Estimation and Validation of Arterial Blood Pressure Using Photoplethysmogram Morphology Features in Conjunction With Pulse Arrival Time in Large Open Databases

Yang, Seung Man; Sohn, Jangjay; Lee, Saram; Lee, Joonnyong; Kim, Hee Chan

doi:10.1109/jbhi.2020.3009658

Cited by 37 publications

(17 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While some authors reported that they accounted for such a subject-dependent split (e.g., [ 6 , 8 , 10 , 12 , 20 ]) others did not and our results suggest that in this case a particularly high prediction performance might partly result from potential train/test split violations. However, it should also be noted that such a careful split is not always possible when using publicly available databases like the MIMIC-II subset where subject specific information is often not provided.…”

Section: Discussioncontrasting

confidence: 74%

“…Yang et al constructed time series from morphological PPG features and PTT values. They divided the time series into high- and low-frequency components and fed them into a DNN and RNN for BP estimation [ 8 ].…”

Section: Related Workmentioning

confidence: 99%

“…In recent years research mainly focused on BP estimation from cuffless multi and single sensor solutions. The former often utilize time or phase differences between different signals (usually ECG and PPG or multiple PPG) related to the blood volume propagation through arteries [ 5 , 6 , 7 , 8 , 9 ]. The latter mainly exploit morphological properties of blood volume dynamics derived from PPG measurements on a particular site [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 ].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Assessment of Non-Invasive Blood Pressure Prediction from PPG and rPPG Signals Using Deep Learning

Schrumpf

Frenzel

Aust

et al. 2021

Sensors

View full text Add to dashboard Cite

Exploiting photoplethysmography signals (PPG) for non-invasive blood pressure (BP) measurement is interesting for various reasons. First, PPG can easily be measured using fingerclip sensors. Second, camera based approaches allow to derive remote PPG (rPPG) signals similar to PPG and therefore provide the opportunity for non-invasive measurements of BP. Various methods relying on machine learning techniques have recently been published. Performances are often reported as the mean average error (MAE) on the data which is problematic. This work aims to analyze the PPG- and rPPG based BP prediction error with respect to the underlying data distribution. First, we train established neural network (NN) architectures and derive an appropriate parameterization of input segments drawn from continuous PPG signals. Second, we use this parameterization to train NNs with a larger PPG dataset and carry out a systematic evaluation of the predicted blood pressure. The analysis revealed a strong systematic increase of the prediction error towards less frequent BP values across NN architectures. Moreover, we tested different train/test set split configurations which underpin the importance of a careful subject-aware dataset assignment to prevent overly optimistic results. Third, we use transfer learning to train the NNs for rPPG based BP prediction. The resulting performances are similar to the PPG-only case. Finally, we apply different personalization techniques and retrain our NNs with subject-specific data for both the PPG-only and rPPG case. Whilst the particular technique is less important, personalization reduces the prediction errors significantly.

show abstract

Section: Discussioncontrasting

confidence: 74%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Assessment of Non-Invasive Blood Pressure Prediction from PPG and rPPG Signals Using Deep Learning

Schrumpf

Frenzel

Aust

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…The three methods mentioned will be explained in more detail in the following subsection. Still, it is relevant to mention that many studies consider PAT as a replacement for PTT [46,47,48,49,50,51,52], but these are not equal, but there is a relationship by the pre-ejection period (PEP) [29], which varies independently per person, hence the discrepancies in the literature about whether PTT is better than PAT or vice versa as an estimator of blood pressure, [53] because the results are still very varied in terms of accuracy [32,46,54,55,56,57].…”

Section: Adaptive Filtersmentioning

confidence: 99%

Continuous Blood Pressure Estimation in Wearable Devices Using Photoplethysmography: A Review

Medina¹,

Lopez²,

Galdos³

et al. 2022

IJETAE

View full text Add to dashboard Cite

Cardiovascular diseases (CVD) are among those with the highest mortality rates, and various wearable devices for continuous monitoring are emerging as a complement to medical procedures. Blood pressure (BP) monitoring in wearable devices, in order to be continuous, must be performed noninvasively, thus involving photoplethysmography (PPG), a technology that has been widely studied in recent years as a non-invasive solution for BP estimation. However, continuous data acquisition in a wearable system is still a challenge, one of the reasons being the noise caused by movement, the correct use of the PPG signal, and the estimation method to be used. This paper reviews the advances in blood pressure estimation based on photoplethysmography, focusing on the analysis of the preprocessing (ICA, FIR, adaptive filters) of the signals. Among the filters reviewed, the most suitable for dealing with Motion Artifacts (MA) of a wearable system are the adaptive filters, because conventional filters are limited to work only in the band for which they are designed, which does not always cover the spectrum of the MA. A review of the estimation methods is also carried out, among them machine learning stands out because it shows greater growth due to the new proposals that use more signals and obtain better results in terms of accuracy. The objective is to know and analyze the appropriate preprocessing filters and estimation methods from the perspective of wearable systems using PPG sensors affected by AM. Keywords— Blood Pressure Estimation, PAT, PTT, Machine Learning, Photoplethysmography, adaptive filtering.

show abstract

“…In recent years research mainly focused on BP estimation from cuff-less multi and single sensor solutions. The former often utilize time or phase differences between different signals (usually ECG and PPG or multiple PPG) related to the blood volume propagation through arteries [5,6,7,8,9]. The latter mainly exploit morphological properties of blood volume dynamics derived from PPG measurements on a particular site [10,11,12,13,14,15,16,17,18,19,20].…”

Section: Introductionmentioning

confidence: 99%

Assessment of deep learning based blood pressure prediction from PPG and rPPG signals

Schrumpf¹,

Frenzel²,

Aust³

et al. 2021

Preprint

View full text Add to dashboard Cite

Exploiting photoplethysmography signals (PPG) for non-invasive blood pressure (BP) measurement is interesting for various reasons. First, PPG can easily be measured using fingerclip sensors. Second, camera-based approaches allow to derive remote PPG (rPPG) signals similar to PPG and therefore provide the opportunity for non-invasive measurements of BP. Various methods relying on machine learning techniques have recently been published. Performances are often reported as the mean average error (MAE) on the data which is problematic. This work aims to analyze the PPG-and rPPG-based BP prediction error with respect to the underlying data distribution. First, we train established neural network (NN) architectures and derive an appropriate parameterization of input segments drawn from continuous PPG signals. Second, we apply this parameterization to a larger PPG dataset and train NNs to predict BP. The resulting prediction errors increase towards less frequent BP values. Third, we use transfer learning to train the NNs for rPPG based BP prediction. The resulting performances are similar to the PPG-only case. Finally, we apply a personalization technique and retrain our NNs with subject-specific data. This slightly reduces the prediction errors.

show abstract

Estimation and Validation of Arterial Blood Pressure Using Photoplethysmogram Morphology Features in Conjunction With Pulse Arrival Time in Large Open Databases

Cited by 37 publications

References 62 publications

Assessment of Non-Invasive Blood Pressure Prediction from PPG and rPPG Signals Using Deep Learning

Assessment of Non-Invasive Blood Pressure Prediction from PPG and rPPG Signals Using Deep Learning

Continuous Blood Pressure Estimation in Wearable Devices Using Photoplethysmography: A Review

Assessment of deep learning based blood pressure prediction from PPG and rPPG signals

Contact Info

Product

Resources

About