Time-Frequency Analysis, Denoising, Compression, Segmentation, and Classification of PCG Signals

Chowdhury, M.; Poudel, Khem Narayan; Hu, Yating

doi:10.1109/access.2020.3020806

Cited by 73 publications

(30 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A unified lookup table (LUT) using the trigonometric approximation method as shown in (3) is adopted in the implementation of Jo et al [15] by (4), where 0 ≤ addr < 2L, L = 32, α = addr AND 110000(2), β = addr AND 001100 (2), and γ = addr AND 000011 (2).…”

Section: B Trigonometric Approximation Methodsmentioning

confidence: 99%

“…Recently, Mel-scale frequency cepstral coefficients (MFCC) [1], [2] have been used to generate the feature vectors of sounds in speech recognition by combining them with convolutional neural network (CNN) and deep neural network (DNN) models [3], [4]. Moreover, the MFCC-based deep learning recognition algorithm has been widely used in heart sound recognition [4], semantic analysis [3], emotion analysis [5], and keyword detection [6], [7], [8]. Both fast Fourier transform (FFT) and discrete cosine transform (DCT) are very computationally intensive in MFCC processing.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hardware Accelerator Design of DCT Algorithm With Unique-Group Cosine Coefficients for Mel-Scale Frequency Cepstral Coefficients

et al. 2022

View full text Add to dashboard Cite

This study presents a compact L-points discrete cosine transform (DCT) hardware accelerator for M-points Mel-scale Frequency Cepstral Coefficients (MFCC). The main contributions of this work can be summarized as 1) proposing an algorithm with lower complexity; 2) achieving higher accuracy performance; 3) implementing a low-cost accelerator with a unique group of cosine coefficients. For algorithm derivation, the proposed method converts the original formula into the type IV of discrete cosine transform (DCT-IV) with a preprocessing procedure. The kernel computation of DCT-IV can be further derived into the same cosine multiplication with the proposed preprocessing. Therefore, a total of (M-1) (L-1) additions, (M-1) L multiplications, and L coefficients are required for the computation. Compared with Jo et al.'s algorithm, the proposed method respectively reduces the number of additions and multiplications by 42.32 % and 41.67 %. Instead, the number of coefficients is increased by 33.33 %. Moreover, the proposed algorithm exhibits a higher peak signal-to-noise ratio (PSNR) value which is achieved at 90.1dB with a 16bit coefficient word length. For hardware realization, the FPGA implementation results show that it can operate at a clock rate of 135.85 MHz and requires only 113 combinational elements, 87 registers, 3 DSP multipliers, 64×16 bits RAM and 32×16 bits ROM. Overall, it would be a good choice for integrating MFCC applications in the future.INDEX TERMS discrete cosine transform (DCT); Hardware Accelerator; Mel-scale Frequency Cepstral Coefficients (MFCC)

show abstract

Section: B Trigonometric Approximation Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Hardware Accelerator Design of DCT Algorithm With Unique-Group Cosine Coefficients for Mel-Scale Frequency Cepstral Coefficients

et al. 2022

View full text Add to dashboard Cite

show abstract

“…This makes it therefore essential to associate a quality value with the computed parameters to better inform the user about the reliability of the measure [ 18 ]. Third, while the use of wearable devices allows for extensive and unique collection of physiological signals, the increasing amount of data requires an ever-increasing degree of automation in processing, highlighting caveats to traditional methods [ 19 ], and paving the way for new data-driven approaches relying upon machine and deep learning [ 20 , 21 , 22 , 23 , 24 ]. In order to address the three above issues, this paper proposes to couple a chest-worn apparatus [ 8 ], called Soundi (Biocubica Srl, Milan, Italy), able to concurrently register electrocardiogram (ECG), photopletismogram (PPG), phonocardiogram (PCG), and seismocardiogram (SCG), with a novel neural architecture, called eMTUnet (enriched multi-task Unet), to identify cardiovascular-related characteristic points.…”

