Primary closure

The first model primary closure model studied displayed the usual geometric features of primary closure, with very limited dilatation of the bulb. Antegrade flow was found at all but 4 of 20 frames of the cardiac cycle, with no discernable flow separation or recirculation, as shown in Figures 5.5, 5.6 and 5.7. During the deceleration of systole a region of flow separation was noted in the bulb on both flow visualisation and CFD predictions in Figure 5.8. The separation region was centred in a slight dilatation immediately distal to a slight stenosis likely to reflect the proximal limit of the endarterectomy.

A further primary closure model was subjected to detailed scrutiny to ensure more robust validation of CFD, as few features of flow disturbance had been identified with the first model. This second model, shown in Figures 5.9-5.12, had a slight stenosis, again around the area of the proximal limit of the endarterectomy, but also a dilatation immediately distal to this in the carotid bulb. Antegrade flow was seen in the CCA during the acceleration phase of systole. In the bulb the velocity decreases within the larger lumen; flow is in the vessel centre with no discernible flow seen opposite the flow divider on either side (Figure5.9). Flow is then clearly antegrade in the ICA. At the peak of systole this pattern is more defined with recirculation seen in the CCA bulb at the ECA aspect and flow separation, though not clearly recirculation, in the ICA aspect (Figure5.10). There is no suggestion of flow disturbance in these areas on the lateral image. The AP image incorporates views of the bulb dilatation not apparent on the lateral image.

During the deceleration of systole the same pattern is seen with recirculation opposite to the flow divider (Figure 5.11). The antegrade flow at the divider appears to have been

directed around the region of flow separation, depicted clearly by both visualisation and CFD. The lateral view has an area of slightly shorter streak lines, implying lower velocity in the centre of the bulb, which reflects the geometry of the bulb dilation, seen on the AP views. During diastole low central antegrade flow was identified with recirculation clearly in both ECA and ICA aspects of the bulb (Figure 5.12). Features of flow disturbance are described in Table 5.1.

Several areas of flow disturbance were visualised and predicted by CFD in the second primary closure model. These areas were in the bulb immediately proximal to the flow divider, just distal to the expected proximal limit of the endarterectomy. Flow predominated close to the flow divider, which was also accurately predicted by CFD; validating the CFD technique. In the model with little carotid bulb dilatation a single feature of disturbed flow was identified which was corroborated by flow separation by flow visualisation.

5.3.3 5mm patch angioplasty

In the 5mm patch angioplasty model studied, with pulsatile flow at 300ml/minute, antegrade flow was identified throughout the carotid bulb during acceleration and peak of systole in both flow visualisation and CFD images as seen in Figures 5.13 and 5.14 on both AP and lateral views. During the deceleration of systole antegrade flow is seen on the lateral view and within the CCA, centre of the bulb and ICA on the AP view, shown in Figure 5.15. At both the ECA and ICA margins of the bulb flow separation was identified on both flow visualisation and CFD, with visualisation confirming recirculation. Diastolic flow was antegrade within the CCA, ICA and close to the flow divider within the bulb, as seen in Figure 5.16. Flow separation was found at the lateral margins of the ICA bulb, opposite to the flow divider. On both flow modalities flow appeared to have been diverted around these areas of separation into the central portion of the vessel in both planes.

Flow visualisation images correlate well with CFD velocity vector data at all stages of the cardiac cycle and thus, the CFD model has been validated, at least at a qualitative value. Images for the other five 5mm patch angioplasties, which had flow visualisation performed, produced similar flow patterns at the same stage of the cardiac cycle, with the exception of one case in which a sudden dilatation was noted at the proximal limit of the patch, as seen in Figure 5.17. The geometry of this model was orientated in such a way that the dilatation was on the ECA aspect of the bulb only; this was at least in part, due to angulation within the CCA. This geometry resulted in a large region of recirculation proximal to the flow divider at the ECA aspect of the carotid bulb.

Figure 5.17 AP view of 5mm patch angioplasty during deceleration of systole

5.3.4 8mm patch angioplasty

In the larger, 8mm patch angioplasty, during acceleration of systole, flow velocity, seen as shorter streaklines and velocity vectors, decreased from the CCA into the CCA bulb, however, no flow separation was observed in either flow visualisation or CFD velocity vector data. This pattern progressed through the bulb, with very low velocity in the bulb region. Velocities were so low at the anterior margin of the bulb that no discernible flow could be seen. In the ICA, beyond the bulb higher flow velocities were seen in Figure 5.18. At the peak of systole, antegrade, unidirectional flow was visualised throughout the CCA, the centre and posterior aspects of the bulb and ICA, seen in Figure 5.19. Flow separation with a possible area of recirculation was seen on visualisation with either no flow or weak separation on CFD. During the deceleration of systole antegrade flow was

seen only within the CCA, the centre of the bulb and the ICA. Posteriorly, anteriorly and at both lateral margins clear flow separation was noted; identified most clearly at the convexities of the model posteriorly and at the lateral margin of the ICA bulb (Figure 5.20). Diastolic flow continued to show these features, though the velocity magnitude appeared lower (Figure 5.21). The agreement between flow visualisation and CFD velocity vector data indicated flow separation in exactly the same regions of the model and stage of the cardiac cycle contribute to the validation of the CFD technique.

In the other 8mm patch models, the patterns seen were very similar to the model described above; again with confirmation from both modalities. Figure 5.22 features a further model, which included a dilatation well within the CCA, presumably at the proximal limit of the patch. This region has clear flow separation with recirculation through all stages of the cardiac cycle (Figure 5.22).

Figure 5.22 AP view of a further 8mm patch angioplasty during deceleration of systole