on a lower limb robot. We describe how to adjust the actuator stiffness to cooperatively work with the adaptive impedance control scheme. Based on our approach, the robot with the impedance-controlled VSA joint can extend the capacity of bandwidth and low output impedance. This is an improvement on the impedance-controlled fixed-compliance joint. The cooperative control framework presented here was tested on an exoskeleton system with two healthy test persons and is also applicable to other actuator prototypes. Future research aims to employ this system for actual patient training.

While bed-integrated ballistocardiography (BCG) has potential clinical applications such as unobtrusive monitoring of patients staying in the general hospital ward, it has so far mainly gained interest in the wellness domain. In this article, the potential of BCG to monitor hospitalized patients after surgical intervention was assessed. Long-term BCG recordings (mean duration 17.7 h) of 14 patients were performed with an EMFit QS bed sensor. In addition, ten healthy subjects were recorded during sleep (mean duration 7.8 h). Using an iterative algorithm, beat-to-beat intervals (BBIs) and the ultra-short-term heart-rate-variability (HRV) parameters standard deviation of NN intervals (SDNN) and root mean square of successive differences (RMSSD) were estimated and compared to an ECG reference in terms of average estimation error and temporal coverage. While the absolute BBI estimation error was found to be higher when full-day patient data was used (16.5 ms), no significant difference between healthy subjects (12.7 ms) and patient nighttime data (11.0 ms) was observed. Nevertheless, temporal coverage of BBI estimation was significantly lower in patients (39.3% overall, 51.7% at night) compared to the healthy sleepers (73.2%). This resulted in reduced HRV estimation coverage (9.7% vs. 37.2%) at comparable estimation error levels.

Stride time variability is an important indicator for the assessment of gait stability. An accurate extraction of the stride intervals is essential for determining stride time variability. Peak detection is a commonly used method for gait segmentation and stride time estimation. Standard peak detection algorithms often fail due to additional movement components and measurement noise. A novel algorithm for robust peak detection in inertial sensor signals was proposed in a previous contribution. In this work, we present a novel approach for estimation of stride time variability based on the formerly proposed peak detection algorithm applied to an unobtrusive sensor setup for motion monitoring. The unobtrusive sensor setup includes a wrist sensor, a pocket or belt sensor, and a necklace sensor, all equipped with both accelerometer and gyroscope. The goal of this work is to implement a generalized approach for accurate and robust stride interval determining algorithm for different sensor locations. Therefore, treadmill and level ground walking experiments were conducted with ten healthy subjects at increasing walking speeds and an age-simulating suit. With the proposed algorithm, we achieved a RMSE of 0.07 s for the stride interval estimation during treadmill walking experiments. The results give promising indications that detection of variation of stride time variability is possible using the proposed unobtrusive sensor setup.

We're very excited about this year's meeting, which is being held on the High Tech Campus in Eindhoven, the Netherlands, a sparkling location originally owned by Philips Research. Due to the open innovation strategy, it has been turned into a rich biotope of R&D companies working in close cooperation with neighboring institutions, universities (including TU Eindhoven and RWTH Aachen University) and academic hospitals.

In this work, we present an in-ear system for physiological and psychological stress detection based on photoplethysmography, acceleration, and temperature measurements. The complete system is used to extract vital signs from healthy subjects, who are exposed to psychologically demanding tasks. The newly developed sensor system is integrated into our IPANEMA body sensor network and, thus, can be used in combination with several sensor modalities. The capability of the stress level estimation is validated in an human stress experiment. To obtain information on the current stress level, several well-known indicators are utilized like the heart rate variability, surgical stress index or the Oliva and Roztocil index.

