Bolus delay and dispersion

Other sources of artifact in deconvolution-based PWI maps include bolus delay and dispersion [80-82]. When a bolus of injected contrast reaches brain tissue via stenotic or collateral arteries, its arrival is delayed. Furthermore, when a multiplicity of pathways is taken, as is especially true in the setting of collateral circulation, there is dispersion of the contrast bolus. The SVD deconvolution algorithm, which is used at the majority of centers where PWI is performed, artifactually underestimates CBF in the setting of bolus delay and/or dispersion (Fig. 8.12). 

The major challenge in using CTP for the assessment of the core infarct is the questionable reliability of dynamic perfusion imaging. CBV measurements are thought to be more robust than CBF and MTT measurements because the calculation of cerebral blood volume is relatively insensitive to bolus delay and dispersion [85, 86]. However, as a calculated parameter, it remains dependent on postprocessing (e.g., partial volume effects) and patient-specific issues (e.g., hematocrit levels) [87-89]. Additionally, as demonstrated in the DIAS-2 study [90], there is poor interrater agreement in assessing the CBV lesion. 

Indirect or summary parameters, such as bolus arrival time, time-to-peak (TTP), and full-width at half maximum concentration, are easy to calculate, but are thought to be less accurate than parameters derived from deconvolution with an arterial input function (AIF). Deconvolution corrects for bolus delay and dispersion up to the level of the AIF [114], However, recent studies suggest that both approaches provide quahtatively similar results [115-117] and may be vaUd for clinical purposes [112,118,119],... 

---