Using imaging Photoplethysmography (iPPG) Signal for Blood Pressure Estimation

Goudarzi, Reza Heydari; Mousavi, Seyedeh Somayyeh; Charmi, Mostafa

doi:10.1109/mvip49855.2020.9116902

Cited by 23 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method reduces the calculation burden of the device, but weakens the overall SNR with large low signal area. The other one is to select the area with intensive blood vessels on the skin as the ROI [15], [19]. In this study, we investigated the SNR maps of the palms of five subjects of different genders and occupations as shown in Fig.5.…”

Section: Methodsmentioning

confidence: 99%

“…A main limitation with this type of rPPG measurement is the environmental noises, including the white noise from the camara, unbalanced illumination, and differences in the transparency of the subjects’ skin. Therefore, techniques such as ICA and PCA [15]–[17], [23], [24] have been frequently used to filter the original data. These algorithms can work properly in many scenarios, but their measurement accuracy will drop significantly when the illuminance changes drastically.…”

Section: Methodsmentioning

confidence: 99%

“…This method can reduce the calculation burden of the device, but it also can weak the overall SNR by bring many parts with low signal content into the measurement area. The other one is to select the position with more blood vessels on the skin as the ROI [17], [21]. But which place is the mostly efficiency parts for the heart rate detection have not been demonstrated.…”

Section: Searching Roimentioning

confidence: 99%

“…It uses a camera to collect the reflected light from the skin and analyzes it to obtain heart rate. Existing rPPG methods use Independent Component Analysis (ICA) [13], Principal Component Analysis (PCA) [12], Butterworth pass band filter [14], Random Forest Regression (RFR) [15], and Fast Fourier Transform (FFT) [11], [16] for heart rate measurement. These methods cannot always effectively detach the heart rate signals from the noises brought by the complicated environment and subjects.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Robust and Remote Measurement of Heart Rate Based on A Surveillance Camera

Yang

Lian

Zhang

et al. 2021

Preprint

View full text Add to dashboard Cite

Arrhythmia is a marked symptom of many cardiovascular diseases. The accurate and in time detection of heart rate can greatly reduce its harm to people. However, it is still a challenge to automatedly and remotely measure the heart rate in daily life, because the environment factor of the measurement changes variously, such as the changing light intensity, the movement of people, and the uncertain distance from sensor to people. In this study, we accurately measured the heart rate of people at the distance of 4.8 meters under different intensity of light just by using a surveillance camera. After a short color video (20 sec) of a person's hand was captured by this camera, a method based on Fast Fourier Transform Algorithm (FFT) is proposed to extract the blood volume pulse wave to calculate the heart rate. By comparing the real heart rate with the results measured by electrocardiography (ECG), the accuracy of heart rate measurement using the method proposed in this study is 98.65% within 4.0 meters, and the accuracy can reach 90% within 5.6 meters. Our experiments also demonstrated that this method can accurately obtain the heart rate even when the intensity of light is below 32 LUX ( office environment 300-500 LUX). The strong environmental suitability makes this method can be applied to many occasions, such as community clinic, old peoples' home, classroom, and other public space.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Searching Roimentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Robust and Remote Measurement of Heart Rate Based on A Surveillance Camera

Yang

Lian

Zhang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…It is also difficult to acquire the ECG. The second detection method is based on the morphological theory of PPG [ 4 , 6 , 7 ]. The most common way to obtain the PPG signal is to use a photoelectric sensor to detect changes in light transmitted or reflected by human skin vessels [ 8 ].…”

Section: Introductionmentioning

confidence: 99%

Using CNN and HHT to Predict Blood Pressure Level Based on Photoplethysmography and Its Derivatives

et al. 2021

View full text Add to dashboard Cite

According to the WTO, there were 1.13 billion hypertension patients worldwide in 2015. The WTO encouraged people to check the blood pressure regularly because a large amount of patients do not have any symptoms. However, traditional cuff measurement results are not enough to represent the patient′s blood pressure status over a period of time. Therefore, there is an urgent need for portable, easy to operate, continuous measurement, and low-cost blood pressure measuring devices. In this paper, we adopted the convolutional neural network (CNN), based on the Hilbert–Huang Transform (HHT) method, to predict blood pressure (BP) risk level using photoplethysmography (PPG). Considering that the PPG′s first and second derivative signals are related to atherosclerosis and vascular elasticity, we created a dataset called PPG+; the images of PPG+ carry information on PPG and its derivatives. We built three classification experiments by collecting 582 data records (the length of each record is 10 s) from the Medical Information Mart for Intensive Care (MIMIC) database: NT (normotension) vs. HT (hypertension), NT vs. PHT (prehypertension), and (NT + PHT) vs. HT; the F1 scores of the PPG + experiments using AlexNet were 98.90%, 85.80%, and 93.54%, respectively. We found that, first, the dataset established by the HHT method performed well in the BP grade prediction experiment. Second, because the Hilbert spectra of the PPG are simple and periodic, AlexNet, which has only 8 layers, got better results. More layers instead increased the cost and difficulty of training.

show abstract