Weichih Hu

Using wavelet transform and fuzzy neural network for VPC detection from the holter ECG

A novel method for detecting ventricular premature contraction (VPC) from the Holter system is proposed using wavelet transform (WT) and fuzzy neural network (FNN). The basic ideal and major advantage of this method is to reuse information that is used during QRS detection, a necessary step for most ECG classification algorithm, for VPC detection. To reduce the influence of different artifacts, the filter bank property of quadratic spline WT is explored. The QRS duration in scale three and the area under the QRS complex in scale four are selected as the characteristic features. It is found that the R wave amplitude has a marked influence on the computation of proposed characteristic features. Thus, it is necessary to normalize these features. This normalization process can reduce the effect of alternating R wave amplitude and achieve reliable VPC detection. After normalization and excluding the left bundle branch block beats, the accuracies for VPC classification using FNN is 99.79%. Features that are extracted using quadratic spline wavelet were used successfully by previous investigators for QRS detection. In this study, using the same wavelet, it is demonstrated that the proposed feature extraction method from different WT scales can effectively eliminate the influence of high and low-frequency noise and achieve reliable VPC classification. The two primary advantages of using same wavelet for QRS detection and VPC classification are less computation and less complexity during actual implementation.

A portable device for real time drowsiness detection using novel active dry electrode system

Electroencephalogram (EEG) signals give important information about the vigilance states of a subject. Therefore, this study constructs a real-time EEG-based system for detecting a drowsy driver. The proposed system uses a novel six channels active dry electrode system to acquire EEG non-invasively. In addition, it uses a TMS320VC5510 DSP chip as the algorithm processor, and a MSP430F149 chip as a controller to achieve a real-time portable system. This study implements stationary wavelet transform to extract two features of EEG signal: integral of EEG and zero crossings as the input to a back propagation neural network for vigilance states classification. This system can discriminate alertness and drowsiness in real-time. The accuracy of the system is 79.1% for alertness and 90.91% for drowsiness states. When the system detects drowsiness, it will warn drivers by using a vibrator and a beeper.

Evaluation of Scalp and Auricular Acupuncture on EEG, HRV, and PRV

In this study, the EEG, ECG and blood-pressure-pulse recorder were employed to evaluate heart rate variability, pulse rate variability, and EEG of 10 adults after scalp (experimental test I) at Sishencong scalp acupoint and auricular (experimental test II) acupuncture at Shenmen auricular acupoint for about 10 min. Comparison of the results between the experimental tests and a control with no stimulation test showed that both the heart rate and pulse rate were decreased, and the blood pressure fell. The high and low frequency power of FFT analysis of heart rate was increased and decreased, respectively; indicating that the parasympathetic nerves were activated and the sympathetic nerves were inhibited. The analysis of the power spectrum of EEG showed that the number of low frequency waves was increased after acupuncture stimulation. Therefore, acupuncture on either Sishencong or Shenmen might calm the mind, slow down the heart rate, and activate the parasympathetic nerves.

ECG events detection and classification using wavelet and neural networks

An integrated system for ECG diagnosis that combines the wavelet transform (WT) for feature extraction and artificial neural network (ANN) models for the classification is proposed. By using the dyadic wavelet transform, the limitations of other methods in detecting ECG features such as QRS complex, the onsets and offsets of P and T waves are overcame. The ECG baseline is approximated using discrete least squares approximation. On classification, two paradigms of learning, supervised and unsupervised, for training the ANN modes are investigated. The backpropagation algorithm and the Kohonens self-organizing feature map algorithm were used for supervised and unsupervised learning, respectively. The system is evaluated using the MITBIH database. The result indicates that the accuracy of diagnosing cardiac disease is above 97.77%. ECG signals can be classified, even with noise and baseline drift.

Portable impedance cardiography system for real-time noninvasive cardiac output measurement

A portable impedance cardiography (ICG) system has been developed. It is battery powered, small size, and in-expensive. Only one 9.6 V Ni/Cd battery is needed to operate the system for one and half hours, continuously. In addition, a digital signal processing (DSP) chip (TMS320C50, National Instrumentation Co.) is used for real-time signal processing, event detection, and stroke volume (SV) calculation. Information such as SV, cardiac output (CO), heart rate (HR) and left ventricular ejection time (LVET) are displayed on a 128*64 dot matrix LCD and updated after each heart beat. This system is also capable of storing the above information in the internal SRAM that can be transmitted to the personal computer through an RS232 interface for further analysis, after monitoring.

The detection of impedance cardiogram characteristic points using wavelet transform

A novel impedance cardiograph event detection method using wavelet transform is proposed. When compared to the C and E points in the pressure-volume loop, the wavelet method performs significantly better than the traditional method (P < 0.05) in the B and X points detection even after the addition of 20% artificial noise into the test signal. Nevertheless, the SVs estimated by ICG are poorly correlated with values measured by the conductance catheter.

