Jeevan K. Pant

Compressive Sensing of Electrocardiogram Signals by Promoting Sparsity on the Second-Order Difference and by Using Dictionary Learning

A new algorithm for the reconstruction of electrocardiogram (ECG) signals and a dictionary learning algorithm for the enhancement of its reconstruction performance for a class of signals are proposed. The signal reconstruction algorithm is based on minimizing the lp pseudo-norm of the second-order difference, called as the lp2d pseudo-norm, of the signal. The optimization involved is carried out using a sequential conjugate-gradient algorithm. The dictionary learning algorithm uses an iterative procedure wherein a signal reconstruction and a dictionary update steps are repeated until a convergence criterion is satisfied. The signal reconstruction step is implemented by using the proposed signal reconstruction algorithm and the dictionary update step is implemented by using the linear least-squares method. Extensive simulation results demonstrate that the proposed algorithm yields improved reconstruction performance for temporally correlated ECG signals relative to the state-of-the-art lp1d-regularized least-squares and Bayesian learning based algorithms. Also for a known class of signals, the reconstruction performance of the proposed algorithm can be improved by applying it in conjunction with a dictionary obtained using the proposed dictionary learning algorithm.

Reconstruction of ECG signals for compressive sensing by promoting sparsity on the gradient

A new algorithm for the reconstruction of signals in compressive sensing framework is proposed. The algorithm is based on a least-squares method which incorporates a regularization to promote sparsity on the gradient of the signal. It uses a sequential basic conjugate-gradient method, and it is especially suited for the reconstruction of signals which exhibit temporal correlation, e.g., electrocardiogram (ECG) signals. Simulation results are presented which demonstrate that the proposed algorithm yields upto 80.28% reduction in mean square error and from 49.95% to 65.64% reduction in the required amount of computation, relative to the state-of-the-art block sparse Bayesian learning bound-optimization algorithm.

Compressive sensing of ECG signals based on mixed pseudonorm of the first- and second-order differences

An improved algorithm for the reconstruction of electrocardiogram signals in compressive sensing is proposed. The algorithm is based on the minimization of a mixed pseudonorm of first- and second-order differences of the signal. Locations of QRS segments are estimated using a technique based on signal derivatives and the Hilbert transform, and they are used to implement the mixed pseudonorm. Simulation results demonstrate that the proposed algorithm offers approximately 23.5%, 11.4%, 4.4%, and 2.1% improvement in signal-to-noise ratio for a compression ratio of 90%, 80%, 70%, and 60%, respectively, relative to several competitive state-of-the-art algorithms.

Foot gait time series estimation based on support vector machine.

A new algorithm for the estimation of stride interval time series from foot gait signals is proposed. The algorithm is based on the detection of beginning of heel strikes in the signal by using the support vector machine. Morphological operations are used to enhance the accuracy of detection. By taking backward differences of the detected beginning of heel strikes, stride interval time series is estimated. Simulation results are presented which shows that the proposed algorithm yields fairly accurate estimation of stride interval time series where estimation error for mean and standard deviation of the time series is of the order of 10-4.

Robust QRS detection for HRV estimation from compressively sensed ECG measurements for remote health-monitoring systems

OBJECTIVE To present a new compressive sensing (CS)-based method for the acquisition of ECG signals and for robust estimation of heart-rate variability (HRV) parameters from compressively sensed measurements with high compression ratio. APPROACH CS is used in the biosensor to compress the ECG signal. Estimation of the locations of QRS segments is carried out by applying two algorithms on the compressed measurements. The first algorithm reconstructs the ECG signal by enforcing a block-sparse structure on the first-order difference of the signal, so the transient QRS segments are significantly emphasized on the first-order difference of the signal. Multiple block-divisions of the signals are carried out with various block lengths, and multiple reconstructed signals are combined to enhance the robustness of the localization of the QRS segments. The second algorithm removes errors in the locations of QRS segments by applying low-pass filtering and morphological operations. MAIN RESULTS The proposed CS-based method is found to be effective for the reconstruction of ECG signals by enforcing transient QRS structures on the first-order difference of the signal. It is demonstrated to be robust not only to high compression ratio but also to various artefacts present in ECG signals acquired by using on-body wireless sensors. HRV parameters computed by using the QRS locations estimated from the signals reconstructed with a compression ratio as high as 90% are comparable with that computed by using QRS locations estimated by using the Pan-Tompkins algorithm. SIGNIFICANCE The proposed method is useful for the realization of long-term HRV monitoring systems by using CS-based low-power wireless on-body biosensors.

Two-pass ℓp-regularized least-squares algorithm for compressive sensing

A two-pass algorithm for signal reconstruction in compressive sensing (CS) is proposed. It is based on using a new regularization in the objective function which elevates functional value at a previously obtained optimal or near-optimal point. Elevation in the objective function causes the optimization to converge to a new solution which would be optimal or near-optimal. Either previously obtained solution or the newly obtained solution is selected as the final solution based on which one yields lower value of the objective function. This algorithm is suitable for nonconvex optimization-based sparse signal reconstruction in CS. Simulation results are presented which indicate that the proposed algorithm is effective for not only improving the percentage of perfect reconstructions from noiseless measurements by upto 3.4% but also offering similar performance improvement for the reconstruction from noisy measurements.

Efficient compressive sensing of ECG segments based on machine learning for QRS-based arrhythmia detection

A novel method for efficient telemonitoring of arrhythmia based on using QRS complexes is proposed. Two features, namely, sum of absolute differences (SAD) and maximum of absolute differences (MAD) are efficiently computed for each ECG segment in the bio-sensor. The computed features can be transmitted from the bio-sensor using wireless channel, and they can be used in the receiver for determining the absence of QRS complex in the segment. By avoiding computationally expensive signal reconstruction for the ECG segments without QRS complex, it is shown, using simulation results, that computation time can be reduced by approximately 7.4% for long-term telemonitoring of QRS-based arrhythmia. Detection of the absence of QRS complex can be carried out in around 7 milliseconds in a standard laptop computer with 2.2GHz processor and 8GB RAM.A novel method for efficient telemonitoring of arrhythmia based on using QRS complexes is proposed. Two features, namely, sum of absolute differences (SAD) and maximum of absolute differences (MAD) are efficiently computed for each ECG segment in the bio-sensor. The computed features can be transmitted from the bio-sensor using wireless channel, and they can be used in the receiver for determining the absence of QRS complex in the segment. By avoiding computationally expensive signal reconstruction for the ECG segments without QRS complex, it is shown, using simulation results, that computation time can be reduced by approximately 7.4% for long-term telemonitoring of QRS-based arrhythmia. Detection of the absence of QRS complex can be carried out in around 7 milliseconds in a standard laptop computer with 2.2GHz processor and 8GB RAM.

Compressive sensing of foot-gait signals by enhancing group block-sparse structure on the first-order difference

A new technique for improving the signal reconstruction performance for compressive sensing of gait signals is proposed. The algorithm is based on the minimization of a pseudo-norm which promotes group-block-sparse structure on the first-order difference of the signal. Signal blocks in foot gait signals occur as groups, and the locations of the group are estimated based on the regularization promoting block-sparse structure. The group locations are used for minimizing the pseudonorm for promoting group-block-sparse structure. Simulation results demonstrate that the proposed technique yields upto 0.76dB improvement in the reconstruction performance for foot-gait signals relative to the algorithms promoting block-sparse structure.