[Show abstract][Hide abstract] ABSTRACT: In this ongoing study we present the preliminary results of a fully automatic sleep stages classification based on acceleration and photoplethysmography signals recorded at wrist. The device consists in a bracelet integrating sensors, processing unit, communication capabilities, and power management. The bracelet has been worn by two healthy volunteers during a night period at hospital in combination to a complete polysomnograph. Spectral analysis of heartbeat intervals in standard HRV frequency bands, as well as movement activity level have been performed and used to differentiate 3 sleep states: WAKE, REM and NREM. The automatic classification has been compared to the hypnogram provided by a professional clinician using standard polysomnography procedure. Classification rates up to 90% have been achieved for NREM state and between 44% and 72% for REM state. High confusion coefficients for WAKE state is reported and results from hypnographic misalignment with the algorithm output.
[Show abstract][Hide abstract] ABSTRACT: Since the introduction of portable devices in 1957 by Dr. Holter, ambulatory electrocardiogram (ECG) monitors have been intensely used. Even with important device improvements such as weight, volume and autonomy, these systems are still associated to clinical/ambulatory cumbersome procedures and, therefore have a limited generalized use. The question of this research is the following: what is the performance of a photoplethysmography (PPG)-based device located at the wrist in terms of heart rate variability (HRV) monitoring?
PPG and ECG signals were recorded simultaneously on patients in clinical conditions. Heartbeat (RR) intervals were estimated from both devices. For PPG signals, an approach based on the detection of local minima of the time-derivative was used to estimate RR time series. For ECG signals, a state-of-the-art approach based on adaptive threshold was used (gold standard). The resulting time series of RR were compared in terms of error (mean ± standard deviation) and distribution using Wilcoxon signed-rank test (hypothesis test for distributions with different median values, rejected with p>0.05). The normalized differences observed on state-of-the-art time-domain and frequency-domain HRV features were also computed (standard deviation of RR intervals (SDRR), powers of very low frequency (VLF), low frequency (LF), high frequency (HF)).
Preliminary results based on 563 minutes of recordings showed an overall agreement 0.17±18.2 ms and a correlation coefficient of 0.987 between PPG and ECG-based RRs (N=28420). The hypothesis that both RR distributions have different median values was rejected (p-value=0.59). Regarding HRV analysis, the difference between the PPG-based and ECG-based temporal and frequency features was the following: -0.02±0.02 for SDRR, -0.05±0.21 for VLF, -0.07±0.19 for LF, and -0.03±0.05 for HF.
In view of these preliminary but promising results, it appears that the pro-posed wrist sensor opens the door towards a new generation of comfortable and easy-to-use cardiac HRV tool especially well adapted for long-term monitoring.
40th Conference on Computing in Cardiology, Zaragoza, Spain; 09/2013
[Show abstract][Hide abstract] ABSTRACT: In this paper we present a personal, cost-effective, multi-parametric monitoring system based on a textile platform and portable sensing devices for the long term and short term acquisition of data from bipolar patients affected by mood disorders. The system allows the early indication and prevention of bipolar state relapse situations. The estimation of the bipolar mood state from several physiological and physical cues such as biochemical markers and voice analysis is described. These features, for this paper extracted from cyclothymic patients, are compared to a behavioural index generated by psychologists during consultations with the patients and reassure the potential of the proposed system.
5th European Conference of the International Federation for Medical and Biological Engineering, Budapest, Hungary; 09/2011
[Show abstract][Hide abstract] ABSTRACT: In this paper we present a personal, cost-effective, multi-parametric monitoring system based on textile platforms and portable sensing devices for the long term and short term acquisition of data from bipolar patients affected by mood disorders. The system allows the early indication and prevention of bipolar state relapse situations. The bipolar mood state of the patients is etimated from several physiogical and physical cues such as biochemical markers, voice analysis and a behavioural index correlated to patient state.
5th International Symposium on Medical Information & Communication Technology, Montreux, Switzerland; 03/2011
[Show abstract][Hide abstract] ABSTRACT: Pulse wave velocity (PWV) is a surrogate of arterial stiffness and represents a non-invasive marker of cardiovascular risk. The non-invasive measurement of PWV requires tracking the arrival time of pressure pulses recorded in vivo, commonly referred to as pulse arrival time (PAT). In the state of the art, PAT is estimated by identifying a characteristic point of the pressure pulse waveform. This paper demonstrates that for ambulatory scenarios, where signal-to-noise ratios are below 10 dB, the performance in terms of repeatability of PAT measurements through characteristic points identification degrades drastically. Hence, we introduce a novel family of PAT estimators based on the parametric modeling of the anacrotic phase of a pressure pulse. In particular, we propose a parametric PAT estimator (TANH) that depicts high correlation with the Complior(R) characteristic point D1 (CC = 0.99), increases noise robustness and reduces by a five-fold factor the number of heartbeats required to obtain reliable PAT measurements.
[Show abstract][Hide abstract] ABSTRACT: 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 is to obtain experience in the field which will be used in preparing a possible manned mission to Mars around the year 2030. This paper describes the LTMS2 system.
[Show abstract][Hide abstract] ABSTRACT: Brain-computer interface (BCI) research aims at developing communication devices for the motor disabled. Such devices are not driven by muscle activity, but by brain activity recorded during different mental tasks. We present here the comparison of phase synchronization and power spectral density (PSD) features, computed from broadband and narrowband filtered EEG signals and their ability to discriminate three mental tasks.
EEG signals were recorded from five subjects while performing left and right hand movement imagination and word generation. We applied a modified Fast Correlation Based Filter (FCBF)  for the purpose of feature selection.