Human slow wave sleep (SWS) during bedtime is paramount for energy conservation and memory consolidation. This study aims at automatically detecting SWS from nocturnal sleep using cardiorespiratory signals that can be acquired with unobtrusive sensors in a home-based scenario. From the signals, time-dependent features are extracted for continuous 30-s epochs. To reduce the measuring noise, body motion artifacts, and/or within-subject variability in physiology conveyed by the features, and thus, enhance the detection performance, we propose to smooth the features over each night using a spline fitting method. In addition, it was found that the changes in cardiorespiratory activity precede the transitions between SWS and the other sleep stages (non-SWS). To this matter, a novel scheme is proposed that performs the SWS detection for each epoch using the feature values prior to that epoch. Experiments were conducted with a large dataset of 325 overnight polysomnography (PSG) recordings using a linear discriminant classifier and tenfold cross validation. Features were selected with a correlation-based method. Results show that the performance in classifying SWS and non-SWS can be significantly improved when smoothing the features and using the preceding feature values of 5-min earlier. We achieved a Cohen's Kappa coefficient of 0.57 (at an accuracy of 88.8%) using only six selected features for 257 recordings with a minimum of 30-min overnight SWS that were considered representative of their habitual sleeping pattern at home. These features included the standard deviation, low-frequency spectral power, and detrended fluctuation of heartbeat intervals as well as the variations of respiratory frequency and upper and lower respiratory envelopes. A marked drop in Kappa to 0.21 was observed for the other nights with SWS time of less than 30 min, which were found to more likely occur in elderly. This will be the future challenge in cardiorespiratory-based SWS detection.

In clinical practice the determination of the heart rate variability (HRV) has become a common measure to investigate the parasympathetic cardiac control. Especially the measurement of the respiratory sinus arrhythmia (RSA) has gained importance to asses the HRV. The RSA can be seen as an indirect parameter for the physiological or psychological stress the patient is currently exposed to. Thus, this parameter is used to identify specific characteristics of disease in a broad field of clinical disciplines. In this contribution, we present a BSN-based approach of assessing the RSA in a long-term evaluation. For this purpose, we use two sensor types: A three channel ECG sensor node which was introduced before and a recently developed respiratory sensor based on conductive yarn. We further implemented an oscillatory model-based Unscented Kalman filter (UKF) to estimate the heart rate as well as the breathing rate and, thus, to calculate the RSA. The algorithm is finally validated by performing deep breathing tests (DBT) on a healthy test subject in order to force an increased occurrence of the RSA. The results of the developed system and proposed algorithm are finally discussed with respect to its applicability in different every days situations.

The accurate, real-time estimation of biomechanical joint parameters bears a potential benefit for many applications. Examples include the assessment of training success in movement therapy, the use as a quantitative clinical scale for joint rigidity or the use in the derivation of control parameters for active, intelligent orthotic or prosthetic devices. Such a realtime assessment system should be as unobtrusive as possible, minimising instrumentation effort for the user or the clinical staff. Towards this goal we have build a body sensor network (BSN) that is able to measure surface electromyogram and 9-degrees of freedom inertial/magnetic data at high sample rates. The measured data is preprocessed and subsequently used in an Unscented Kalman Filter in a model-based approach employing the nonlinear dynamics of the human knee kinematics. The derivation of biomechanical joint parameters, in our case the knee stiffness, can then be readily obtained from the nonlinear model. To validate BSN measurements, we present a novel test-bench and its corresponding nonlinear model. The biomechanical parameter estimator is validated in pendulum like motions on the test-bench and in experiments where the test subject is undergoing co-activation of extensor and flexor muscles acting on the knee.

Spasticity is a common disorder of the skeletal muscle with a high incidence in industrialised countries. A quantitative measure of spasticity using body-worn sensors is important in order to assess rehabilitative motor training and to adjust the rehabilitative therapy accordingly. We present a new approach to spasticity detection using the Integrated Posture and Activity Network by Medit Aachen body sensor network (BSN). For this, a new electromyography (EMG) sensor node was developed and employed in human locomotion. Following an analysis of the clinical gait data of patients with unilateral cerebral palsy, a novel algorithm was developed based on the idea to detect coactivation of antagonistic muscle groups as observed in the exaggerated stretch reflex with associated joint rigidity. The algorithm applies a cross-correlation function to the EMG signals of two antagonistically working muscles and subsequent weighting using a Blackman window. The result is a coactivation index which is also weighted by the signal equivalent energy to exclude positive detection of inactive muscles. Our experimental study indicates good performance in the detection of coactive muscles associated with spasticity from clinical data as well as measurements from a BSN in qualitative comparison with the Modified Ashworth Scale as classified by clinical experts. Possible applications of the new algorithm include (but are not limited to) use in robotic sensorimotor therapy to reduce the effect of spasticity.

