Abstract:Patients with surgical, pulmonary, and cardiac problems, continual monitoring of Oxygen Saturation of a Person (SpO2) and Respiratory Rate (RR) is essential. Similarly, the persons with cardiopulmonary health issues, RR estimation is crucial. The performance of the ventilator assistance and lung medicines are evaluated using SpO2 and RR. For the persons, those who are living alone with respiratory illnesses need a compulsory estimation of RR. In case of serious illness, the RR might face abrupt changes. The immobility of the disturbance and RR makes the RR evaluation from the PhotoPlethysmoGraphic (PPG) signals is a difficult challenge. So, an efficient RR and SpO2 estimation framework from the PPG signal using the deep learning method is developed in this paper. At first, the PPG signal is collected from standard data sources. The collected PPG signals undergo signal pre-processing. The pre-processing procedures include Motion Artifacts (MA) removal and filtering techniques. The pre-processed signals are split into distinct windows. From the split windows of the signals, the spectral features, RR, and Respiratory Peak Variance (RPV) features are extracted. The retrieved features are selected optimally with the help of Advanced Golden Tortoise Beetle Optimizer (AGTBO). The weights are chosen optimally with the same AGTBO. The optimally selected features are fused with the optimal features to get the weighted optimal features. These weighted optimal features are fed into the Ensemble Learning-based RR and SpO2 Estimation Network (ELRR-SpO2EN). The ensemble learning model is developed by combining Multilayer Perceptron (MLP), AdaBoost, and Attention-based Long Short Term Memory (A-LSTM). The performance of the developed RR and SpO2 estimation model is compared with other existing techniques. The experimental analysis results revealed that the proposed AGTBO-ELRR-SpO2EN model attained 96 % accuracy for the second dataset, which is higher than the conventional models such as MLP (90 %), Adaboost (92 %), A-LSTM (92 %), and MLP-ADA-ALSTM (94 %). Thus, it has been confirmed that the designed RR and SpO2 estimation framework from PPG signals is more efficient than the other conventional models.

i-PRExT: Photoplethysmography Derived Respiration Signal Extraction and Respiratory Rate Tracking Using Neural Networks

RespNet: A deep learning model for extraction of respiration from photoplethysmogram

PPG2RespNet: a deep learning model for respirational signal synthesis and monitoring from photoplethysmography (PPG) signal

An efficient model for extracting respiratory and blood oxygen saturation data from photoplethysmogram signals by removing motion artifacts using heuristic-aided ensemble learning model

Rapid Extraction of Respiratory Waveforms from Photoplethysmography: A Deep Encoder Approach

Energy-Efficient PPG-Based Respiratory Rate Estimation Using Spiking Neural Networks

Lightweight End-to-End Deep Learning Solution for Estimating the Respiration Rate from Photoplethysmogram Signal

Multitask Network for Respiration Rate Estimation -- A Practical Perspective

Deep Recurrent Neural Network for Extracting Pulse Rate Variability from Photoplethysmography During Strenuous Physical Exercise

Improving Photoplethysmographic Measurements Under Motion Artifacts Using Artificial Neural Network for Personal Healthcare

Method of Remote Photoplethysmography Robust to Interference in Video Registration of Human Facial Skin

A Novel Respiratory Rate Estimation Algorithm from Photoplethysmogram Using Deep Learning Model

Photoplethysmography-Based Respiratory Rate Estimation Algorithm for Health Monitoring Applications

A Novel Non-Invasive Estimation of Respiration Rate from Photoplethysmograph Signal Using Machine Learning Model

Continuous Patient-Independent Estimation of Respiratory Rate and Blood Pressure Using Robust Spectro-Temporal Features Derived From Photoplethysmogram Only

An Evaluation of Non-Contact Photoplethysmography-Based Methods for Remote Respiratory Rate Estimation

CorNET: Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment

Heart Rate Estimation From Facial Videos Using a Spatiotemporal Representation With Convolutional Neural Networks

A Feasibility Study of a Video-Based Heart Rate Estimation Method with Convolutional Neural Networks

Neural signals-based respiratory motion tracking: a proof-of-concept study

On-Device Reliability Assessment and Prediction of Missing Photoplethysmographic Data Using Deep Neural Networks