Full Hematocrit–Viscosity Curve Identification Using Three-Dataset Krieger–Dougherty Regression

Blood viscosity is strongly dependent on hematocrit, and the hematocrit–viscosity relationship is an important determinant of blood rheology under physiological and pathological conditions. However, obtaining a full hematocrit–viscosity curve requires multiple measurements over a wide hematocrit range. In this study, a simple method is proposed to reconstruct the full hematocrit–viscosity curve using only three-dataset Krieger–Dougherty (K–D) regression as $\mu ={{\mu }_0(1-{\displaystyle \frac{\varphi }{{\varphi }_m}})}^{-\alpha {\varphi }_m}$. Based on suspended blood, RBC-rich blood and RBC-depleted blood are prepared after centrifugation. The hematocrit of each type of blood is measured using a micro-hemocytometer. Simultaneously, the blood viscosity of each type of blood is measured using the coflowing streams method. The proposed method is evaluated sequentially using reference datasets and hematocrit–viscosity datasets of control blood. According to results, the full hematocrit–viscosity curve obtained from three selected datasets is in good agreement with the experimental data and yields a lower root-mean-square error than conventional methods using all datasets. The exponent of the K–D model is strongly influenced by the midpoint dataset, whereas μ₀ is mainly affected by the suspending medium (dextran solution). In contrast, GA-induced rigidified RBCs do not significantly affect μ₀ within a 0.15% concentration. In conclusion, the proposed method provides a simple, efficient, and reliable approach for estimating the full hematocrit–viscosity curve.