Effect of hemolysis on oxygen and hematocrit measurements by near infrared reflectance spectroscopy.

A short review of the principles of near infrared reflectance spectroscopy (NIRS) in whole blood is followed by a discussion on the influence of hemolysis. The increase of free plasmahemoglobin (PHb) has a strong influence on the continuous measurement of hematocrit and oxygen saturation (O(2)S) by NIRS. In view of the relative stability of hematocrit values in vivo this effect may be used to detect a change of the hemolysis rate induced by blood pumps in case of malfunction. The aim of this study is, therefore, the assessment of the hemolysis rate within an in vitro mock loop comprising a rotary blood pump by means of NIRS at constant hematocrit levels compared to the photometric reference method. Reflected light is measured by an integrated optical sensor working at three wavelengths (660 nm, 730 nm, and 830 nm). The experimental results demonstrate that the increase of free hemoglobin in plasma due to mechanical pumping leads to a decrease of detected reflected light at all three wavelengths. Influencing parameters such as adhering proteins on the sensor surface and the blood flow rate are briefly discussed. Finally, the possibility of using NIRS sensors for detecting malfunctions of blood pumps in vitro and in vivo is discussed, together with the option of using these sensors for supervision of long-term implantable pumps.

Simulation of the BP-80 blood pump.

In this study, computational fluid dynamics (CFD) analysis was applied to investigate the flow within a commercially available Biopump, BP-80 (Medtronic, Minneapolis, MN, U.S.A.). The Biopump was selected because, for this purpose, a great number of experimental hemolysis data is available. The process of geometry representation and grid generation was focused on, due to its high impact on the numerical results. This process incorporated the use of three commercially available software packages for three-dimensional computer-aided design (3D-CAD), grid generation, and solving, respectively. For the purpose of validation, the head/flow characteristics of the pump were experimentally obtained and compared to the computed data. The results showed a rough agreement between CFD data and experimental data. Further investigations should cover detailed shear stress analyses and computation of other hemolysis-related quantities.

Computational fluid dynamics and experimental validation of a microaxial blood pump.

Intravascular application of microaxial blood pumps as heart assist devices requires a maximum in size reduction of the pump components. These limitations affect the design process in many ways and restrict the number of applicable experimental procedures, but a detailed knowledge of the hemodynamics of the pump is of great interest for efficiency enhancement and reduction of blood trauma and thrombus formation. Computational fluid dynamics (CFD) offers a convenient approach to this goal. In this study, the inlet, vane, and outlet regions of a microaxial blood pump used as an intraaortic left ventricular assist device are analyzed by CFD and 3-dimensional (3-D) particle tracking velocimetry (PTV). For this purpose, a mock loop is set up that facilitates 3-D flow visualization. Flow in the main part of this testing device is modeled and computed by means of CFD. Pump head/flow (HQ) characteristics, axial pressure distribution, and particle images are then compared with numerical flow data. Results show that the pump performance characteristics, as well as inlet and outlet swirl predicted by the CFD model, are quite accurate compared with measured data. Proper boundary condition definitions and spatial discretization topology requirements for satisfactory results are discussed.