A Novel Non-Invasive Estimation of Respiration Rate From Motion Corrupted Photoplethysmograph Signal Using Machine Learning Model

Respiratory ailments such as asthma, chronic obstructive pulmonary disease (COPD), pneumonia, and lung cancer are life-threatening. Respiration rate (RR) is a vital indicator of the wellness patient. Continuous monitoring RR can provide early indication thereby save lives. However, real-time continuous facility only available at intensive care unit (ICU) due to size cost equipment. Recent researches have proposed Photoplethysmogram (PPG) and/ Electrocardiogram (ECG) signals for estimation however, usage ECG limited unavailability it in wearable devices. Due advent smartwatches with built-in PPG sensors, now being considered RR. This paper describes novel approach using motion artifact correction machine learning (ML) models signal features. Feature selection algorithms were used reduce computational complexity chance overfitting. The best ML model feature algorithm combination fine-tuned optimize its performance hyperparameter optimization. Gaussian Process Regression (GPR) Fit process regression (Fitrgp) outperformed all other combinations exhibits root mean squared error (RMSE), absolute (MAE), two-standard deviation (2SD) 2.63, 1.97, 5.25 breaths per minute, respectively. Patients would be able track lower less inconvenience if extracted efficiently reliably from signal.