Abstract: PO0305
A Simplified Fluid Dynamics Model of Ultrafiltration
Session Information
- Bioengineering
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Bioengineering
- 300 Bioengineering
Authors
- Cardimino, Christopher R., University of Massachusetts Amherst, Amherst, Massachusetts, United States
- Germain, Michael J., Renal and Transplant Associates of New England, Springfield, Massachusetts, United States
- Chait, Yossi, University of Massachusetts Amherst, Amherst, Massachusetts, United States
Background
We recently presented a novel approach for the design of personalized ultrafiltration rate (UFR) profiles during hemodialysis (HD) treatments. The success of this approach depends on an accurate parameter estimation of a simplified fluid volume dynamics during ultrafiltration.
Methods
We used a simplified model derived from a validated fluid volume model during HD comprising intravascular and interstitial pools, microvascular refilling/filtration, and lymphatic flow. Input data used for parameter estimation are UFR profile and hematocrit (HCT) from CLIC obtained during actual HD treatments. Estimation was based on initial 30-min segment of the data and the model was validated based on the subsequent 30-min response. Model time constant and steady-state gain were obtained for a single patient at 5 treatments over a 3-week period.
Results
Estimation/validation results (Figure 1) demonstrate reasonable accuracy of the simplified fluid dynamics model. Underlying model parameters of a single patient exhibit significant variability between similar days of treatment and between treatment days (Figure 2). Both HCT response to same UFR profile (steady-state gain) and response time (time constant) vary by as much as 100%.
Conclusion
Successful estimation of fluid volume model parameters during HD is feasible which supports the concept of online design of personalized UFR profiles, A non-negligible variability of a patient’s model parameters may complicate the design of personalized UFR profiles.
Figure 1
Figure 2