Mixing properties of coronary infusion catheters assessed by in vitro experiments and computational fluid dynamics

Abstract

Aims: Continuous infusion thermodilution is an established technique for the assessment of absolute coronary blood flow and microvascular resistance due to its proven accuracy and reproducibility. However, for this technique to yield reliable measurements, direct and homogenous mixing of injected saline and blood is mandatory. This study aimed to assess and compare the mixing properties of two different microcatheters, namely the Rayflow® (with sideholes for infusion) and the Finecross® catheter (single end-hole for infusion), which are commonly used in the catheterization laboratory.

Methods and results: The study employed three different methods to evaluate the mixing properties of the catheters. Firstly, a qualitative assessment of mixing was performed using ink injections in an in vitro bench model of a coronary artery. Secondly, in analogy to the human catheterization laboratory, mixing properties over the length of the coronary artery were assessed semi-quantitatively by temperature measurements in the bench model. Lastly, a quantitative assessment was performed by 3D computational fluid dynamics, where the standard deviation and entropy ratio of the temperature over the cross-section in the coronary artery model were calculated for both catheters.

Conclusion: All three evaluation methods demonstrated that the Rayflow catheter's specific design leads to a more optimal, homogeneous mixture of blood and saline over both the cross-section and length of a coronary vessel, as compared with the standard end-hole catheter.

Keywords: Computational fluid dynamics; Continuous thermodilution; Coronary physiology; Microvascular resistance.

Graphical Abstract

Methods and main results are shown for in vitro measurements and 3D computational fluid dynamics.

Figure 1

Illustration of the bench model mimicking the coronary artery and physiology and schematic display of the setup (LifeTec Group, Eindhoven, The Netherlands).

Figure 2

Graphical overview of the CFD geometries for the Rayflow catheter (with the pressure/temperature wire inside the catheter) and the Finecross catheter (with the pressure/temperature wire alongside the catheter) where the upper enlarged frame shows a close-up of the Rayflow geometry and the lower enlarged frame embarks the Finecross geometry. The small dots indicate the positions of the outlets.

Figure 3

Cross-section sample points for the simulations of the Rayflow catheter (top) and Finecross catheter (bottom). Note that the sample points indicated in asterisks represent the location of the pressure/temperature (pt) wire and are excluded from the time-averaged standard deviation (SD_t) and the entropy ratio (E).

Figure 4

Still frames for both the Finecross (top) and the Rayflow (bottom) at 10 s after the start of the ink injection. The tip of the infusion catheter is indicated by the arrow in each panel, and the distances can be observed from the liner in the bottom. The FC₅ (top) shows incomplete mixing (‘fail’), whereas the RF₅ catheter (bottom) shows complete mixing and thus ‘pass’.

Figure 5

Example of relative deviations of the normalized temperatures Rayflow (indicated by RF1 and RF2) and Finecross (indicated by FC1 and FC2) infusion catheters. On the x-axis, the distance to the tip in cm is depicted from 6 cm distal to the tip (6 cm) to the tip itself (0 cm) and on the y-axis the relative deviation (in %) with respect to the mean temperature.

Figure 6

(A) Time-averaged standard deviation (SD_t) and entropy ratio (E) (B) at different distances from the outlet during infusion for a simulation of coronary blood flow of 150 mL/min and an infusion rate of 15 mL/min for both the Rayflow (being the bottom line in panel A and upper line in panel B) and Finecross (being the upper line in panel A and the lower line in panel B) infusion catheters.

Figure A1

Example of percentage of data points in predefined ranges.

Figure A2

Relative deviations of the normalized measurements of temperature pullbacks for each combination of coronary flow rate and infusion rate for both catheters (Rayflow catheter in green and Finecross catheter in purple). The blue bar represents an area of 20% deviation, which is considered clinically acceptable.