ABSTRACT
In a previous paper, a new dielectric technique was used to estimate hematocrit (HTC) in extracorporeal blood circulation systems, independently of plasma conductivity or osmolarity. Although many impedance techniques have been formerly proposed in the literature, none has been evaluated against plasma conductivity and osmolarity. Herein, we estimate HTC based on permittivity changes and also with other four techniques found in the literature. Besides, the error incurred in each is also studied when plasma conductivity and osmolarity changed as much as 1 mS/cm and 50 mOsm/kg, respectively. The dielectric (permittivity) technique has an error close to 5.4%, while the others showed both tendencies, i.e., lower error (2.5%, two of them) and higher error (8.6% and 16.3%, the other two). The dielectric technique, even though did not produce the lowest error, provides a well-described physical model along with simple instrumentation.
Subject(s)
Algorithms , Diagnosis, Computer-Assisted/methods , Electric Impedance , Hematocrit/methods , Humans , Reproducibility of Results , Sensitivity and SpecificityABSTRACT
Based on permittivity changes, a new method to measure hematocrit (HCT) in extracorporeal blood systems is presented. Human blood samples were tested at different HCT levels pairing the values of permittivity change, obtained by means of a commercial impedance analyzer, with traditional centrifugation measurements. Data were correlated using both linear and nonlinear regression. When using the lineal model, the comparison yielded a high correlation coefficient (r = 0.99). Theoretical simplifications suggest that the method is independent of changes in the conductivities of the intracellular and extracellular compartments. The influence of osmolarity and conductivity of the extracellular compartment was analyzed. It is shown that HCT can be predicted within an error lower than 5% when those parameters changed as much as 1 mS/cm and 50 mOsm/kg, respectively. Thus, the method appears as valid and viable showing good possibilities in applications such as renal dialysis.