Section: Introductionmentioning

confidence: 99%

“…In order to detect artifacts in the PPG signal, 1-D CNN was proposed [ 23 ], the localization of relevant points in the waveform was not carried out though. In PCG, conventional signal processing based on the Mel-scale wavelet transform and Shannon energy has been shown to be effective only for steady-state acquisitions [ 19 , 20 ]. The processing of simultaneous PPG and PCG signals was exploited in [ 32 ] by integrating the PPG waveform into the calculation of the heart sound envelopes using Shannon entropy.…”

Section: Introductionmentioning

confidence: 99%

Identification of Characteristic Points in Multivariate Physiological Signals by Sensor Fusion and Multi-Task Deep Networks

Rossi

Alessandrelli

Dombrovschi

et al. 2022

Sensors

View full text Add to dashboard Cite

Identification of characteristic points in physiological signals, such as the peak of the R wave in the electrocardiogram and the peak of the systolic wave of the photopletismogram, is a fundamental step for the quantification of clinical parameters, such as the pulse transit time. In this work, we presented a novel neural architecture, called eMTUnet, to automate point identification in multivariate signals acquired with a chest-worn device. The eMTUnet consists of a single deep network capable of performing three tasks simultaneously: (i) localization in time of characteristic points (labeling task), (ii) evaluation of the quality of signals (classification task); (iii) estimation of the reliability of classification (reliability task). Preliminary results in overnight monitoring showcased the ability to detect characteristic points in the four signals with a recall index of about 1.00, 0.90, 0.90, and 0.80, respectively. The accuracy of the signal quality classification was about 0.90, on average over four different classes. The average confidence of the correctly classified signals, against the misclassifications, was 0.93 vs. 0.52, proving the worthiness of the confidence index, which may better qualify the point identification. From the achieved outcomes, we point out that high-quality segmentation and classification are both ensured, which brings the use of a multi-modal framework, composed of wearable sensors and artificial intelligence, incrementally closer to clinical translation.

show abstract

“…For an urban sound event, several typical sound characteristics like mel-frequency coefficients [24], zero-crossing [25], and wavelet transformation [26] have been used as sound features representation. Many classifiers are commonly applied in sound-related classification problems, for instance, support vector machines, extreme learning machines, and Gaussian mixture, in addition to standard machine learning algorithms such as knearest neighbors [27][28][29][30].…”

Section: Introductionmentioning

confidence: 99%

An Ensemble One Dimensional Convolutional Neural Network with Bayesian Optimization for Environmental Sound Classification

et al. 2021

View full text Add to dashboard Cite

With the growth of deep learning in various classification problems, many researchers have used deep learning methods in environmental sound classification tasks. This paper introduces an end-to-end method for environmental sound classification based on a one-dimensional convolution neural network with Bayesian optimization and ensemble learning, which directly learns features representation from the audio signal. Several convolutional layers were used to capture the signal and learn various filters relevant to the classification problem. Our proposed method can deal with any audio signal length, as a sliding window divides the signal into overlapped frames. Bayesian optimization accomplished hyperparameter selection and model evaluation with cross-validation. Multiple models with different settings have been developed based on Bayesian optimization to ensure network convergence in both convex and non-convex optimization. An UrbanSound8K dataset was evaluated for the performance of the proposed end-to-end model. The experimental results achieved a classification accuracy of 94.46%, which is 5% higher than existing end-to-end approaches with fewer trainable parameters. Four measurement indices, namely: sensitivity, specificity, accuracy, precision, recall, F-measure, area under ROC curve, and the area under the precision-recall curve were used to measure the model performance. The proposed approach outperformed state-of-the-art end-to-end approaches that use hand-crafted features as input in selected measurement indices and time complexity.

show abstract

Time-Frequency Analysis, Denoising, Compression, Segmentation, and Classification of PCG Signals

Cited by 73 publications

References 21 publications

Hardware Accelerator Design of DCT Algorithm With Unique-Group Cosine Coefficients for Mel-Scale Frequency Cepstral Coefficients

Hardware Accelerator Design of DCT Algorithm With Unique-Group Cosine Coefficients for Mel-Scale Frequency Cepstral Coefficients

Identification of Characteristic Points in Multivariate Physiological Signals by Sensor Fusion and Multi-Task Deep Networks

An Ensemble One Dimensional Convolutional Neural Network with Bayesian Optimization for Environmental Sound Classification

Contact Info

Product

Resources

About