Abstract—Textile-based sensors offer an unobtrusive method of continually monitoring physiologica... more Abstract—Textile-based sensors offer an unobtrusive method of continually monitoring physiological parameters during daily ac-tivities. Chemical analysis of body fluids, noninvasively, is a novel and exciting area of personalized wearable healthcare systems. BIOTEX was an EU-funded project that aimed to develop tex-tile sensors to measure physiological parameters and the chemical composition of body fluids, with a particular interest in sweat. A wearable sensing system has been developed that integrates a textile-based fluid handling system for sample collection and trans-port with a number of sensors including sodium, conductivity, and pH sensors. Sensors for sweat rate, ECG, respiration, and blood oxygenation were also developed. For the first time, it has been possible to monitor a number of physiological parameters together with sweat composition in real time. This has been carried out via a network of wearable sensors distributed around the body of a
A monitoring device worn by a user comprising a heart rate sensor for delivering a sensor cardiov... more A monitoring device worn by a user comprising a heart rate sensor for delivering a sensor cardiovascular (CV) signal, a signal processing device for correcting artifacts due to user's motion in said sensor CV signal, and a motion sensor delivering a motion reference signal the motion sensor having a main function for determining and outputting a signal representing the level of activity of the user, and an auxiliary function for controlling said signal processing device. The monitoring device may be disposed in a headband worn on the user's head. The monitoring device may comprise the monitoring of other physiological signals such as SPO 2.
PORTABLE-POLYTRONIC 2008 - 2nd IEEE International Interdisciplinary Conference on Portable Information Devices and the 2008 7th IEEE Conference on Polymers and Adhesives in Microelectronics and Photonics, 2008
The European Space Agency (ESA) commissioned CSEM to design, build, validate and deliver one full... more The European Space Agency (ESA) commissioned CSEM to design, build, validate and deliver one fully operational ground prototype of a system (LTMS2) measuring physiological parameters. The system was shipped and will be used during 2008 at Concordia station (www.concordiastation.com) in Antarctica to study the physiological adaptation of crews to remote, isolated and extreme environments. The underlying long-term objective for ESA
ABSTRACT We present a clinically validated patient monitoring system composed of a wearable sensi... more ABSTRACT We present a clinically validated patient monitoring system composed of a wearable sensing unit, a portable data recorder and a remote station with software for acquiring, processing, archiving and visualizing recorded physiological data. The sensing unit simultaneously measures ECG, pulse oximetry, blood pressure, activity/posture, temperature, and respiration rate in an unobtrusive, noninvasive and modular way by integrating and synchronizing proprietary and commercial sensors. The system received Swissmedic’s authorization to be clinically validated at the Department of Intensive Care Medicine of Bern University Hospital. It was tested in a controlled environment and the recorded signals were compared to medical reference measurements from the devices of the Department of Intensive Care Medicine. Three healthy subjects were monitored for three sessions of 17 hours and one of 24 hours. All sessions were composed of usual daily activities such as office work, walking, jogging, and night sleeping.
2007 Ph.D Research in Microelectronics and Electronics Conference, 2007
Sampling is commonly retained as a critical step in any mixed-signal system. High-speed analog-to... more Sampling is commonly retained as a critical step in any mixed-signal system. High-speed analog-to-digital converter sampling jitter limits all-over performance of these systems introducing a signal dependent noise in the sampled signal. In most environments it is desirable to reduce sampling clock jitter, however there are cases where designers are forced to introduce or cope with this undesirable noise effect.
The present disclosure relates to a method for estimating blood constituent concentration of a us... more The present disclosure relates to a method for estimating blood constituent concentration of a user under low perfusion conditions using a spectrophotometry-based monitoring device; the method comprising: measuring a plurality of photoplethysmographic (PPG) signals; measuring a cardio-synchronous (CV) signal; detecting an instantaneous heart rate and determining a heart rate variability from the CV signal; selecting reliable projected PPG signals; estimating a value of said blood constituent concentration from the magnitude of ...
Résumé À cause des singularités de la fonction de vraisemblance, l'approche par maximum de v... more Résumé À cause des singularités de la fonction de vraisemblance, l'approche par maximum de vraisemblance pour l'estimation des paramètres d'un mélange gaussien est connu pour être un problème d'optimisation mal posé. Nous proposons dans cette communication, une pénalisation de la fonction de vraisemblance par une distribution a priori de type gamma inverse qui élimine les singularités et rend ainsi ce problème bien posé. Une conséquence algorithmique intéressante d'un tel choix est de fournir une version pénalisée de l'algorithme EM qui conserve une structure de remise à jour explicite et qui garantit que les estimées ne sont pas singulières. Un exemple numérique met en evidence cette dernière propriété. Abstract Due to singularities of the likelihood function, the maximum likelihood approach for the estimation of the parameters of normal mixture models is an acknowledged ill posed optimization problem. Ill posedness is solved by penalizing the likeliho...
