Application Of A Prediction And Smoothing Algorithm To Non-invasive blood Pressure Measurement

This paper presents an approach to the problem of artifact sensitivity in non-invasive-blood-pressure (NIBP) measurement using the oscillometric technique. An algorithm based on the Kalman filter with fixed lag smoothing was employed to predict the next oscillometric pulse amplitude and cuff pressure and to generate a smooth 'curve" from which the patient's blood pressure is computed. The algorithm has been implemented in a transport vital signs monitor. Clinical testing has demonstrated that the accuracy of the blood pressure measurements are well within AAMI specifications. Use in patient environments, including ambulance and helicopter transport, has demonstrated that the system is very robust in the presence of patient motion.