We found that the features were selected from electrode signals corresponding to neurophysiological evidence, i.e. electrodes lying over the motor cortex. PSD and phase locking value (PLV) features were more discriminative when computed from narrowband (8-12 Hz) and broadband (8-30 Hz) filtered signals respectively.
The generalization performance is as good as the one obtained with SVM-rfe, but this algorithm is faster and selects fewer features. These properties may make FCBF a valuable tool for further improvement of BCIs.
Methods of Information in Medicine 02/2007; 46(2):160-3. · 1.08 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To allow motor-disabled people for communication, Brain–Computer Interfaces (BCIs) are being developed. Such a communication does not depend on the brain's normal output pathways of peripheral nerves and muscles, but is based on analysis of recorded brain activity. In this paper, we compare the performance of power features and phase locking values (PLVs) computed from broadband and narrowband filtered EEG signals for discriminating 3 mental tasks in the framework of a BCI. EEG signals were recorded from 5 subjects while performing the 3 mental tasks left- and right-hand movement imagination and word generation. To reduce the total amount of features, the most discriminative features were selected in a 2-step feature selection procedure by SVM-based recursive feature elimination.Significance tests demonstrated that band power features were more discriminative when they were computed in the narrower frequency band 8–12 Hz. In case of PLV features, the discrimination of mental tasks was significantly better when they were computed from the broader 8–30 Hz frequency band, as compared to the narrower bands 8–12, 13–18 and 19–30 Hz.
[Show abstract][Hide abstract] ABSTRACT: An improved algorithm for the non-invasive estimation of the autonomous nervous profile is presented. It is based on a previously published method using blind source separation on inter-beat intervals and systolic pressure time series. Improvements focus on robust extraction of cardiovascular parameters and management of singular solutions of blind source separation algorithm. Clinical validation has been performed successfully in 104 experiments on 38 subjects. The method has been finally applied to estimate the autonomic nervous profile during vasovagal syncope, and results have been compared to classical heart rate variability methods.
Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual International Conference of the IEEE; 10/2003
[Show abstract][Hide abstract] ABSTRACT: Portable pulse rate detecting device for contact with human body tissue, including a light-emitting source for emitting radiant energy directed at through human body tissue; at least first and second light detectors for detecting intensity of radiant energy after propagation through human body tissue and for providing first and second input signals as a function of such propagation, a detecting device for providing a motion reference signal, and processing means for removing motion-related contributions from the first and second input signals and subtracting a calculated model based on the motion reference signal from each of the first and second input signals, wherein the processing means is also for removing measurement noise and residual non-modeled contributions from the first and second enhanced signals using a noise reduction algorithm.
[Show abstract][Hide abstract] ABSTRACT: We present a new integrated device for monitoring heart rate at the wrist using an optical measure. Motion robustness is obtained by using accurate motion reference signals of 3D low noise accelerometers together with dual channel optical sensing. Nonlinear modelling allows to remove the motion contributions in the optical signals and the spatial diversity of the sensors is used to remove reciprocal contributions in the two channels. Finally a statistical estimation, based on physiological properties of the heart, gives a robust estimation of the heart rate. Qualitative and quantitative performance evaluation of the performances on real signals clearly show that the proposed system gives an accurate estimation of the heart rate, even under intense physical activity.
Engineering in Medicine and Biology Society, 2001. Proceedings of the 23rd Annual International Conference of the IEEE; 02/2001
[Show abstract][Hide abstract] ABSTRACT: This paper addresses the problem of speech recognition in noisy conditions when low complexity is required like in embed- ded systems. In such systems, vector quantization is generally used to reduce the complexity of the recognition systems (e.g. HMMs). A novel approach for vector quantization based on the miss- ing data theory is proposed. This approach allows to increase the robustness of the system against the noise perturbations with only a small increase of the computational requirements. The proposed algorithm is composed of two parts. The first part consists in dividing the spectral temporal features of the noisy signal into two subspaces: the unreliable (or missing) features and the reliable (or present) features. The second part of the proposed approach consists in defining a robust distance mea- sure for vector quantization that compensates for the unreliable features. The proposed approach obtains similar results in noisy con- ditions than a more classical approach that consists in adapting the codebook of the vector quantization to the noisy conditions using model compensation. However the computation require- ments are lower in the proposed approach and it is more suitable for a low complexity speech recognition system.
EUROSPEECH 2001 Scandinavia, 7th European Conference on Speech Communication and Technology, 2nd INTERSPEECH Event, Aalborg, Denmark, September 3-7, 2001; 01/2001
[Show abstract][Hide abstract] ABSTRACT: This paper addresses the problem of robust speech recognition in
noisy conditions in the framework of hidden Markov models (HMMs) and
missing feature techniques. It presents a new statistical approach to
detection and estimation of unreliable features based on a probabilistic
measure and Gaussian mixture model (GMM). In the estimation process, the
GMM is compensated using parameters of the statistical model of additive
background noise. The GMM means are used to replace the unreliable
features. The GMM based technique is less complex than the corresponding
HMM based estimation and gives similar improvement in the recognition
performance. Once unreliable features are replaced by the estimated
clean speech features, the entire set of spectral features can be
transformed to the other feature domain characterized by higher baseline
recognition rate (e.g. MFCCs) for final recognition using continuous
density hidden Markov models (CDHMMs) with diagonal covariance matrices
Acoustics, Speech, and Signal Processing, 2000. ICASSP '00. Proceedings. 2000 IEEE International Conference on; 02/2000