Vinod K. Pandey

A wavelet based technique for suppression of EMG noise and motion artifact in ambulatory ECG

A wavelet-based denoising technique is investigated for suppressing EMG noise and motion artifact in ambulatory ECG. EMG noise is reduced by thresholding the wavelet coefficients using an improved thresholding function combining the features of hard and soft thresholding. Motion artifact is reduced by limiting the wavelet coefficients. Thresholds for both the denoising steps are estimated using the statistics of the noisy signal. Denoising of simulated noisy ECG signals resulted in an average SNR improvement of 11.4 dB, and its application on ambulatory ECG recordings resulted in L2 norm and max-min based improvement indices close to one. It significantly improved R-peak detection in both the cases.

Wavelet Based Cancellation of Respiratory Artifacts in Impedance Cardiography

Impedance cardiography is a noninvasive technique for monitoring stroke volume, based on sensing variation in the thoracic impedance, z(t), due to blood flow. In this technique, first derivative of impedance, dzjdt, is used to calculate two parameters: ventricular ejection time and (-dzjdt)max. Respiration and motion artifacts cause base line drift in the sensed impedance waveform, particularly during exercise, and this drift results in errors in the estimation of the two parameters. Ensemble averaging of dz/dt signal suppresses respiratory and motion artifacts but it introduces distortion in the signal. In this study, a wavelet based denoising method is used to cancel corrupting respiratory artifacts from the dz/dt and z(t) signals in order to make a reliable estimate of cardiac output.

Use of Wavelet Transformation for Geophysical Well-Log Data Analysis

Geophysical well-log (bore-hole) data represent the rock physical properties as a function of depth measured in a well. They aid in demarcating the subsurface horizons, identifying abrupt changes in physical properties of rocks and locating cyclicity (if any) in stratal succession. Rocks document their depositional history, which geoscientists unfold by analyzing different geophysical data, such as well-log data, seismic data, etc. We have applied continuous wavelet transformation technique on one type of well-log data (gamma ray log) from three wells located in Bombay High oil field, India. Our analyses have helped to identify the discontinuities, detect the horizons and check the presence of cyclic pattern (if any) in deposition. Among various wavelets that we have used, the Morlet wavelet has proved to be most appropriate to resolve the marker horizons in gamma ray log data. We also conclude that the choice of wavelet is dependent upon the rock intrinsic properties that vary greatly in the space domain (with varying depth).

Adaptive filtering for suppression of respiratory artifact in impedance cardiography

Impedance cardiography is a noninvasive technique, based on sensing the variation in the electrical impedance of the thorax caused by variation in the blood volume during the cardiac cycle, for monitoring the stroke volume and some other cardiovascular indices. Respiratory and motion artifacts cause baseline drift in the sensed impedance waveform, particularly during or after exercise. This paper presents an LMS-based adaptive filtering technique for suppression of respiratory artifact for improving the estimation of the indices without smearing the beat-to-beat variations. It uses a reference signal, closely related to the respiratory artifact, obtained by a least-squares approximation based B-spline fitting on the contaminated impedance cardiogram in synchronism with the respiratory phases. The technique is evaluated on signals with simulated artifacts and on signals from nine healthy subjects and five patients with cardiovascular disorders. The values of the stroke volume, estimated on beat-to-beat basis, after suppression of respiratory artifact showed a good agreement with those from Doppler echocardiography.

Knotless spline noise removal technique for improved OHCR

Noise in on-line hand written characters due to natural shaking of the hand and noise due to the process of digitization is inherent and this can lead to a degraded performance of character recognition system. In this paper, we propose a noise removal technique based on knotless spline. We first show that its noise removal property is independent of the amount of noise unlike the traditional Gaussian smoothing and further show that the noise removal can significantly enhance the performance of the character recognition algorithm. We specifically perform experiments with Devanagari data and show that the noise removal can enhance the performance of character recognition by as much as 10%.

Adaptive filtering for baseline wander removal in ECG

It is imperative to remove baseline wander from Electrocardiogram (ECG) before utilizing it for diagnostic purpose. In this paper, an adaptive filtering based technique is investigated for removal of baseline wander from the ECG signals obtained from a Holter ECG monitor. It uses a reference signal, obtained by adding absolute power spectrum of 3-channel accelerometer output connected to an ECG electrode. The technique is quantitatively evaluated on the ECG signals by estimating an improvement index based on second order signal statistics. Removal of the actual baseline wander is qualitatively evaluated by applying the technique on the signals acquired from several subjects during post-exercise condition.

Evaluating the Performance of a Speech Recognition Based System

Speech based solutions have taken center stage with growth in the services industry where there is a need to cater to a very large number of people from all strata of the society. While natural language speech interfaces are the talk in the research community, yet in practice, menu based speech solutions thrive. Typically in a menu based speech solution the user is required to respond by speaking from a closed set of words when prompted by the system. A sequence of human speech response to the IVR prompts results in the completion of a transaction. A transaction is deemed successful if the speech solution can correctly recognize all the spoken utterances of the user whenever prompted by the system. The usual mechanism to evaluate the performance of a speech solution is to do an extensive test of the system by putting it to actual people use and then evaluating the performance by analyzing the logs for successful transactions. This kind of evaluation could lead to dissatisfied test users especially if the performance of the system were to result in a poor transaction completion rate. To negate this the Wizard of Oz approach is adopted during evaluation of a speech system. Overall this kind of evaluations is an expensive proposition both in terms of time and cost. In this paper, we propose a method to evaluate the performance of a speech solution without actually putting it to people use. We first describe the methodology and then show experimentally that this can be used to identify the performance bottlenecks of the speech solution even before the system is actually used thus saving evaluation time and expenses.

An Impedance Detector for Glottography

The first article, “An impedance Detector for Glottography,” presents a circuit for the detection of impedance at the sides of the human larynx. It provides a noninvasive technique to monitor the vibrations of vocal folds during voice production, for the diagnosis of laryngeal diseases. A current source-based impedance sensor is described along with the AM detection module. The results have been analyzed and compared with a standard pitch detector instrument.AbstractImpedance glottography, or electroglottography, is a noninvasive technique for monitoring the variation of the degree of contact between the vibrating vocal cords during voice production. Electrical impedance is sensed using a pair of electrodes placed on either side of the larynx, by injecting a low-level high-frequency current. A wide-band impedance detector circuit has been developed using a precision rectifier based on voltage feedback clamping amplifier and approximately linear-phase lowpass filter. The circuit can be used for obtaining electroglottogram signal for diagnosis of vocal fold disorders.