Electromyography (EMG) is used as medical tool to display muscle activity and gain information about the health status of the patients muscle function, which may be affected by many kind of diseases. Spasticity is caused by injuries of the central nervous system, which may occur in consequence of stroke or as concomitant of multiple sclerosis. If the muscle function is influenced by spasticity, there are different types of therapy to regain muscle control. For robotic supported rehabilitation, such as provided by diverse exoskeleton applications, it is important to identify spastic muscle activity patterns, in order to protect patients against mechanical injury. Therefore the EMG data of a hemiplegic patient was analysed, in order to find characteristic features of affected muscle activity and combine them to a characteristic feature vector. To classify the different states of muscle activity a Support Vector Machine (SVM) is used, trained with the feature vector space, which was created from the given EMG data. After that, the developed SVM was applied to data sets of patients also affected by spasticity in order to compare the obtained results to those estimated by a previously used algorithm for spasticity detection. Subsequently, the recognition capability of the implemented SVM was validated by a newly developed EMG sensor node for the IPANEMA Body Sensor Network (BSN).

In this work, we report on human trials with the MedIT in-ear photoplethysmography (PPG) measurement system. The system is evaluated with healthy subjects and people suffering from heart insufficiency, respectively. Physiological heart activity can be measured with a minimal error of 1.2 heartbeats per minute and a regression coefficient of 0.9975 compared with standard ECG. Respiration related information was extracted by combining PPG amplitude analysis and car-diorespirational coupling (cardiorespiratory sinus arrhythmia). The moments of inspiration and expiration were estimated with a Naive Bayes' classifier with high sensitivity and specificity of 81,4% and 86%, respectively. For automatic cardiological alarming, a feature space is defined that clearly demonstrates the separability of normal heart rhythm and heart insufficiency. The results demonstrate a promising perspective for a mobile and long-term cardiorespiratory monitoring and alarming with an unobtrusive and inexspensive PPG measurement technique that is fully compatible to modern communication devices.

The aim of this paper is to present and evaluate algorithms for heartbeat interval estimation from multiple spatially distributed force sensors integrated into a bed. Moreover, the benefit of using multichannel systems as opposed to a single sensor is investigated. While it might seem intuitive that multiple channels are superior to a single channel, the main challenge lies in finding suitable methods to actually leverage this potential. To this end, two algorithms for heart rate estimation from multichannel vibration signals are presented and compared against a single-channel sensing solution. The first method operates by analyzing the cepstrum computed from the average spectra of the individual channels, while the second method applies Bayesian fusion to three interval estimators, such as the autocorrelation, which are applied to each channel. This evaluation is based on 28 night-long sleep lab recordings during which an eight-channel polyvinylidene fluoride-based sensor array was used to acquire cardiac vibration signals. The recruited patients suffered from different sleep disorders of varying severity. From the sensor array data, a virtual single-channel signal was also derived for comparison by averaging the channels. The single-channel results achieved a beat-to-beat interval error of 2.2% with a coverage (i.e., percentage of the recording which could be analyzed) of 68.7%. In comparison, the best multichannel results attained a mean error and coverage of 1.0% and 81.0%, respectively. These results present statistically significant improvements of both metrics over the single-channel results (p < 0.05).

A mobile device is presented for monitoring both respiration and pulse. The device is developed as a bendable/flexible inlay that can be placed in a shirt pocket or the inside pocket of a jacket. To achieve optimum monitoring performance, the device combines two sensor principles, which work in a safe noncontact way through several layers of cotton or other textiles. One sensor, based on magnetic induction, is intended for respiratory monitoring, and the other is a reflective photoplethysmography sensor intended for pulse detection. Because each sensor signal has some dependence on both physiological parameters, fusing the sensor signals allows enhanced signal coverage.