This communication presents a non-supervised three-dimensional segmentation method based upon a d... more This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameteriza...
This communication presents a non-supervised three-dimensional segmentation method based upon a d... more This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameteriza...
In [1] a sampling theorem for a certain class of signals with finite rate of innovation (which in... more In [1] a sampling theorem for a certain class of signals with finite rate of innovation (which includes for example stream of Diracs) has been developed. In essence, such non bandlimited signals can be sampled at or above the rate of innovation. In the present paper, we consider the case of such signals when noise is present. Clearly, the finite rate of innovation property is lost, but if the signal-to-noise ratio (SNR) is sufficient, several methods are possible to reconstruct the signal while sampling well below the Nyquist rate. We thus explore the trade-offs between SNR, sampling rate, computational complexity and reconstruction quality. Applications of such methods can be found in acquisition and processing of signals in high bandwidth communications, like ultra wide band communication [2].
This paper deals with the statistical analysis of subspace identification methods, and more preci... more This paper deals with the statistical analysis of subspace identification methods, and more precisely, with the analysis of model uncertainty evaluation. The use of the statistical tool of bootstrap in evaluating model uncertainty is validated on a theoretical basis, by proving the consistency of the “bootstrapped” sample variance of the frequency response of the estimated model.
AbstractIn [1] a sampling theorem for a certain class of signals with finite rate of innovation (w... more AbstractIn [1] a sampling theorem for a certain class of signals with finite rate of innovation (which includesfor example stream of Diracs) has been developed. In essence, such non band-limited signals can besampled at or above the rate of innovation. In the present paper, we consider the case of such signalswhen noise is present. Clearly, the finite rate of innovation property is lost, but if the signal-to-noise ratio(SNR) is sufficient, several methods are possible to reconstruct the signal while sampling well below theNyquist rate. We thus explore the trade-offs between SNR, sampling rate, computational complexity andreconstruction quality. Applications of such methods can be found in acquisition and processing of signalsin high bandwidth communications, like ultra wide band communication [2]. I. I NTRODUCTION Band-limited functions are just an example of signals that are specified by a fixed number ofsamples per unit of time, namely, if a signal x(t) is band-limited to [ ! m ;! m ],...
Mixture models form the essential basis of data clustering within a statistical framework. Here, ... more Mixture models form the essential basis of data clustering within a statistical framework. Here, the estimation of the parameters of a mixture of Gaussian densities is considered. In this particular context, it is well known that the maximum likelihood approach is statistically ill posed, i.e. the likelihood function is not bounded above, because of singularities at the boundary of the parameter domain. We show that such a degeneracy can be avoided by penalizing the likelihood function using a suited type of penalty function. Recently, the resulting penalized maximum likelihood estimator has been proved to be asymptotically well-behaved. Local maximization of the likelihood function can be performed by mean of Green's modified EM algorithm: provided that an inverse gamma is chosen as penalty function, EM re-estimation equations are still explicit and automatically ensure that the estimates are not singular. Numerical examples are provided in the finite data case, showing the per...
ABSTRACT This communication presents a non-supervised three-dimensional segmentation method based... more ABSTRACT This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameterization ensures that some important properties of the field (e.g., stationarity) do hold. On a practical level, it reduces the computational complexity of the algorithm used in the segmentation and parameter estimation stages of the procedure. In addition, it decreases the number of model parameters that must be estimated, thereby improving convergence speed and accuracy of the corresponding estimation method.
Estimating SpO2 from photoplethysmographic measurements at the sternum is a challenging task give... more Estimating SpO2 from photoplethysmographic measurements at the sternum is a challenging task given sternum’s very low perfusion. Reliable SpO2 estimation requires to specifically tackle the problem of very low SNR by simultaneously increasing the available information and reducing the noise. We propose a novel algorithm that: 1) Intelligently takes advantage of multichannel photoplethysmography to increase the available information; 2) Uses
2006 IEEE International Conference on Acoustics Speed and Signal Processing Proceedings, 2006
In this paper, we model a spatially varying channel where a source is moving along a random traje... more In this paper, we model a spatially varying channel where a source is moving along a random trajectory with respect to a fixed receiver. The aim is to compute the power spectral density corresponding to the channel impulse response as a function of temporal and spatial frequencies. The trajectory of the source follows an auto regressive model where the poles
Abstract—Textile-based sensors offer an unobtrusive method of continually monitoring physiologica... more Abstract—Textile-based sensors offer an unobtrusive method of continually monitoring physiological parameters during daily ac-tivities. Chemical analysis of body fluids, noninvasively, is a novel and exciting area of personalized wearable healthcare systems. BIOTEX was an EU-funded project that aimed to develop tex-tile sensors to measure physiological parameters and the chemical composition of body fluids, with a particular interest in sweat. A wearable sensing system has been developed that integrates a textile-based fluid handling system for sample collection and trans-port with a number of sensors including sodium, conductivity, and pH sensors. Sensors for sweat rate, ECG, respiration, and blood oxygenation were also developed. For the first time, it has been possible to monitor a number of physiological parameters together with sweat composition in real time. This has been carried out via a network of wearable sensors distributed around the body of a
A monitoring device worn by a user comprising a heart rate sensor for delivering a sensor cardiov... more A monitoring device worn by a user comprising a heart rate sensor for delivering a sensor cardiovascular (CV) signal, a signal processing device for correcting artifacts due to user's motion in said sensor CV signal, and a motion sensor delivering a motion reference signal the motion sensor having a main function for determining and outputting a signal representing the level of activity of the user, and an auxiliary function for controlling said signal processing device. The monitoring device may be disposed in a headband worn on the user's head. The monitoring device may comprise the monitoring of other physiological signals such as SPO 2.
