A Novel Breathing Rate Estimation Framework Using 3 Distinct Types of Signals

Breathing Rate is a vital biological characteristic that needs to be assessed in a variety of therapeutic scenarios. It is still routinely monitored in a manual fashion. Consequently, a unique methodology for estimating the BR from a PPG- Pho-toplethysmogram, ECG- electrocardiogram, or blood pressure (BP) signal is presented in this research. To extract respiration-orientated signals, the system employs DWT- Discrete Wavelet Transform as well as EMD- Empirical Mode Decomposition algorithms, utilizing both the frequency as well as time domain data. Our approach was able to obtain satisfactory efficiency even during severely distorted portions of the signals by integrating an SQI- Signal Quality Index as well as an RKF - Robust Kalman Filter. The resultant signals are merged via the fusion of the state vector, and the BR is then calculated. The methodology was evaluated on 2 general people-accessible healthcare datasets, namely, the BIDMC as well as the MIT-BIH Polysomnographic datasets. The MAPE- Mean Absolute Percentage Error was used to assess performance. The outcomes also showed that the system was resilient to noise which is almost OdB. As a result, this system may be useful for BR surveillance in raised noise environments.