Evaluating Quality of Photoplethymographic Signal on Wearable Forehead Pulse Oximeter With Supervised Classification Approaches

Abstract

Pulse oximeter is a common and important instrument in medical clinic, which uses the photoplethysmography to measure oxygen saturation ratio (SpO2). However, the photoplethysmographic (PPG) signal is easily corrupted by the motion artifact when SpO2 is measured in a dynamic scenario. Moreover, the probe of most pulse oximeters available in the market is finger-type clip, which is only suitable in the static scenario. This study developed a wearable forehead pulse oximeter which could be used in a dynamic scenario. Because PPG signals with high signal quality index (SQI) produce accurate SpO2 values, to improve the accuracy of SpO2, the quality of PPG signals were evaluated by two supervised classification approaches, support vector machine (SVM) and convolutional neural network (CNN). The SVM classified the SQI of PPG signal by twelve statistic features calculated from 7-second PPG segment. In the CNN approach, the PPG signals converted to an image was used as the input. The VGG-19 model, was then used to evaluate the SQI of PPG segments. Twenty subjects were recruited to perform the static and dynamic experiments. For the dynamic experiment, subjects ran on a treadmill with three different speeds, 3 km/hour, 5 km/hour and 7 km/hour. Experimental results indicated that the accuracies of SVM and CNN for SQI classifications were 89.9 ± 1.18 % (mean ± deviation %) and 88.7 ±1.54%, respectively. Then, the dynamic data were used to test the two classification models which were trained by the static data. The accuracies of SVM and CNN for SQI classifications were 89.5 ± 3.87% and 86.2 ±4.28%, respectively. The error ratios of SpO2 in the case of static condition before and after the SQI classification with the SVM were respectively 5.6 ± 6.6% and 1.9 ± 1.1%. The results suggested that the performance of SVM was better than CNN.

Influence of cuff pressures of automatic sphygmomanometers on pulse oximetry measurements

“…A low-pass filter was applied to process the PPG signal. The cutoff frequency was set to 30 Hz which allows to remove ambient light [30] and power line interference [31] . Small changes in signal were eliminated using the moving average function.…”

Estimation of Left Ventricular Ejection Fraction Using Cardiovascular Hemodynamic Parameters and Pulse Morphological Characteristics with Machine Learning Algorithms

“…A decision rule for the signal quality was designed to select the pulse waves with good quality. The PCA was used to extract ten hemodynamic parameters and seventeen morphological parameters from the high-quality pulses [ 24 ]. Six parameters of bodily information were included.…”

“…In the duration of PCA, the pulse wave is easily coupled with the artificial motion when the cuff pressure is held at about 55 mmHg. The pulse quality would affect the accuracy of physiological measurement [ 24 , 38 , 39 ]. Thus, we proposed a decision rule to evaluate the quality of each pulse wave in the duration of PCA.…”

“…Machine learning (ML) algorithms have been widely used in physiological measurements for estimating the physiological parameters, such as blood pressure [ 11 , 17 ], muscle mass [ 18 ], calories [ 19 , 20 ], glucose [ 21 ], stroke volume [ 22 ], classifying the signal qualities of electrocardiogram [ 23 ] and photoplethysmogram [ 24 , 25 ], detecting arrythmia [ 26 ] and risky activities in daily life [ 27 ]. When using an ML method to process the regression or classification problem, searching the major features and finding the appropriate ML algorithms will depend on the collected data [ 28 , 29 ].…”

Artifacts classification and apnea events detection in neck photoplethysmography signals

“…Following this approach, several machine learning algorithms have been proposed in the literature to discriminate artifacts from clean PPG. Examples of signal processing techniques used in these algorithms include: decision lists [39][40][41][42][43], decision trees [44,45], naïve Bayes classifiers [46], support vector machines (SVM) [36,[47][48][49][50], multi-layered perceptrons [51], personalized neural networks (NN) [52], and 1-D CNNs [53,54].…”