Blind recovery of cardiac and respiratory sounds using non-negative matrix factorization &amp;amp; time-frequency masking

Shah, Ghafoor; Papadias, Constantinos B.

doi:10.1109/bibe.2013.6701542

Cited by 9 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nonetheless, adaptive filters do not completely remove heart sound due to their non-stationary nature which makes time alignment between the primary and reference signals difficult to apply. 3,37,38 Furthermore, it is not always possible to obtain an adequate reference signal that allows achieving separation. 17,31 In some cases, an electrocardiogram (ECG) 36,39 is used as the reference signal, which may not always be available.…”

Section: Related Workmentioning

confidence: 99%

“…43 Also, depending on the database there are mother Wavelets, thresholds and decomposition levels that are better adapted, generating an additional challenge in real scenarios. 3,37 Some methods used time-frequency representations in order to recognize the segments where the heart sound is located to eliminate that segment and to reconstruct it based on the adjacent information, which should correspond to a pure lung sound. In 11 the Poincaré recurrence, Takens' theorem and a trajectory algorithm are used.…”

Section: Related Workmentioning

confidence: 99%

“…However, most modern digital stethoscopes allow you to record an observation on a single channel. 37 Non-negative Matrix Factorization (NMF) is another type of BSS approach that uses a time-frequency representation of the cardio-respiratory signal. In 26 the use of NMF is proposed using information from the Constant-Q Transform (CQT) of the signal for the construction of a binary mask.…”

Section: Related Workmentioning

confidence: 99%

“…In 37 a staged semi-supervised NMF problem is proposed, in which initially an estimated heart sound decomposition is obtained, then the cardio-respiratory sound and finally the cardio-respiratory sound with noise. In works such as 13,37 clustering criteria is proposed to classify each obtained component. One of the main difficulties with NMF is that there is no a systematic method to determine the quantity of components that optimizes the separation process.…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Source separation for single channel thoracic cardio-respiratory sounds applying Non-negative Matrix Factorization (NMF) using a focused strategy on heart sound positions

Arce¹,

Cuadra

2023

18th International Symposium on Medical Information Processing and Analysis

View full text Add to dashboard Cite

Auscultation using stethoscopes allows the diagnosis of respiratory and cardiac diseases. However, these sounds interfere with each other both in time and frequency. In the case of recording heart sounds, it is possible to ask the patient to stop their breathing to perform auscultation and obtain a pure heart sound. But, in the case of lung sounds it is impossible to do the same. In this paper, a source separation method based on Non-negative Matrix Factorization (NMF) is used to decompose a signal into different c omponents. T he m ethod proposed uses information from the estimated lung sound to reinsert the segments of interest into the original signal. The objective of this approximation is not to distort the segments of pure respiratory sound (free of heart sound). This method is compared to a base case of NMF decomposition on the raw signal. Three criteria for classifying the components based on the literature are also proposed, which will allow to indicate which component corresponds to each sound. The results were evaluated using temporal and spectral correlations, mean square error (MSE) and signal to distortion ratio (SDR) between the original respiratory signal and the respiratory signal estimated through the algorithm. It is shown that the best approximation is the NMF decomposition on the entire signal & replacing segments under different p arameter variations.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Source separation for single channel thoracic cardio-respiratory sounds applying Non-negative Matrix Factorization (NMF) using a focused strategy on heart sound positions

Arce¹,

Cuadra

2023

18th International Symposium on Medical Information Processing and Analysis

View full text Add to dashboard Cite

show abstract

“…Source separation techniques based on non-negative matrix factorization have been used in single-channel source separation [16] and recovery of cardiac and respiratory sounds [17]. In this study, we use a source separation method based on the amplitude spectrogram (AS) to isolate a signal and estimate heart rate (HR) [18].…”

Section: Introductionmentioning

confidence: 99%

Heart pulse demodulation from Emfit mattress sensor using spectral and source separation techniques.

Perez-Macias¹,

Värri²,

Himanen³

et al. 2022

Computing in Cardiology Conference (CinC)

View full text Add to dashboard Cite

Introduction:Unobtrusive bed mattress sensors such us the electromechanical film transducer (Emfit) record the heart's activity, breathing, and body movements with clinical applications ranging from epilepsy to sleep disorders. The mechanical activity of the heart recorded using these sensors is known as ballistocardiogram (BCG). However, BCG shape changes on position, patient, and mattress. Aim: In this study, we isolate a positionindependent heart signal designed to retrieve the heart activity recorded using an Emfit mattress sensor using extended polysomnography (PSG) recordings as a reference. Methods: We used spectral and source separation techniques to infer the heart's Emfit Pulse (EP). We validated the resulting signal using representative 10minute normal breathing epochs extracted from the PSG from 33 subjects and by estimating the heart rate (HR) from the EP and compared it against the electrocardiogram (ECG) using non-overlapping oneminute windows. Results: Results show a signal similar in shape to a photoplethysmogram (PPG) with different timings in relation to the ECG's R-peak. We found good agreement between HR with a mean absolute error (MAE) of 2.4 beats per minute (bpm) with standard deviation of 4.6bpm. Conclusions: A position independent heart pulse signal from the Emfit mattress sensor was obtained in a range of subjects and validated by means of HR analysis.

show abstract

Heart-lung sound separation by nonnegative matrix factorization and deep learning

Wang

Qin

et al. 2023

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Blind recovery of cardiac and respiratory sounds using non-negative matrix factorization & time-frequency masking

Cited by 9 publications

References 9 publications

Source separation for single channel thoracic cardio-respiratory sounds applying Non-negative Matrix Factorization (NMF) using a focused strategy on heart sound positions

Source separation for single channel thoracic cardio-respiratory sounds applying Non-negative Matrix Factorization (NMF) using a focused strategy on heart sound positions

Heart pulse demodulation from Emfit mattress sensor using spectral and source separation techniques.

Heart-lung sound separation by nonnegative matrix factorization and deep learning

Contact Info

Product

Resources

About