This paper presents a mobile device for monitoring respiration and pulse. The device is realized as a bendable inlay, which can be put into a shirt pocket or the inside pocket of a jacket. In order to achieve optimum monitoring performance, the device combines two sensor principles both working in a noncontact way through several layers of cotton or other textiles. One sensor is based on magnetic induction and is intended for respiratory monitoring, the other one is a reflective photoplethysmographic sensor intended for pulse detection. Furthermore, since each sensor signal shows a certain dependence on both physiological parameters, fusing the sensor signals renders the possibility of signal coverage enhancement.

For many years, pulse oximetry has been widely used in the clinical environment for a reliable monitoring of oxygen saturation ( SpO2) and heart rate. But since common sensors are mainly placed to peripheral body parts as finger or earlobe, it is still highly susceptible to reduced peripheral perfusion, e.g., due to centralization. Therefore, a novel in-ear pulse oximetric sensor (placed against the tragus) was presented in a prior work which is deemed to be independent from perfusion fluctuations due to its proximity to the trunk. Having demonstrated the feasibility of in-ear SpO2 measurement with reliable specificity in a laboratory setting, we now report results from a study on in-ear SpO2 in a clinical setting. For this, trials were performed on 29 adult patients undergoing surgery. In-ear SpO2 data are compared with SaO2 data obtained by blood gas analysis, and with three reference pulse oximeters applied to the finger, ear lobe, and forehead. In addition, we derived an SpO2-independent perfusion index by means of the wavelengths used. The feasibility and robustness of in-ear SpO2 measurement is demonstrated under challenging clinical conditions. SpO2 shows good accordance with SaO2, a high level of comparability with the reference pulse oximeters, and was significantly improved by introducing a new algorithm for artifact reduction. The perfusion index also shows a good correlation with the reference data.

Contactless vital sign measurement technologies often have the drawback of severe motion artifacts and periods in which no signal is available. However, using several identical or physically different sensors, redundancy can be used to decrease the error in noncontact heart rate estimation, while increasing the time period during which reliable data are available. In this paper, we show for the first time two major results in case of contactless heart rate measurements deduced from a capacitive ECG and optical pulse signals. First, an artifact detection is an essential preprocessing step to allow a reliable fusion. Second, the robust but computationally efficient median already provides good results; however, using a Bayesian approach, and a short time estimation of the variance, best results in terms of difference to reference heart rate and temporal coverage can be achieved. In this paper, six sensor signals were used and coverage increased from 0-90% to 80-94%, while the difference between the estimated heart rate and the gold standard was less than ±2 BPM.

The current rise in popularity of ballisto-cardiography-related research has led to the development of new sensor concepts and recording methods. Measuring the ballistocardiogram using bed mounted pressure sensors opens up new possibilities for home monitoring applications. The signals measured with these sensors contain a mixture of cardiac and respiratory components, which can be used for detection of comorbidities of heart failure like apnea or arrhythmia. However, the separation of the cardiac and respiratory components has proven to be difficult, since there is significant overlap in the spectra of both components. In this paper, an algorithm for the separation task is presented, which can overcome the problem of overlapping spectra. Additionally, a model has been developed for the generation of artificial ballistocardiograms, which are used to analyze the separation performance. Furthermore, the algorithm is tested on preliminary data from a clinical study.

The drainage control of cerebrosprinal fluid, as needed in the treatment of increased intracranial pressure, is to this day achieved manually by using an external ventricular drainage system. The manual control procedure bears several risks for the patient among which the rapid decrease of intracranial pressure induced by patient's upper body inclination angle change is the most prominent. An automatic controller is suggested to increase patient's safety and the quality of the therapy by delivering a continuous pressure control and the rejection of disturbances like cerebrospinal fluid production rate and inclination angle changes. In this contribution an intracranial pressure controller is designed using the robust control methodology and subsequently validated in nonlinear simulations and in a hydrodynamic human cerebral simulator. The controller is designed using a mixed uncertainty approach accounting for uncertainties in the technical and the physiological system. A self-scheduled implementation of the controller guarantees the compensation of the nonlinear input function for the process, being of Hammer stein structure. The controller shows stable response in simulations and Mock experiments for various operating points and disturbance rejection tests.

