Pith. sign in

REVIEW 5 major objections 5 minor 29 references

A Simultaneous ECG-PCG Acquisition System with Real-Time Burst-Adaptive Noise Cancellation

T0 review · 5 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A portable ECG–PCG device with burst-adaptive noise cancellation lifts real-world hospital signal-to-noise ratios by 30.32 dB (ECG) and 37.01 dB (PCG).

desk verdict Solid prototype and honest simulation, but the headline real-world SNR numbers are likely a fixed-filter artifact and the ANC contribution is unproven in the field. read the letter →

arxiv 2510.23819 v3 pith:CVCY7TAK submitted 2025-10-27 eess.SY cs.SYeess.SP

Avishka Herath , Malith Jayalath , Kumudu Kaushalya , Sanjana Kapukotuwa , Chathuni Wijegunawardena , Pahan Mendis , Kithmin Wickremasinghe , Duminda Samarasinghe
show 2 more authors
Wageesha N. Manamperi Chamira U. S. Edussooriya
This is my paper · ORCID
classification eess.SYcs.SYeess.SP
keywords adaptivenoisecancellationphonocardiogramelectrocardiogramburstnormalizedleastmeansquarereal-timedenoisingportablecardiacscreeninghospital
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that a low-cost, portable device can simultaneously record ECG and heart-sound (PCG) signals and clean them in real time even in loud hospital wards. The key move is a burst-adaptive normalized least-mean-square filter that speeds up adaptation whenever the ambient-noise reference jumps in energy, addressing the non-stationary bursts that defeat standard adaptive filters. On real hospital recordings the authors report SNR improvements of 30.32 dB for ECG and 37.01 dB for PCG. If correct, the system makes reliable cardiac screening feasible in resource-constrained settings without expensive commercial digital stethoscopes.

What carries the argument

The central mechanism is the burst-adaptive NLMS update: an energy tracker maintains a moving average of reference power; whenever instantaneous reference power exceeds a threshold times that average, the normalized step size is multiplied by a scaling factor, so the filter reacts quickly to noise bursts, then resumes small-step operation. The reference comes from a quad-microphone array facing outward on the stethoscope body, intended to capture ambient noise uncorrelated with heart sounds.

What would settle it

Compute the coherence between the ambient reference and the noise component in the stethoscope microphone during hospital recordings; if it is low across 20–200 Hz, the reported 37 dB gain would not generalize to other hardware placements.

Watch

Extended reading notes

Core claim

The authors set out to show that a single portable device can acquire ECG and PCG simultaneously and clean both in real time with modest compute. The specific discovery is that a normalized LMS filter, modified to boost its step size during detected energy bursts, adapts to non-stationary hospital noise faster than standard LMS/NLMS while staying stable in quiet intervals. On hospital recordings made with the prototype, the PCG pipeline improved SNR by 37.01 dB and the ECG pipeline by 30.32 dB; on a synthetic noisy PCG dataset, the burst-adaptive filter beat both LMS and NLMS on all four metrics for normal and abnormal recordings.

Load-bearing premise

The result rests on the assumption that the outward-facing ambient microphones provide a noise reference closely correlated with the noise at the stethoscope and free of heart sounds; the paper does not measure that correlation.

Editorial extensions

If this is right

  • Real-time cleaning of both modalities can move onto an embedded microcontroller, so synchronized ECG–PCG screening no longer depends on offline processing or bulky equipment.
  • A 37.01 dB PCG gain means murmur features that a human or algorithm needs can survive in a busy ward, not just in a quiet exam room.
  • A 30.32 dB ECG gain preserves the timing landmarks needed to place S1 and S2 within the cardiac cycle when the two signals are read together.
  • On burst-heavy noise, the kind produced by alarms, trolleys, and conversations, the burst-adaptive filter outperforms conventional LMS and NLMS, indicating the approach targets the noise profile that actually limits auscultation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper does not report the coherence between the ambient reference and the noise reaching the chest microphone; measuring that coherence would show whether the 37 dB gain transfers to other stethoscope geometries and body positions.
  • The ECG gain comes from fixed elliptic filters rather than adaptive cancellation, so adding an adaptive ECG stage or motion-artifact rejection is a natural extension for ambulatory use.
  • A testable consequence of the burst-adaptive idea is that the improvement over standard NLMS should grow with the proportion of time spent in burst events; a dataset with controlled burst density could verify that dose–response.
  • Because the device uses dry electrodes and outward-facing mics, performance during physical motion is an open question; a movement-artifact protocol would map that boundary.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. This paper presents a portable simultaneous ECG-PCG acquisition device with a real-time adaptive noise cancellation pipeline. The PCG channel uses a proposed burst-adaptive NLMS (BA-NLMS) algorithm that temporarily increases the step size when the reference-signal energy exceeds a moving average; the ECG channel uses a cascade of fixed elliptic filters (0.5 Hz high-pass, 150 Hz low-pass, 49.5–50.5 Hz bandstop). The BA-NLMS is evaluated on 3934 simulated noisy PCG recordings (per class) built from the AUTHHSDB dataset and hospital noise, comparing against LMS and NLMS. The complete prototype is then evaluated on a single 14-second hospital recording, reporting ΔSNR of 37.01 dB for PCG and 30.32 dB for ECG using a hand-defined SNR metric (Eq. 6). The abstract additionally states that a complexity analysis confirms the pipeline's suitability for embedded implementation, but no such analysis appears in the manuscript.