PORTABLE-POLYTRONIC 2008 - 2nd IEEE International Interdisciplinary Conference on Portable Information Devices and the 2008 7th IEEE Conference on Polymers and Adhesives in Microelectronics and Photonics, 2008
The European Space Agency (ESA) commissioned CSEM to design, build, validate and deliver one full... more The European Space Agency (ESA) commissioned CSEM to design, build, validate and deliver one fully operational ground prototype of a system (LTMS2) measuring physiological parameters. The system was shipped and will be used during 2008 at Concordia station (www.concordiastation.com) in Antarctica to study the physiological adaptation of crews to remote, isolated and extreme environments. The underlying long-term objective for ESA
ABSTRACT We present a clinically validated patient monitoring system composed of a wearable sensi... more ABSTRACT We present a clinically validated patient monitoring system composed of a wearable sensing unit, a portable data recorder and a remote station with software for acquiring, processing, archiving and visualizing recorded physiological data. The sensing unit simultaneously measures ECG, pulse oximetry, blood pressure, activity/posture, temperature, and respiration rate in an unobtrusive, noninvasive and modular way by integrating and synchronizing proprietary and commercial sensors. The system received Swissmedic’s authorization to be clinically validated at the Department of Intensive Care Medicine of Bern University Hospital. It was tested in a controlled environment and the recorded signals were compared to medical reference measurements from the devices of the Department of Intensive Care Medicine. Three healthy subjects were monitored for three sessions of 17 hours and one of 24 hours. All sessions were composed of usual daily activities such as office work, walking, jogging, and night sleeping.
2007 Ph.D Research in Microelectronics and Electronics Conference, 2007
Sampling is commonly retained as a critical step in any mixed-signal system. High-speed analog-to... more Sampling is commonly retained as a critical step in any mixed-signal system. High-speed analog-to-digital converter sampling jitter limits all-over performance of these systems introducing a signal dependent noise in the sampled signal. In most environments it is desirable to reduce sampling clock jitter, however there are cases where designers are forced to introduce or cope with this undesirable noise effect.
The present disclosure relates to a method for estimating blood constituent concentration of a us... more The present disclosure relates to a method for estimating blood constituent concentration of a user under low perfusion conditions using a spectrophotometry-based monitoring device; the method comprising: measuring a plurality of photoplethysmographic (PPG) signals; measuring a cardio-synchronous (CV) signal; detecting an instantaneous heart rate and determining a heart rate variability from the CV signal; selecting reliable projected PPG signals; estimating a value of said blood constituent concentration from the magnitude of ...
Résumé À cause des singularités de la fonction de vraisemblance, l'approche par maximum de v... more Résumé À cause des singularités de la fonction de vraisemblance, l'approche par maximum de vraisemblance pour l'estimation des paramètres d'un mélange gaussien est connu pour être un problème d'optimisation mal posé. Nous proposons dans cette communication, une pénalisation de la fonction de vraisemblance par une distribution a priori de type gamma inverse qui élimine les singularités et rend ainsi ce problème bien posé. Une conséquence algorithmique intéressante d'un tel choix est de fournir une version pénalisée de l'algorithme EM qui conserve une structure de remise à jour explicite et qui garantit que les estimées ne sont pas singulières. Un exemple numérique met en evidence cette dernière propriété. Abstract Due to singularities of the likelihood function, the maximum likelihood approach for the estimation of the parameters of normal mixture models is an acknowledged ill posed optimization problem. Ill posedness is solved by penalizing the likeliho...
This communication presents a non-supervised three-dimensional segmentation method based upon a d... more This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameteriza...
This communication presents a non-supervised three-dimensional segmentation method based upon a d... more This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameteriza...
