Integrating particle tracking with computational fluid dynamics to assess haemodynamic perturbation by coronary artery stents -Particle tracking - CHROMA stent - data 1.2b

2.3 Particle tracking

Flow patterns and particle accumulation were visualised by tracking the motion of suspended particles through the model. As per the cell migration studies, the protocol detailed in this section was carried out for a variety of coronary and flow diverter stents, as well as empty non-stented vessels as controls. However, 1.5 mm diameter models were not included in this study.

2.3.1 Set up

Vessels were connected to a fluid reservoir and peristaltic pump, as described in Section 2.2.5.1. Any models previously used with HUVEC were first cleaned of debris. Reservoirs were filled with 12 ml of room temperature 70% industrial methylated spirit (IMS). A suspension of 10 µm polystyrene particles (FluoSpheres, Fisher Scientific) with 3.6 x 106 particles/ml was agitated via vortex mixer (Lab Dancer S42, IKA). 40 µl of the suspension was then transferred to the reservoir at a 250:1 fluid to particle volume ratio. The pump was briefly set to maximum power, to circulate the fluid and distribute the particles around the system. The flow rate was then reduced to 6.5 ml/min for 2 mm diameter vessels, and 12.8 ml/min for 2.5 mm diameter vessels. At these rates, the Reynolds number was 43 and 68 respectively, matching that in blood flow of 1 Pa wall shear stress within the same model.

2.3.2 Image acquisition

Model vessels were placed onto the stage of an inverted microscope (AE2000, Motic), with a connected camera (Moticam 2300, Motic). A 4x objective lens was used to ensure the full diameter of the vessel was in view and a live image was displayed via Images Plus software (Motic). The vessel was oriented so that the leading edge of the stent was within the image, with flow moving horizontally from left to right. The bottommost stent struts were brought into focus, before light intensity, aperture diaphragm opening and camera exposure were adjusted until moving particles could be seen.

The camera software was used to record a 2 minute video. The vessel was then repositioned so that the field of view was moved downstream, with a small amount of overlap with the previous field. Another 2 minute video was recorded and the vessel repositioned again. This process was repeated until the downstream end of the stent was captured in a final field of view.

Once the procedure was complete, particle accumulation was recorded (see Section 2.3.5). Following this, the model was flushed, removed from the system and rinsed with acetone to dissolve and remove any remaining particles. The model was rotated by 90° to expose different stent geometry and reconnected to the pump with a fresh reservoir. The above process was then repeated.

2.3.3 Image processing

Video recordings were split into image sequences using Video to JPG Converter software (DVDVideoSoft). These images were, if required, reoriented to ensure that the direction of flow was exactly horizontal and that adjacent fields of view could be aligned. The software returned images at a frame rate of 20 frames per second (fps), and every second frame was selected. For each field of view two sets of 300 images, each representing roughly 30 seconds of flow, were selected for tracking.

One frame from each image sequence was used to isolate the struts by removing the PDMS wall, along with any attached particles or debris, providing a clear image of stent geometry onto which tracks could be overlaid.