Significance. If the real-world SNR improvements were fully supported, this device would be a meaningful low-cost cardiac screening tool for noisy, resource-constrained settings. The simulation study is a genuine strength: it uses a public dataset, a large number of recordings, several power-independent metrics (NMSE, ΔSNR, CC, NMAE), and direct comparisons against LMS and NLMS. The BA-NLMS shows consistent but modest gains over NLMS in simulation (about 0.5 dB in the full noisy set and about 1.9 dB on burst-only recordings). However, the headline real-world numbers are not currently supported: the frequency bands in the SNR metric are never specified, there is no ablation isolating the adaptive contribution from fixed band-pass filtering, the ECG improvement comes entirely from fixed filters rather than adaptive cancellation, and the validation rests on a single 14-second trace. The missing complexity analysis also undercuts the embedded-suitability claim. With a proper ablation, more recordings, and a defined evaluation protocol, this work could become a useful contribution.

major comments (5)
  1. [Section III-B, Eq. (6)] The real-world validation rests on a single 14-s recording and an undefined SNR metric. Eq. (6) defines SNR = 10 log10( P_LF / (P_LFN + P_HF) ), but the frequency ranges LF, LFN, and HF are never specified. Since the PCG pipeline ends with a 20–200 Hz band-pass filter (Section III-B), a fixed filter that removes energy outside the assumed 'signal band' will increase this ratio by construction. The reported 37.01 dB improvement may therefore largely reflect band-limiting rather than the BA-NLMS algorithm. Please specify the frequency bands, report the ΔSNR achieved by the fixed filters alone, and evaluate on more than one recording. Ideally, use a metric anchored to a known clean reference or a blind quality measure.
  2. [Section II-C] The ECG improvement of 30.32 dB is not attributable to the proposed adaptive noise-cancellation pipeline. Section II-C describes only a cascade of fixed elliptic IIR filters (0.5 Hz high-pass, 150 Hz low-pass, 49.5–50.5 Hz bandstop); there is no adaptive component in the ECG path. The abstract and conclusion attribute this improvement to the proposed system's noise-cancellation capabilities. Please provide an ablation showing the contribution of each stage, or clearly state that the ECG ΔSNR is a fixed-filter result and not the output of the adaptive algorithm.
  3. [Abstract / Section IV] The abstract states that a complexity analysis confirms the pipeline's suitability for embedded implementation, but no such analysis appears in Sections II–IV. In fact, Section IV lists 'implementing the denoising algorithms on the microcontroller for fully embedded operation' as future work, and Section II-A says data are streamed to a computer for real-time ANC. Thus the claims of an integrated, embedded, real-time system are not currently supported. Either add the complexity analysis and clarify where processing is performed, or temper the claims accordingly.
  4. [Section II-B] The central ANC assumption is unverified on the prototype. Section II-B states that the reference signal r(n) is uncorrelated with heart sounds but closely correlated with the noise reaching the stethoscope microphone. No measurement of that correlation is presented. In the synthetic evaluation, the reference is generated from the same noise source through a moving-average filter (Section III-A), so it is highly correlated by construction and does not test the real-world assumption. Please report the measured coherence/correlation between the ambient microphone array and the noise component in the stethoscope channel, and evaluate the algorithm under a decorrelated reference.
  5. [Section III-A] The BA-NLMS parameters (μ0 = 0.05, α = 0.99, η = 5, β = 6, ε = 0.001, filter length 10) are stated without a fitting procedure or sensitivity analysis. The burst-detection mechanism depends critically on η and β, and no evidence is provided that these values generalize beyond the selected AUTHHSDB subset. Please include a sensitivity study or justify the values from a principled criterion; otherwise the robustness of the simulated and real-world gains is unclear.
minor comments (5)
  1. [Section III-A] The sentence 'This process generated 3934 noisy recordings, each for normal and abnormal' is ambiguous. The numbers suggest 7 clean normal × 562 noise = 3934 per class; please state explicitly whether the total is 3934 or 7868.
  2. [Equation (6), Fig. 4] The input SNR values reported for the real-world recordings (e.g., -9.08 dB for ECG and -2.08 dB for PCG) are not defined. State whether they use the same Eq. (6) and, if so, give the exact frequency bands used.
  3. [Section II-B, Eqs. (2) and (5)] Notation is inconsistent: Eq. (2) uses ||r(n)||² while Eq. (5) uses r^T(n)·r(n) for the same quantity. Unify the notation.
  4. [Abstract] The abstract in the submission says 'the proposed system achieved signal-to-noise ratio improvements,' whereas the abstract in the full text says 'the proposed algorithms achieved.' Please use consistent wording, especially because the ECG path is not adaptive.
  5. [Figure 4] The time-domain plots and spectrograms lack axis labels/units (time in seconds, amplitude in what units) and the spectrogram color scale is not described. Please add these details.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: BA-NLMS is benchmarked on external AUTHHSDB data; the weak real-world SNR metric is a validation concern, not a definitional identity.