For a clinical application, oxygenation and etCO2are required to be regulated to a specific value for minimizing the risk of hypoxia and hypercapnia or hypocapnia. To realize these tasks, a knowledge-based controller, a fuzzy controller and a model-based H∞controller are proposed in this article for controlling the complex nonlinear cardiopulmonary system. The selection of controlled variables is critical for each control objective: PEEP and FiO2for oxygenation and minute ventilation (MV) for etCO2. In this article, the new and concrete results of animal experiments are presented for a knowledge-based controller and a fuzzy controller. In addition, a model-based approach using H∞loop-shaping technique is proposed for the control of etCO2as a pilot study. System identification is carried out to determine the model structure of the porcine dynamics. Based on a numerical study, the second order system with one zero describes the system with the best RMSE criteria, corresponding to the model obtained by human models. Based on the simulation result, the model-based H∞loop-shaping technique is a distinguished approach for the control of etCO2.

We present a study on the feasibility of the automatic detection of atrial fibrillation (AF) from cardiac vibration signals (ballistocardiograms/BCGs) recorded by unobtrusive bedmounted sensors. The proposed system is intended as a screening and monitoring tool in home-healthcare applications and not as a replacement for ECG-based methods used in clinical environments. Based on BCG data recorded in a study with 10 AF patients, we evaluate and rank seven popular machine learning algorithms (naive Bayes, linear and quadratic discriminant analysis, support vector machines, random forests as well as bagged and boosted trees) for their performance in separating 30 s long BCG epochs into one of three classes: sinus rhythm, atrial fibrillation, and artifact. For each algorithm, feature subsets of a set of statistical time-frequency-domain and time-domain features were selected based on the mutual information between features and class labels as well as first- and second-order interactions among features. The classifiers were evaluated on a set of 856 epochs by means of 10-fold cross-validation. The best algorithm (random forests) achieved a Matthews correlation coefficient, mean sensitivity, and mean specificity of 0.921, 0.938, and 0.982, respectively.

This paper presents a newly developed Transcutaneous Energy Transfer (TET) System to supply an electromechanical implant with energy. The system is capable of delivering a power of 1-5 W to the implant over a distance of up to 5 cm via an inductive link with a frequency of 100 kHz. Additionally, the inductive link incorporates a data link which allows transmission of measurement data and information regarding the link quality. Because of the integration of power transfer and data transfer the system is thus energy saving in comparison to most TET systems, which often need an additional second dedicated radio communication channel. For the data transmission from the energy transmitter to the implant frequency shift keying and from the implant to the energy transmitter load modulation has been implemented.

This study investigates the feasibility of using bioelectrical impedance measurements to \rev{detect} the body fluid status. The multi-frequency impedance measurements were performed in combination with an animal experiment with five female pigs. For this purpose, the fluid balances of these animals, which were connected to an extracorporeal membrane oxygenation, were recorded. The ECMO circuit needs a high blood flow from the venous system and in order to avoid vasoconstriction in the femoral vein, blood-thinning infusions were injected. The quantity of injected fluid and the quantity of urine were recorded to monitor the fluid balance of each animal. These balances were compared with the intracellular and extra cellular tissue resistance, which was measured by bioelectrical-impedance spectroscopy. The experimental results strongly support the clinical benefit of the BIS for the assessment of changes in the hydration status.

Due to the demographic change and increased cost pressure, contactless measurement methods such as capacitive ECG become more and more interesting in research and industry. As these electrodes shall be integrated in objects of daily life to unobtrusively monitor patients (e.g. a chair or a bed), one major problem arises: no medical expert attaches the electrodes and it is a priori not known to which electrodes the patient will have the best contact. Hence, several electrodes are usually integrated and this paper proposes a simple but robust method to automatically select the best coupled electrodes for the deduction of an ECG. It uses the driven right leg (DRL)-electrode to capacitively inject a high frequency voltage into the patient which is then capacitively received and analysed by each electrode. The electrodes with the largest high frequency signal are then used for the deduction of the ECG.

