Evaluation of Computational Models for Hemolysis Estimation

Abstract

The potential for mechanical erythrocyte damage, or hemolysis, in heart valves and blood pumps has been estimated using computational fluid dynamics (CFD) analysis, combined with mathematical models of red cell damage mechanisms. To date, these prediction approaches have not been compared with each other in common benchmark cases, nor have they been evaluated in radically different flow geometries. In this study, four test devices, including a hemoresistometer, a spinning disk, a capillary tube, and a concentric cylinder viscometer were hemolysis tested and computationally simulated. A number of existing models were used to predict blood damage and these models were compared with each other and with actual measurement of hemolysis. The study indicates that the effectiveness of blood damage prediction from the existing models is similar and can potentially be improved with the consideration of a proposed repeated flow passage effect. It also indicates that the improved models can be used to more effectively predict blood damage in widely different situations.

Significance of Extensional Stresses to Red Blood Cell Lysis in a Shearing Flow

“…Some investigators have recognized the potential importance of the history of mechanical loading in blood trauma and its combination with resistance of cells to accumulated damage. 8,46 As such, the inability of current power-law models to account for the past shear history of a cell adds to the inaccuracy of their predictions, 20,22 where, for example, sub-hemolytic trauma to cells can leave them more susceptible to rupture when subjected to shear yet again. Antaki 1 also discussed the utility of power-law models and concluded that they are unable to capture the effects of subtle features of the blood contacting geometry, which have been shown to largely affect hemolysis.…”

“…Results from application of correlations to predict hemolysis quantitatively have been disappointing, and for various reasons some researchers have proposed amendments of the power-law model or the use of entirely different models. 4,5,10,18,21,22,26,28,37 One of the problems with these empirical models relates to the cell's exposure. Differences in the nature of shear stress exposure in controlled laboratory conditions compared to the reality of clinical settings contribute to inaccuracies of the power-law model of hemolysis estimation.…”

“…Antaki 1 also discussed the utility of power-law models and concluded that they are unable to capture the effects of subtle features of the blood contacting geometry, which have been shown to largely affect hemolysis. Previous researchers have attempted to use current blood damage power-law models to estimate hemolysis from CFD results, often using a Lagrangian approach, 22,33,45 where damage was calculated along multiple streamlines in the domain of interest. These conventional models provide poor estimates of hemolysis when compared to experimental hemolysis data in the same flow system, thus these investigators have illustrated the inaccuracy of current models for hemolysis prediction.…”

Estimation of Viscous Dissipative Stresses Induced by a Mechanical Heart Valve Using PIV Data

“…They then integrated shear stress and exposure time along the stream line to obtain the Index of Hemolysis as an index of improvement for artificial organs. 17 In turbulent flow, RBCs would be damaged only when the scale of smallest vortices (Kolmogorov scale) was similar to the scale of RBCs. However, turbulent Reynolds stress calculated in the flow field belongs to macro scale, and this Reynolds stress should not be considered the real stress causing hemolysis.…”

Modelagem matemática da fluidodinâmica não-newtoniana e bifásica simplificada da hemólise induzida mecanicamente em sistemas de bombeamento centrífugo de sangue

“…Alguns aprimoramentos baseados nos modelos de lei de potências foram propostos. Uma das vertentes destes aprimoramentos consiste em acompanhar hemácias ao longo de suas trajetórias no interior de dispositivos artificiais, calculando-se a hemólise induzida mecanicamente acumulada ao longo desse caminho, através de uma integração numérica linear , (Bludszuweit, 1995a), (Bludszuweit, 1995b), (Chan et al, 2002), (De Wachter e Verdonck, 2002), , , , (Song et al, 2003b), (Goubergrits e Affeld, 2004), (Arvand et al, 2005), (Gu e Smith, 2005), (Grigioni et al, 2005), (Goubergrits, 2006), (Throckmorton e Untaroiu, 2008), (Arwatz e Smits, 2013).…”