full rationale

The paper's central algorithmic claim is the BA-NLMS filter, evaluated in Section III-A on clean AUTHHSDB PCG signals corrupted with hospital noise and compared against standard LMS/NLMS using NMSE, ΔSNR, CC, and NMAE. These metrics are not defined in terms of the filter's parameters, and the noise/reference construction follows an external MathWorks procedure; the parameter values (μ0=0.05, α=0.99, η=5, β=6, ε=0.001) are fixed and reported, not fitted to the test set, so no prediction reduces to an input by construction. The real-world validation (Section III-B) is methodologically weak: Eq. (6) leaves the LF/LFN/HF bands unspecified, the PCG pipeline includes a fixed 20-200 Hz band-pass that could alone raise a broadband-power-ratio SNR measure, and the ECG pipeline contains only fixed elliptic filters, so attributing the 30.32 dB ECG gain to the 'proposed system' is overstated. These are evaluation-design and reporting issues that affect the evidentiary value of the headline numbers; they are not circular derivations, and no equation in the paper reduces to its own inputs. The only same-group citation ([12]) is background on ECG-PCG classification and is not load-bearing. No circular step is therefore identified.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central system rests on five hand-set filter parameters, the SNR metric's unspecified frequency bands, and the physical assumption that a separate ambient mic array provides a clean noise reference. No new physical entities are introduced.

free parameters (7)
  • μ0 = 0.05
    Base NLMS step size; set manually, no sensitivity analysis.
  • α = 0.99
    Smoothing coefficient for running average energy estimate.
  • η = 5
    Burst detection threshold (instantaneous-to-average energy ratio).
  • β = 6
    Step-size boost during detected bursts.
  • ε = 0.001
    Regularization in NLMS normalization.
  • Filter length M = 10
    FIR filter order; stated without justification.
  • SNR band definitions (LF, LFN, HF) = not specified in text
    SNR metric defining signal and noise frequency ranges; numerical values not reported.
assumptions (4)
  • domain assumption Reference signal r(n) is uncorrelated with heart sounds and closely correlated with the noise at the stethoscope microphone (Sec. II-B).
    If the ambient microphone reference misses the actual noise path or picks up heart sounds, the ANC will distort the PCG. No measurement of reference correlation is provided.
  • domain assumption The moving-average/autoregressive procedure in the simulated test produces a valid noise-reference pair (Sec. III-A, following [29]).
    Simulated environment may not match acoustic coupling in the real device.
  • standard math Standard NLMS convergence theory applies (Eq. 1-2, citing [26],[27]).
    Background assumption of adaptive filter convergence.
  • domain assumption Hospital noise is dominated by sudden high-energy bursts, motivating the threshold-based step control (Eq. 4).
    The burst detector assumes this noise character; no characterization of hospital noise statistics is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Simultaneous ECG-PCG Acquisition System with Real-Time Burst-Adaptive Noise Cancellation." pith.science (2026). https://pith.science/paper/CVCY7TAK

