Calculation of the mean pressure with less delay for real-time clinical monitoring

The mean of cardiovascular pressure signals is an important metric in patient monitoring applications for many types of diseases and injuries. It is typically calculated with a moving- average of 3-8s of the pulsatile signal. This method of calculating the mean introduces a delay of 1.5-4s. We demonstrate that an FIR filter with coefficients calculated with a least squares error (LSE) estimator can reduce this delay without a clinically significant impact on the accuracy of the displayed signal. Preliminary results with intracranial pressure signals show that the delay can be completely eliminated with a maximum root-mean-square error of approximately 1 mmHg. Reduction or elimination of this delay could permit patient monitors to display the mean in real time and permit clinicians to respond to acute events more rapidly.