An Extended Kalman Filter for Retained Volume Estimation in Peritoneal Perfusion Applications

This paper develops and validates an extended Kalman filter for estimating the fluid volume retained in a patient's or a lab animal's peritoneal cavity during perfusion. Such estimation is potentially valuable for monitoring perfusion effectiveness and patient condition during medical interventions such as peritoneal dialysis. Monitoring intra-abdominal pressure and volume is particularly important for preventing undesirable or unsafe conditions such as intra-abdominal hypertension, abdominal compartment syndrome, etc. The literature already models the dynamics of peritoneal cavity pressure during perfusion. However, to the best of the authors' knowledge, the use of such models for online perfused volume estimation is a novel contribution of this paper. Specifically, the paper presents a simple model of cavity pressure dynamics, parameterizes it from a perfusion experiment on a large laboratory animal (namely, a Yorkshire swine), and uses it to design an extended Kalman filter for online retained volume estimation. The paper validates its approach by comparing the perfused volume estimate to a benchmark estimate generated using canister fluid level sensors embedded in the perfusion setup. The results of this validation are very encouraging, with 97% correlation between the two volume estimates.