@misc{pith2026251023819,
  author       = {Pith},
  title        = {Pith review of: A Simultaneous ECG-PCG Acquisition System with Real-Time Burst-Adaptive Noise Cancellation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CVCY7TAK}},
  note         = {Machine review of arXiv:2510.23819}
}
read the original abstract

Cardiac auscultation is an essential clinical skill, requiring excellent hearing to distinguish subtle differences in timing and pitch of heart sounds. However, diagnosing solely from these sounds is often challenging due to interference from surrounding noise, and the information may be limited. Most of the existing solutions that adaptively cancel external noise are either non-real-time or computationally intensive, making them unsuitable for implementation in a portable system. This work proposes an end-to-end system with a real-time adaptive noise cancellation pipeline integrated into a device that simultaneously acquires electrocardiogram (ECG) and phonocardiogram (PCG) signals. We employ a burst-adaptive normalized least mean square algorithm that adjusts its adaptation in response to high-energy, non-stationary hospital noise. The algorithm's performance was initially assessed using datasets with artificially induced noise. Subsequently, the complete end-to-end system was validated using real-world hospital recordings captured with the dual-modality device. For ECG and PCG signals recorded from the device in noisy hospital settings, the proposed system achieved signal-to-noise ratio improvements of 30.32 dB and 37.01 dB, respectively. Furthermore, complexity analysis confirms the pipeline's suitability for embedded implementation. These results demonstrate the system's effectiveness in enabling reliable and accessible cardiac screening in noisy hospital environments typical of resource-constrained settings.

Figures

Figures reproduced from arXiv: 2510.23819 by the authors.

Figure 1
Figure 1. Block diagram of the overall system architecture. Signals from dry Ag/AgCl electrodes (acquiring ECG) and digital microelectromechanical system (MEMS) microphones (acquiring PCG) are captured, processed by a microcontroller unit (MCU), and transmitted wirelessly to a computer for real-time ANC and signal visualization. tively denoising PCG signals in real-time while maintaining low computational complexity for embed… view at source ↗
Figure 2
Figure 2. Photographs of the prototype with the design features annotated. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Denoising performance of the Burst-Adaptive NLMS filter on a PCG [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Denoising performance of the proposed real-time filter algorithms on recorded noisy ECG and PCG signals. Spectrograms (left) and time-domain [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references

  1. [1]

    Global birth prevalence of congenital heart defects 1970- 2017: updated systematic review and meta-analysis of 260 studies,

    Y . Liu, S. Chen, L. Z ¨uhlke, G. C. Black, M. K. Choy, N. Li, and B. D. Keavney, “Global birth prevalence of congenital heart defects 1970- 2017: updated systematic review and meta-analysis of 260 studies,” International Journal of Epidemiology, vol. 48, no. 2, pp. 455–463, 2019

    1970

  2. [2]

    Global, regional, and national burden of congenital heart disease, 1990–2017: a systematic analysis for the global burden of disease study 2017,

    C. A. Sable, “Global, regional, and national burden of congenital heart disease, 1990–2017: a systematic analysis for the global burden of disease study 2017,”Lancet Child & Adolescent Health, vol. 4, pp. 185–200, 2020

    1990

  3. [3]

    Global, regional, and national epidemiology of congenital heart disease in children from 1990 to 2021,

    Jiaoli Xu, Qinhong Li, Lili Deng, Jingxuan Xiong, Zugen Cheng, and Caixia Ye, “Global, regional, and national epidemiology of congenital heart disease in children from 1990 to 2021,”Frontiers in Cardiovas- cular Medicine, vol. 12, pp. 1522644, 2025

    1990

  4. [4]

    Cardiovascular diseases (CVDs) — fact sheet,

    World Health Organization, “Cardiovascular diseases (CVDs) — fact sheet,” https://www.who.int/news-room/fact-sheets/detail/ cardiovascular-diseases-(cvds), Jul 2025, Accessed: October 20, 2025

    2025

  5. [5]

    An intelligent stethoscope with ECG and heart sound synchronous display,

    Y . Lin, C. Chuang, C. Yen, S. Huang, P. Huang, J. Chen, and S. Lee, “An intelligent stethoscope with ECG and heart sound synchronous display,” in2019 IEEE International Symposium on Circuits and Systems (ISCAS), May 2019, pp. 1–4

    2019

  6. [6]

