Evaluation of Objective Distortion Measures for Automatic Quality Assessment of Processed PPG Signals for Real-Time Health Monitoring Devices

Abstract

Real-time photoplethysmogram (PPG) denoising and data compression has become most essential requirements for accurately measuring vital parameters and efficient data transmission but that may introduce different kinds of waveform distortions due to the lossy processing techniques. Subjective quality assessment tests are the most reliable way to assess the quality, but they are time expensive and also cannot be incorporated with quality-driven compression mechanism. Thus, finding a best objective distortion measure is highly demanded for automatically evaluating quality of reconstructed PPG signal that must be subjectively meaningful and simple. In this paper, we present four types of objective distortion measures and evaluate their performance in terms of quality prediction accuracy, Pearson correlation coefficient and computational time. The performance evaluation is performed on different kinds of PPG waveform distortions introduced by the predictive coding, compressed sampling, discrete cosine transform and discrete wavelet transform. On the normal and abnormal PPG signals taken from five standard databases, evaluation results showed that different subjective quality evaluation groups (5-point, 3-point and 2-point rating scale) had different best objective distortion measures in terms of prediction accuracy and Pearson correlation coefficient. Moreover, selection of a best objective distortion measure depends upon type of PPG features that need to be preserved in the reconstructed signal.

A Study of Brain Function Characteristics of Service Members at High Risk for Accidents in the Military

“…Representative single-pulse values were obtained from the values of the pulse-to-pulse interval by calculating the average of each time point relative to the pulse onset. The second and third derivatives were used to obtain SDPPG and TDPPG from the single-pulse PPG signal of each subject [ 133 , 140 , 141 ].…”

Photoplethysmogram signal reconstruction through integrated compression sensing and basis function aware shallow learning

“…Figure 1. Illustration of the bio-sensing wireless device or PPG monitoring wearable equipment Increased complexity of relations is learned using deep learning methods for PPG input signals and its output that is compressed [10]. This leads to increased rates of compression as well as higher conservation of diagnosed data in the signal compressed.…”

A method for portable PPG signal dynamic acquisition and denoise based on CEEMDAN and wavelet transform

“…So far, it has been accepted the methods to eliminate noise from PPG signal of the locomotor state can be grouped as three kinds which are based on adaptive filtering [13,14], wavelet transform (WT) [15]- [17] and empirical mode decomposition (EMD) [18,19]. Adaptive filtering depends on extra reference acceleration signals simultaneously measured by accelerometers during collecting PPG signal, [13,14] which increases the cost of the test equipment.…”

“…Adaptive filtering depends on extra reference acceleration signals simultaneously measured by accelerometers during collecting PPG signal, [13,14] which increases the cost of the test equipment. To get rid of the dependence on reference signal WT based approaches were proposed by providing wavelet base to decompose the signal into different wavelet coefficients and selecting an appropriate threshold value and method to obtain purified PPG signal, [15]- [17] it can have a good denoise effect [18] but may introduce phase distortions in PPG data [19]. EMD based methods are similar to WT based ones, with the difference in the base functions which are obtained from the extreme envelope of PPG signal itself, not artificially selected in WT based methods [20].…”