Optical blood pressure estimation with photoplethysmography and FFT-based neural networks

Abstract

We introduce and validate a beat-to-beat optical blood pressure (BP) estimation paradigm using only photoplethysmogram (PPG) signal from finger tips. The scheme determines subject-specific contribution to PPG signal and removes most of its influence by proper normalization. Key features such as amplitudes and phases of cardiac components were extracted by a fast Fourier transform and were used to train an artificial neural network, which was then used to estimate BP from PPG. Validation was done on 69 patients from the MIMIC II database plus 23 volunteers. All estimations showed a good correlation with the reference values. This method is fast and robust, and can potentially be used to perform pulse wave analysis in addition to BP estimation.

Continuous Blood Pressure Measurement From Invasive to Unobtrusive: Celebration of 200th Birth Anniversary of Carl Ludwig

“…With the ever-increasing computational power and development of big data technology, big data analysis using the machine learning method for cuffless BP measurement has gained increasing attention [99][100][101][102][103]. The general idea with this technique is to initially extract surrogate cardiovascular indexes from physiological signals, then use machine learning to train this data to adapt to the system, and finally predict BP using the trained model.…”

Método Não Invasivo Para Predição Da Pressão Arterial Com Pletismografia E Inteligência Computacional

“…Do ponto de vista do tratamento dos dados e medição da pressão sistólica e diastólica por meio de PPG, o trabalho desenvolvido por Xing e Sun, utiliza transformada rápida de Fourrier (FFT) para modelagem dos dados (Xing and Sun, 2016). O trabalho elabora um modelo matemático que associa o sinal PPG ao volume da artéria na região compreendida pelo sensor.…”

Non-Invasive Continuous Blood-Pressure Monitoring Models Based on Photoplethysmography and Electrocardiography

“…The structures of the regression models were similar to the PWTT-based models, while the features changed from univariate to multivariate and the accuracy improved. With the development of machine learning, the characteristic parameters of models were further enriched, including the amplitude, phase characteristics of pulse waves extracted with fast Fourier transform [23], spectral characteristics [24], and the features of the photoplethysmography (PPG) waveform and related first and second (time) derivatives [25,26]. Moreover, the model construction methods were expanded, such as neural network [24,[27][28][29], support vector machine [30], adaptive boosting regression [31], and random forest algorithm [32].…”