    A low-cost 3-d printed stethoscope connected to a smartphone,

    C. Aguilera-Astudillo, M. Chavez-Campos, A. Gonzalez-Suarez, and J. L. Garcia-Cordero, “A low-cost 3-d printed stethoscope connected to a smartphone,” inAnnual International Conference of the IEEE Engineering in Medicine and Biology, August 2016, pp. 4365–4368

    2016

  7. [7]

    Features for heartbeat sound signal normal and pathological,

    D. S. M. Amin and B. R. Fethi, “Features for heartbeat sound signal normal and pathological,”Recent Patents on Computer Science (Discontinued), vol. 1, no. 1, pp. 1–8, 2008

    2008

  8. [8]

    Michael Klum, Mike Urban, Timo Tigges, Alexandru-Gabriel Pielmus, Aarne Feldheiser, Theresa Schmitt, and Reinhold Orglmeister, “Wear- able cardiorespiratory monitoring employing a multimodal digital patch stethoscope: Estimation of ECG, PEP, LVET and respiration using a 55 mm single-lead ECG and phonocardiogram,”Sensors, vol. 20, no. 7, 2020

    2020

Show all 29 references
  1. [9]

    Assessing PEP and LVET from heart sounds: algorithms and evaluation,

    Rui Pedro Paiva, P Carvalho, Xavier Aubert, Jens Muehlsteff, Jorge Henriques, and Manuel Antunes, “Assessing PEP and LVET from heart sounds: algorithms and evaluation,” in2009 Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE, 2009, ...

    2009

  2. [10]

    Timing the opening and closure of the aortic valve using a phonocardiogram envelope: a performance test for systolic time intervals measurement,

    A ´ıda Jim ´enez-Gonz´alez, “Timing the opening and closure of the aortic valve using a phonocardiogram envelope: a performance test for systolic time intervals measurement,”Physiological measurement, vol. 42, no. 2, pp. 025004, 2021

    2021

  3. [11]

    NMCSE: Noise-robust multi-modal coupling signal estimation method via optimal transport for cardiovascular disease detection,

    Peihong Zhang, Zhixin Li, Rui Sang, Yuxuan Liu, Yiqiang Cai, Yizhou Tan, and Shengchen Li, “NMCSE: Noise-robust multi-modal coupling signal estimation method via optimal transport for cardiovascular disease detection,”arXiv preprint arXiv:2505.18174, 2025

    2025

  4. [12]

    A novel transfer learning-based approach for screening pre-existing heart diseases using synchronized ECG signals and heart sounds,

    Ramith Hettiarachchi, Udith Haputhanthri, Kithmini Herath, Hasindu Kariyawasam, Shehan Munasinghe, Kithmin Wickramasinghe, Duminda Samarasinghe, Anjula De Silva, and Chamira U. S. Edussooriya, “A novel transfer learning-based approach for screening pre-existing heart diseases ...

    2021

  5. [13]

    Multi-modal cardiac function signals classification algorithm based on improved DS evidence theory,

    Jinghui Li, Li Ke, Qiang Du, Xiangmin Chen, and Xiaodi Ding, “Multi-modal cardiac function signals classification algorithm based on improved DS evidence theory,”Biomedical Signal Processing and Control, vol. 71, pp. 103078, 2022

    2022

  6. [14]

    Co-learning-assisted progressive dense fusion network for cardiovascular disease detection using ECG and PCG signals,

    Haobo Zhang, Peng Zhang, Fan Lin, Lianying Chao, Zhiwei Wang, Fei Ma, and Qiang Li, “Co-learning-assisted progressive dense fusion network for cardiovascular disease detection using ECG and PCG signals,”Expert Systems with Applications, vol. 238, pp. 122144, 2024

    2024

  7. [15]

    Available: https://www

    3M Littmann Stethoscopes, “3M,” [Online]. Available: https://www. littmann.com/3M/en US/littmann-stethoscopes/, [Accessed: October 21, 2025]

    2025

  8. [16]

    Thinklabs one digital stethoscope,

    Thinklabs Medical LLC, “Thinklabs one digital stethoscope,” [Online]. Available: https://www.thinklabs.com/, [Accessed: October 21, 2025]

    2025

  9. [17]

    Available: https://www.ekohealth

    Eko Devices Inc., “Eko,” [Online]. Available: https://www.ekohealth. com/, [Accessed: October 21, 2025]

    2025

  10. [18]

    Ayu devices,

    Ayu Devices Pvt. Ltd., “Ayu devices,” [Online]. Available: https://www. ayudevices.store/, [Accessed: October 21, 2025]

