Zhe Hu

Hi,
is there an official reference or documentation for this great work?

This model appears to be based on Mauro Ursino's paper "Interaction between carotid baroregulation and the pulsating heart: a mathematical model," published in 1998 by the American Physiological Society. I was interested in this model because I built my own according to the same paper and then discovered this one, which was supposed to be the exact same thing. In comparing the two, I found a few problems with this one.

1. The calculation of Vu (total unstressed volume, equation 10 in the paper) is a little off. This can be found in the analysis & control system of the model. This model includes Vulv and Vurv in its calculation of Vu, which makes sense on its own but is not appropriate where Vu is used, which explains why Ursino's calculation of Vu does not include those quantities. Vu is subtracted, along with many other volume quantities, from the total blood volume of the system in order to find Psv (equation 9). Among the other volume quantities are Vlv and Vrv, which are total volumes that already include their own unstressed volumes. Therefore, to include Vulv and Vurv in the calculation of Vu is to double-count them.

2. In the systemic circulation, this model subtracts Vusv from Vsv in order to calculate Psv correctly (equation 9). This is incorrect, as Vusv is included in Vu, which was already subtracted.

3. I have reason to suspect other problems in this model related to the initial conditions (e.g. many volumes start out negative). I will not detail these here because most of them seem to get worked out of the model as the simulation runs for a little while. Two that are worth mentioning, though, are the calculations of Vlv and Vrv in the left heart and right heart, respectively. At first I could see that subtracting Vulv or Vurv from the integral of the net flow into the ventricle would justify error #1 above. The problem with this, however, is that the real Vlv and Vrv quantities are needed in two of the heart pump functions (equations 18 and 29) which can be found in Control/Left Ventricle and Control/Right Ventricle of the model. This model is giving the heart pump functions [Vlv - Vulv] and [Vrv - Vurv] instead of just plain Vlv and Vrv. This leads to incorrect simulation results.

Mistakes aside, I commend the author of this model because it runs several times faster than my own model, which does the exact same thing.