A ballistocardiograph records the mechanical activity of the heart. We present a novel algorithm for the detection of individual heart beats and beat-to-beat interval lengths in ballistocardiograms (BCGs) from healthy subjects. An automatic training step based on unsupervised learning techniques is used to extract the shape of a single heart beat from the BCG. Using the learned parameters, the occurrence of individual heart beats in the signal is detected. A final refinement step improves the accuracy of the estimated beat-to-beat interval lengths. Compared to many existing algorithms, the new approach offers heart rate estimates on a beat-to-beat basis. The agreement of the proposed algorithm with an ECG reference has been evaluated. A relative beat-to-beat interval error of 1.79% with a coverage of 95.94% was achieved on recordings from 16 subjects.

The coordination between locomotion and respiration of Parkinson's disease (PD) patients is reduced or even absent. The degree of this disturbance is assumed to be associated with the disease severity [S. Schiermeier, D. Schäfer, T. Schäfer, W. Greulich, and M. E. Schläfke, "Breathing and locomotion in patients with Parkinson's disease," Eur. J. Physiol., vol. 443, No. 1, pp. 67-71, Jul. 2001]. To enable a long-term and online analysis of the locomotion-respiration coordination for scientific purpose, we have developed a distributed wireless communicating network. We aim to integrate biofeedback protocols with the real-time analysis of the locomotion-respiration coordination in the system to aid rehabilitation of PD patients. The network of sensor nodes is composed of intelligent network operating devices (iNODEs). The miniaturized iNODE contains a continuous data acquisition system based on microcontroller, local data storage, capability of on-sensor digital signal processing in real time, and wireless communication based on IEEE 802.15.4. Force sensing resistors and respiratory inductive plethysmography are applied for motion and respiration sensing, respectively. A number of experiments have been undertaken in clinic and laboratory to test the system. It shall facilitate identification of therapeutic effects on PD, allowing to measure the patients' health status, and to aid in the rehabilitation of PD patients.

During the last years the significance of RFID systems has increased rapidly. In fact, its use is expected to increase by a factor of twenty until 2016. Medical applications are expected to be a main cause of this development. Currently, RFID systems are primary identification systems, based on the transmission of a key number or little further information. Using new technical developements and higher integration levels, it may be possible to extend these systems to measure and transmit data. Such RFID systems can be used in different areas of medical technology, because of the combination of application flexibility with the possibility of identification and data collection. For example, they can be used to identify patients or pharmaceuticals, monitor blood preservations or for medical implant communication.

Early detection of lung edema for patients suffering from chronic heart disease improves the medical treatment and can avoid committal of the patient to an intensive care unit. Therefore, an early warning system monitoring the amount of fluid in the lungs by measuring trans-thoracic bioimpedance outside the body has been developed. The proposed system(TiBIS) consists of a textile integrated measurement module and a Personal Digital Assistant for signal processing and user interaction.

Cardiovascular diseases are among the most common causes of death in industrial countries. In order to take preventive actions, it is of great interest, to both physicians and patients, to determine cardiovascular risk factors early. To address this problem, a wearable in-ear measuring system (IN-MONIT) for 24/7 monitoring of vital parameters has been developed. The central component is a microoptic reflective sensor located inside the auditory canal. From the measured photoplethysmographic curves, heart activity and heart rate can be derived. In this paper, we describe the optoelectronic sensor concept and the autonomous design of the IN-MONIT measurement system. For the assessment of heart rate, different algorithms are introduced and the performance of the developed sensor system is evaluated in relation to conventional systems. In addition, the robustness to external artifacts is evaluated and artifact reduction strategies are considered.

Health and mobility of elderly people is gaining importance in aging societies. New communication-based methods to provide health services with personal health care devices are considered promising elements of first-class medical care services for everybody. To achieve this vision, several technological issues have to be solved: (i) body sensors to monitor vital functions have to be developed; (ii) these sensors should be integrated into textile structures to guarantee ease of use and patient acceptance; (iii)the collected sensor data has to be analyzed to detect emergency situations and to reduce the data volume; (iv) relevant data has to be integrated with other information systems in the work environment of medical experts. These challenges are addressed within the HealthNet project at RWTH Aachen University. The goal of the project is to develop a framework in which health professional scan remotely monitor and diagnose mobile patients. The described demonstration presents our results of the first three issues mentioned above while focusing on the employed data mining and management techniques.