Parametric Classification of Recoverable Radar-Assessed Respiratory Rate Data

A static algorithm-based method is described here to differentiate between recoverable sedentary respiratory rate data extraneous motion segments measured using Doppler radar. Extraneous motion such as locomotion and fidgeting can cause drastic changes in dc offset and SNR of the received signal. Such extraneous data may not be excluded and can lead to an erroneous assessment of the respiration rate. In some cases, however, moderate distinct extraneous motion does not completely occlude the measurement of respiratory torso motion, allowing for respiration rate recovery. This work focuses on the accurate classification of data which is suitable for respiration rate analysis in the presence of locomotion and small extraneous movements. The proposed algorithm has been demonstrated to be accurate for classifying data with recoverable respiratory rates for 2 subjects and 3 types of fidgets with 99.4% accuracy on average.