ccoolbaugh
commented
8 years ago Resolution
Removed assignment of end diastole time to a default value and created a new filter that removes a cardiac cycle if its duration differs by more than 30% from the mean heart rate.
Additional Modifications
Fixed a bug with loading the platform calibration settings. Removed references to mean blood flow calculation from the validation study.
Current Status
The current cardiac cycle windowing scheme uses physiological limits (i.e. heart rate between 40 and 220 bpm) to remove aberrant cycles in the pulse wave data. If a cardiac cycle is removed from the analysis, the time point associated with end diastole for the preceding heart beat is assigned according to the mean RR interval time for the selected data epoch. This feature of the cardiac cycle algorithm is indicated in lines 881-890 of the FloWaveUS.m file.
Problem
Assignment of a default time for the end of diastole can result in a large deviation in the cardiac cycle duration compared to the neighboring cardiac cycles. The subsequent calculation of blood flow is then noisy due to the substantial difference in integration from the start of systole to the end of diastole as shown in the spikes in the figure below.
Is use of physiological limits to filter the heart rate appropriate if the cardiac cycle duration does not dynamically change in the experiment?