Wrist pulse signal acquisition and analysis for disease diagnosis: A review

Guo, Chaoxun; He, Haoze; Liao, Yining; Zhang, David

doi:10.1016/j.compbiomed.2022.105312

Cited by 32 publications

(18 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a weak physiological signal, wrist pulse can be easily contaminated by various kinds of interference, including power frequency disturbances, amplitude oscillation caused breathing, muscle contraction and limb vibration [6], [33]. Previous studies have shown that the wrist pulse signal is usually located at low frequencies.…”

Section: Data Prep-processmentioning

confidence: 99%

“…However, long-term training and high reliance on the practi-tioner's subjective experience limit its effectiveness in practice [2]. Over the past decade, instead of relying on subjective perception, increasing interests have been focused on developing pattern recognition techniques to perform health diagnosis in terms of measured pulse signals with advanced sensors [3]- [6], which is referred as computational pulse analysis [7].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

NA-MEMD Is Actually What You Need for Computational Pulse Analysis

Tian

Qia

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

Section: Data Prep-processmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

NA-MEMD Is Actually What You Need for Computational Pulse Analysis

Tian

Qia

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Pulse diagnosis (PD), which evaluates health status by analyzing tactile radial arterial palpation with trained fingertips, plays an important role in oriental medicine including traditional Chinese medicine (TCM) and traditional Korean medicine (TKM) [ 4 ]. In practice, however, its effectiveness is limited by its reliance on long-term training and rich experience, which can lead to significant differences in diagnostic results among different doctors [ 5 , 6 ].…”

Section: Introductionmentioning

confidence: 99%

“…With the advantages of its adaptive feature learning capacity and multi-layer nonlinear mapping ability, deep learning models frequently outperform classic shallow methods as an end-to-end process [ 20 ]. For wrist pulse recognition, recently a series of work on deep learning has also been employed [ 5 ]. To classify pulse signals without extracting complicated features, a nine-layer deep convolutional neural network was proposed [ 21 ].…”

Section: Introductionmentioning

confidence: 99%

Towards Generating Realistic Wrist Pulse Signals Using Enhanced One Dimensional Wasserstein GAN

Chang

et al. 2023

Sensors

View full text Add to dashboard Cite

For the past several years, there has been an increasing focus on deep learning methods applied into computational pulse diagnosis. However, one factor restraining its development lies in the small wrist pulse dataset, due to privacy risks or lengthy experiments cost. In this study, for the first time, we address the challenging by presenting a novel one-dimension generative adversarial networks (GAN) for generating wrist pulse signals, which manages to learn a mapping strategy from a random noise space to the original wrist pulse data distribution automatically. Concretely, Wasserstein GAN with gradient penalty (WGAN-GP) is employed to alleviate the mode collapse problem of vanilla GANs, which could be able to further enhance the performance of the generated pulse data. We compared our proposed model performance with several typical GAN models, including vanilla GAN, deep convolutional GAN (DCGAN) and Wasserstein GAN (WGAN). To verify the feasibility of the proposed algorithm, we trained our model with a dataset of real recorded wrist pulse signals. In conducted experiments, qualitative visual inspection and several quantitative metrics, such as maximum mean deviation (MMD), sliced Wasserstein distance (SWD) and percent root mean square difference (PRD), are examined to measure performance comprehensively. Overall, WGAN-GP achieves the best performance and quantitative results show that the above three metrics can be as low as 0.2325, 0.0112 and 5.8748, respectively. The positive results support that generating wrist pulse data from a small ground truth is possible. Consequently, our proposed WGAN-GP model offers a potential innovative solution to address data scarcity challenge for researchers working with computational pulse diagnosis, which are expected to improve the performance of pulse diagnosis algorithms in the future.

show abstract

“…Here, we studied, for the first time, the dynamics of pulse-transit times, from heart to finger, in a similar way as previously done for inter-heartbeat intervals (Peng et al, 1993a;Bunde et al, 2000;Schumann et al, 2010), respiratory intervals (Kantelhardt et al, 2003;Schumann et al, 2010) and brain-wave amplitudes (Kantelhardt et al, 2015). In particular, the dynamics of the control of the pulse-wave propagation-e.g., blood pressure, arterial stiffness, etc.-can potentially be studied (Guo et al, 2022). Changes of the scaling behavior in some subjects can also be used as early indicators or diagnostic tools for pathologies that affect one or many of the studied organ systems (Ivanov et al, 1999a;Goldberger et al, 2002).…”

Section: Introductionmentioning

confidence: 99%

Long- and short-term fluctuations compared for several organ systems across sleep stages

Zschocke

Bartsch

Glos

et al. 2022

Front. Netw. Physiol.

View full text Add to dashboard Cite

Some details of cardiovascular and cardio-respiratory regulation and their changes during different sleep stages remain still unknown. In this paper we compared the fluctuations of heart rate, pulse rate, respiration frequency, and pulse transit times as well as EEG alpha-band power on time scales from 6 to 200 s during different sleep stages in order to better understand regulatory pathways. The five considered time series were derived from ECG, photoplethysmogram, nasal air flow, and central electrode EEG measurements from full-night polysomnography recordings of 246 subjects with suspected sleep disorders. We applied detrended fluctuation analysis, distinguishing between short-term (6–16 s) and long-term (50–200 s) correlations, i.e., scaling behavior characterized by the fluctuation exponents α1 and α2 related with parasympathetic and sympathetic control, respectively. While heart rate (and pulse rate) are characterized by sex and age-dependent short-term correlations, their long-term correlations exhibit the well-known sleep stage dependence: weak long-term correlations during non-REM sleep and pronounced long-term correlations during REM sleep and wakefulness. In contrast, pulse transit times, which are believed to be mainly affected by blood pressure and arterial stiffness, do not show differences between short-term and long-term exponents. This is in constrast to previous results for blood pressure time series, where α1 was much larger than α2, and therefore questions a very close relation between pulse transit times and blood pressure values. Nevertheless, very similar sleep-stage dependent differences are observed for the long-term fluctuation exponent α2 in all considered signals including EEG alpha-band power. In conclusion, we found that the observed fluctuation exponents are very robust and hardly modified by body mass index, alcohol consumption, smoking, or sleep disorders. The long-term fluctuations of all observed systems seem to be modulated by patterns following sleep stages generated in the brain and thus regulated in a similar manner, while short-term regulations differ between the organ systems. Deviations from the reported dependence in any of the signals should be indicative of problems in the function of the particular organ system or its control mechanisms.

show abstract

Wrist pulse signal acquisition and analysis for disease diagnosis: A review

Cited by 32 publications

References 78 publications

NA-MEMD Is Actually What You Need for Computational Pulse Analysis

NA-MEMD Is Actually What You Need for Computational Pulse Analysis

Towards Generating Realistic Wrist Pulse Signals Using Enhanced One Dimensional Wasserstein GAN

Long- and short-term fluctuations compared for several organ systems across sleep stages

Contact Info

Product

Resources

About