Two-Cuff Noninvasive Blood Pressure Waveform Monitoring System for Dynamic Blood Vessel Characteristic Study

This study proposes and constructs a two-cuff non-invasive blood pressure waveform monitor system. The proposed system uses dynamic feedback to maintain constant low cuff pressure at 40 mmHg. Thus, long term continuous blood pressure waveform monitoring is possible. Additionally, the two blood pressure waveforms can be used to compute the system function, frequency response and impulse response of vessel system for system dynamic studies. Ten normal young males were recruited for this study. Continuous blood pressure signals from upper arm and wrist were obtained by the proposed system. During steady state, correlation coefficients of systolic pressure and diastolic pressure, between upper arm and wrist, are 0.83plusmn0.15 and 0.81plusmn0.14, respectively. During cold pressor, correlations decrease slightly. Additionally, the blood vessel system model was evaluated using autoregressive exogenous model (ARX) model. The dynamic changes of blood vessel system during task can be seen from the 3D time-varying impulse responses. In order to assist the observation of the dynamic characteristic change of blood vessel system during different physiological conditions, the 90% energy curve was proposed and used. Before the end of cold pressor, the 90% energy curve reviewed that system energy accumulates faster than steady state indicating the vessel system shifted to higher frequency. On the other hand, the system shifted to slower system during Valsalva maneuver

The use of thoracic-abdominal transfer function in extracting fetal electrocardiogram

The technique that uses the thoracic abdominal transfer function for fetal electrocardiogram (FECG) extraction is proposed and tested. Signals obtained by means of surface electrodes during non-invasive FECG monitoring are contaminated by maternal electrocardiogram (MECG). The proposed method calculates the transfer function between thoracic and abdominal MECG. Upon acquiring this transfer function, the abdominal MECG can be estimated using simultaneously recorded thoracic MECG for cancellation. Both simulated and real signals were used to evaluate the performance of this method. It is concluded that this method provides good to excellent results depend on the quality of incoming signals.

A wavelet-based high-frequency analysis of fragmented QRS complexes in patients with myocardial infarction

Fragmented QRS (fQRS) is an important and noninvasive marker for evaluating myocardial scar in patients with coronary artery disease, which is defined as additional spikes within the QRS wave. It is not easy to detect the fQRS accurately because of a variety of fQRS morphologies. This study is to analyze the high-frequency (HF) potentials of fQRS complexes using a continuous wavelet transform-based method in patients with myocardial infarction (MI). The HF parameter is defined as the root-mean-square (RMS) value of wavelet coefficients at the central frequencies of 100Hz, 150Hz, 200Hz, or 250Hz further normalized by the RMS value of the entire QRS complex, which is defined as the HF ratio. There were 76 MI patients and 43 Normal subjects recruited in this study. All of the ECG recordings were obtained from the PTB Diagnostic ECG Database including the conventional 12-lead and Frank XYZ lead ECGs. A signal averaging technology was adopted to reduce the background noise. The fQRS complexes were defined by the presence of an additional R wave, or notching in the nadir of the S wave, notching of the R wave, or the presence of more than one R prime. All of the mean HF ratios of the fQRS complexes are significantly larger than those of the non-fQRS complexes (p<;0.001). The total accuracy of the HF ratio for detecting the fQRS complex is about 80% (specificity 84% and sensitivity 60%) in the 12-lead ECGs, and about 84% (specificity 88% and sensitivity 65%) in the Frank lead ECGs.

Development of a cuffless blood pressure measurement system

This study constructs a novel blood pressure measurement device without the air cuff to overcome the problem of discomfort and portability. The proposed device measures the blood pressure through a mechanism that is made of silicon rubber and pressure transducer. The system uses a microcontroller to control the measurement procedure and to perform the necessary computation. To verify the feasibility of the constructed device, ten young volunteers were recruited. Ten blood pressure readings were obtained using the new system and were compared with ten blood pressure readings from bedside monitor (Spacelabs Medical, model 90367). The results indicated that, when all the readings were included, the mean pressure, systolic pressure and diastolic pressure from the new system were all higher than those from bedside monitor. The correlation coefficients between these two were 0.15, 0.18 and 0.29, for mean, systolic and diastolic pressures, respectively. After excluding irregular apparatus utilization, the correlation coefficient increased to 0.71, 0.60 and 0.41 for diastolic pressure, mean pressure and systolic pressure, respectively. We can conclude from these results that the accuracy can be improved effectively by defining the user regulation more precisely. The above mentioned irregular apparatus utilization factors can be identified and eliminated by the microprocessor to provide a reliable blood pressure measurement in practical applications in the future.