    2025

  11. [19]

    A novel wearable device integrating ECG and PCG for cardiac health monitoring,

    Junbin Zang, Qi An, Bo Li, Zhidong Zhang, Libo Gao, and Chenyang Xue, “A novel wearable device integrating ECG and PCG for cardiac health monitoring,”Microsystems & Nanoengineering, vol. 11, no. 1, pp. 7, 2025

    2025

  12. [20]

    Wireless broadband acousto-mechanical sens- ing system for continuous physiological monitoring,

    Jae-Young Yoo, Seyong Oh, Wissam Shalish, Woo-Youl Maeng, Emily Cerier, Emily Jeanne, Myung-Kun Chung, Shasha Lv, Yunyun Wu, Seonggwang Yoo, et al., “Wireless broadband acousto-mechanical sens- ing system for continuous physiological monitoring,”Nature Medicine, vol. 29, no. 1...

    2023

  13. [21]

    Portable ECG and PCG wireless acquisition system and multiscale CNN feature fusion Bi-LSTM network for coronary artery disease diagnosis,

    Junye Lin, Shaokui Wang, Weipeng Xuan, Ding Chen, Fuhai Liu, Jinkai Chen, Shudong Xia, Shurong Dong, and Jikui Luo, “Portable ECG and PCG wireless acquisition system and multiscale CNN feature fusion Bi-LSTM network for coronary artery disease diagnosis,”Computers in Biology a...

    2025

  14. [22]

    Design and test of a PCG-ECG syn- chronous auscultation system,

    Pengcheng Shi, Xinyi Wang, Guojun Zhang, Wendong Zhang, Jiangong Cui, Hua Tian, and Guofu Wang, “Design and test of a PCG-ECG syn- chronous auscultation system,”IEEE Transactions on Instrumentation and Measurement, vol. 73, pp. 1–14, 2024

    2024

  15. [23]

    A robust low- cost adaptive filtering technique for phonocardiogram signal denoising,

    S Hannah Pauline and Samiappan Dhanalakshmi, “A robust low- cost adaptive filtering technique for phonocardiogram signal denoising,” Signal Processing, vol. 201, pp. 108688, 2022

    2022

  16. [24]

    An end-to-end deep learning framework for real-time denoising of heart sounds for cardiac disease detection in unseen noise,

    Shams Nafisa Ali, Samiul Based Shuvo, Muhammad Ishtiaque Sayeed Al-Manzo, Anwarul Hasan, and Taufiq Hasan, “An end-to-end deep learning framework for real-time denoising of heart sounds for cardiac disease detection in unseen noise,”IEEE Access, vol. 11, pp. 87887– 87901, 2023

    2023

  17. [25]

    Enhancing PCG signal quality through cascaded adaptive noise cancelling with metaheuristic optimization,

    Madhava Rao Alla and Chandan Nayak, “Enhancing PCG signal quality through cascaded adaptive noise cancelling with metaheuristic optimization,”Circuits, Systems, and Signal Processing, pp. 1–40, 2025

    2025

  18. [26]

    Filtering for Removal of Artifacts,

    Rangaraj M. Rangayyan, “Filtering for Removal of Artifacts,” in Biomedical Signal Analysis: A Case-Study Approach, pp. 73–176. IEEE, 2002

    2002

  19. [27]

    On the convergence behavior of the LMS and the normalized LMS algorithms,

    D.T.M. Slock, “On the convergence behavior of the LMS and the normalized LMS algorithms,”IEEE Transactions on Signal Processing, vol. 41, no. 9, pp. 2811–2825, 1993

    1993

  20. [28]

    An open access database for the evaluation of heart sound algorithms,

    Chengyu Liu, David Springer, Qiao Li, Benjamin Moody, Ricardo Abad Juan, Francisco J Chorro, Francisco Castells, Jos ´e Millet Roig, Ikaro Silva, Alistair E W Johnson, Zeeshan Syed, Samuel E Schmidt, Chrysa D Papadaniil, Leontios Hadjileontiadis, Hosein Naseri, Ali Moukadem, A...

    2016

  21. [29]

    Enhance a signal using LMS and normalized LMS algorithms,

    The MathWorks Inc., “Enhance a signal using LMS and normalized LMS algorithms,” [Online]. Available: https://mathworks.com/help/dsp/ ug/enhance-a-signal-using-lms-and-normalized-lms-algorithms.html, Accessed: October 20, 2025

    2025

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.