In [1] a sampling theorem for a certain class of signals with finite rate of innovation (which in... more In [1] a sampling theorem for a certain class of signals with finite rate of innovation (which includes for example stream of Diracs) has been developed. In essence, such non bandlimited signals can be sampled at or above the rate of innovation. In the present paper, we consider the case of such signals when noise is present. Clearly, the finite rate of innovation property is lost, but if the signal-to-noise ratio (SNR) is sufficient, several methods are possible to reconstruct the signal while sampling well below the Nyquist rate. We thus explore the trade-offs between SNR, sampling rate, computational complexity and reconstruction quality. Applications of such methods can be found in acquisition and processing of signals in high bandwidth communications, like ultra wide band communication [2].
This paper deals with the statistical analysis of subspace identification methods, and more preci... more This paper deals with the statistical analysis of subspace identification methods, and more precisely, with the analysis of model uncertainty evaluation. The use of the statistical tool of bootstrap in evaluating model uncertainty is validated on a theoretical basis, by proving the consistency of the “bootstrapped” sample variance of the frequency response of the estimated model.
AbstractIn [1] a sampling theorem for a certain class of signals with finite rate of innovation (w... more AbstractIn [1] a sampling theorem for a certain class of signals with finite rate of innovation (which includesfor example stream of Diracs) has been developed. In essence, such non band-limited signals can besampled at or above the rate of innovation. In the present paper, we consider the case of such signalswhen noise is present. Clearly, the finite rate of innovation property is lost, but if the signal-to-noise ratio(SNR) is sufficient, several methods are possible to reconstruct the signal while sampling well below theNyquist rate. We thus explore the trade-offs between SNR, sampling rate, computational complexity andreconstruction quality. Applications of such methods can be found in acquisition and processing of signalsin high bandwidth communications, like ultra wide band communication [2]. I. I NTRODUCTION Band-limited functions are just an example of signals that are specified by a fixed number ofsamples per unit of time, namely, if a signal x(t) is band-limited to [ ! m ;! m ],...
Mixture models form the essential basis of data clustering within a statistical framework. Here, ... more Mixture models form the essential basis of data clustering within a statistical framework. Here, the estimation of the parameters of a mixture of Gaussian densities is considered. In this particular context, it is well known that the maximum likelihood approach is statistically ill posed, i.e. the likelihood function is not bounded above, because of singularities at the boundary of the parameter domain. We show that such a degeneracy can be avoided by penalizing the likelihood function using a suited type of penalty function. Recently, the resulting penalized maximum likelihood estimator has been proved to be asymptotically well-behaved. Local maximization of the likelihood function can be performed by mean of Green's modified EM algorithm: provided that an inverse gamma is chosen as penalty function, EM re-estimation equations are still explicit and automatically ensure that the estimates are not singular. Numerical examples are provided in the finite data case, showing the per...
ABSTRACT This communication presents a non-supervised three-dimensional segmentation method based... more ABSTRACT This communication presents a non-supervised three-dimensional segmentation method based upon a discrete-level unilateral Markov field model for the labels and conditionaly Gaussian densities for the observed voxels. Such models have been shown to yield numerically efficient algorithms, for segmentation and for estimation of the model parameters as well. Our contribution is twofold. First, we deal with the degeneracy of the likelihood function with respect to the parameters of the Gaussian densities, which is a well-known problem for such mixture models. We introduce a bounded penalized likelihood function that has been recently shown to provide a consistent estimator in the simpler cases of independent Gaussian mixtures. On the other hand, implementation with EM reestimation formulas remains possible with only limited changes with respect to the standard case. Second, we propose a telegraphic parameterization of the unilateral Markov field. On a theoretical level, this parameterization ensures that some important properties of the field (e.g., stationarity) do hold. On a practical level, it reduces the computational complexity of the algorithm used in the segmentation and parameter estimation stages of the procedure. In addition, it decreases the number of model parameters that must be estimated, thereby improving convergence speed and accuracy of the corresponding estimation method.
Estimating SpO2 from photoplethysmographic measurements at the sternum is a challenging task give... more Estimating SpO2 from photoplethysmographic measurements at the sternum is a challenging task given sternum’s very low perfusion. Reliable SpO2 estimation requires to specifically tackle the problem of very low SNR by simultaneously increasing the available information and reducing the noise. We propose a novel algorithm that: 1) Intelligently takes advantage of multichannel photoplethysmography to increase the available information; 2) Uses
2006 IEEE International Conference on Acoustics Speed and Signal Processing Proceedings, 2006
In this paper, we model a spatially varying channel where a source is moving along a random traje... more In this paper, we model a spatially varying channel where a source is moving along a random trajectory with respect to a fixed receiver. The aim is to compute the power spectral density corresponding to the channel impulse response as a function of temporal and spatial frequencies. The trajectory of the source follows an auto regressive model where the poles
Uploads
Papers