Page 96

proefschrift gommer inhoud met kaft.indd

Chapter 5 Limitations With TCD only blood flow velocity can be measured and no absolute flow due to unknown blood vessel diameter. Because NVC responses are converted into relative responses this is no issue, as long as the vessel diameter remains constant during the evoked responses. PCA diameter is approximately 2.5 mm. Huberet al 11 showed via angiography that cerebral arteries with a diameter of at least 2.5 mm do not show changes in diameter with hypercapnia, hyperventilation, hyper- tonic glucose and administration of papaverine. We assumed also visually evoked flow enhancement does not influence PCA diameter. The combined RGCA model used, which is an addition of an NVC model and a CA model, might not be optimal. Both sub-systems, i.e. the NVC and CA system, are modeled with a second order model. For each system individually this was shown to be adequate 20, 29. However, if both systems work (partly) through the same mechanisms, which could well be the case, the combined system might be over dimensioned and less parameters might be necessary to describe the evoked response with the same precision. Currently the combined RGCA model consists of eight parameters. This is the same amount as the ARMA model of Paneraiet al 17, but as already mentioned, they also incorporated the influence of CO2 from a third input. Improvements might be achieved by stepwise down- sizing the model and evaluating if parameter reliability improves with a more simpler model. However, what is clear from our results is that parameter reliabil- ity improved in the RGCA model compared to the RG model. Our data consists of a heterogeneous set from different patient groups and healthy elderly. Our aim was to study the influence of blood pressure related changes in CBFV in relation to quantification of NVC, which is a purely meth- odological approach. Since all data analysis is linear, all data should equally be effected by these methodological differences. Conclusion Blood pressure correction of NVC responses by including cerebral autoregula- tion in model fitting of averaged VEFR responses results in significantly lower fit errors. Blood pressure correction is more effective when mean instead of systolic CBFV responses are used. A substantial amount of NVC parameters change after blood pressure correction resulting in more narrow confidence bounds and due to that into more reliably estimates. Measurement and quantification of NVC should include beat-to-beat blood pressure measurement. 94


proefschrift gommer inhoud met kaft.indd
To see the actual publication please follow the link above