Learn More
We report the implementation of a text input application (speller) based on the P300 event related potential. We obtain high accuracies by using an SVM classifier and a novel feature. These techniques enable us to maintain fast performance without sacrificing the accuracy, thus making the speller usable in an online mode. In order to further improve the(More)
There has been an urgent need for an effective and efficient upper limb rehabilitation method for poststroke patients. We present a Micro-Sensor-based Upper Limb rehabilitation System for poststroke patients. The wearable motion capture units are attached to upper limb segments embedded in the fabric of garments. The body segment orientation relative to the(More)
QRS detection is important because R peak is recognized as the most useful fiducial point in ECG segmentation. Despite of lot of research effort, the accuracy and robustness of QRS detection still remain open problems. Here we propose a novel QRS detector, which uses the Discrete Wavelet Transform (DWT) and Cubic Spline Interpolation as preprocessor,(More)
Automatic detection of life threatening abnormal beats in electrocardiogram (ECG) signal is of importance in many healthcare applications. The ECG beat signal variations in both shape and time impose great challenges to automatic detection tasks. To address those challenges and for high accuracy automatic detection, we present here a two stage abnormal(More)
Motion estimation drift has been a challenge in inertial sensor motion capture research. This paper presents a novel biomechanical model-based multi-sensor motion estimation method working on a group of sensor units attached to a limb. In this method, biomechanical model provides constraints and defines relationships among sensors. The motion parameters of(More)
Due to the complexity and time-variation of wireless channel, it is a great challenge to measure and estimate the channel quality accurately in wireless ad hoc network. To address this problem, we propose a link quality estimator which use the packet loss rate (PLR), round trip time (RTT), available bandwidth (ABW) at the sender and the feedback information(More)
Body sensor networks provide a platform for ubiquitous healthcare, driving the diagnosis in hospital static environment to the daily life dynamic context. We realized the importance of sensing of activities, which is not only a dimension of human health but also important context information for diagnosis based on the physiologic data. This paper presents(More)
To develop effective learning algorithms for continuous prediction of cursor movement using EEG signals is a challenging research issue in Brain-Computer Interface (BCI). To train a classifier for continuous prediction, trials in training dataset are first divided into segments. The difficulty is that the actual intention (label) at each time interval(More)
QRS detection is an important step in electrocardiogram signal processing and analysis. Despite a lot of research effort, robustness and high detection accuracy still remain open problems. Here we present a real-time QRS detector, based on wavelet decomposition and spline interpolation, which is working in our portable health monitor system (PHMS). The(More)
To enhance the performance of hidden Markov models for EEG signal classification, we present here a new model referred to as kernel based hidden Markov model (KHMM). Due to the embedded HMM structure, this model is capable of capturing well the temporal change of a time-series signal. Furthermore, KHMM has better discrimination and generalization capability(More)