Sensitivity analysis study on the effect of the fluid mechanics assumptions for the computation of electrical conductivity of flowing human blood

• Published in: Reliability engineering & system safety Vol. 213; p. 107663
• Main Authors: Melito, Gian Marco, Müller, Thomas Stephan, Badeli, Vahid, Ellermann, Katrin, Brenn, Günter, Reinbacher-Köstinger, Alice
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.09.2021; Elsevier BV
• Subjects: Blood conductivity; Cardiac output; Computational fluid dynamics; Electrical conductivity; Electrical resistivity; Erythrocytes; Fluid flow; Fluid mechanics; Heart rate; Mathematical models; Model accuracy; Parameters; Polynomial chaos expansion; Reliability engineering; Rheological properties; Sensitivity analysis; Stroke volume; Fluid mechanics; Sensitivity analysis; Polynomial chaos expansion; Blood conductivity
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2021.107663

Impedance cardiography is a non-invasive methodology for measuring cardiodynamic parameters, such as stroke volume and heart rate, as well as cardiac output. For the measurement, the electric conductivity of blood is important. The conductivity of blood depends on various parameters, such as the haematocrit value as well as the red blood cells’ (RBC) shape and orientation. In models, the response is usually affected by uncertainty, which may lead to inaccurate medical diagnosis. Therefore, a ranking of the influence of the model’s input factors may be necessary. Also, physically and physiologically correct assumptions are fundamental for the accuracy of the model. The basis for predicting the conductivity of blood in this study is the Maxwell–Fricke theory, which allows computing the electrical bulk conductivity of quiescent blood. For flowing blood, fluid mechanics has to be coupled in the modelling phase. Nevertheless, some assumptions may lead to invalid or inaccurate results. Based on a global sensitivity analysis, this work shows which fluid mechanical assumptions are incorrect and should be avoided. Moreover, positive effects based on accurate rheological modelling of the fluid properties are shown, and the factors with a decisive influence on the computed conductivity change of flowing blood are illustrated. •Electrical conductivity of human blood is a cornerstone for impedance cardiography.•Trigger variable selects model assumptions in polynomial chaos expansion metamodel.•Sensitivity analysis between variables and among model assumptions is performed.•Shear-thinning generalized model better represents the rheology of blood.•Blood conductivity is sensitive to flow rate ratio and red blood cells aspect ratio.