First-pass contrast-enhanced dynamic perfusion imaging (DSC MRI)

Principles

Gadolinium chelates are used in perfusion MRI for their magnetic susceptibility effect at high concentration: the heterogeneities of the magnetic field created by the presence of the contrast agent in the vessels lead to a decrease in the relaxation times T2 and T2* of the surrounding tissue.

The reduction in the signal recorded during the first pass of the contrast agent will depend on its concentration in the vessel, the number and diameter of vessels per volume unit, and the type of signal weighting. T2-weighted sequences (SE-EPI) are more specific to the microvascular compartment than T2*-weighted sequences (GE-EPI) which also take into account larger vessels.

The injection is carried out at fast rate to deliver a quantity of contrast agent as quickly as possible (as close as possible to the notion of a bolus, corresponding to an instantaneous delivery). It is accompanied by a dynamic acquisition in echo planar sequence (SE-EPI, GE-EPI) to measure the signal before (baseline), during and after the first pass of the bolus.

 

 

The perfusion parameters are extracted by post-processing the signal curves, and appear in the form of parametric images. Two main techniques are employed to do this:

  • Modeling the signal curve (gamma-fit) to eliminate the contribution of tracer recirculation: this provides only a relative quantification of the perfusion data.
  • Deconvolution by the arterial input function, to take into account the dispersion of the bolus in time and the patient’s cardiac function (albeit with errors, linked to the greater distal spread of the bolus and the hematocrit differences between small and large vessels, causing variations in the proportion of plasma volume in which the tracer is distributed).

 

Measured parameters
  • TA: time of arrival of the contrast agent in the slice after injection
  • TTP (Time To Peak): time corresponding to the maximum contrast variation
  • MTT (mean transit time)
  • Peak amplitude: percentage loss of intensity of the maximal signal
  • rCBV (regional cerebral blood volume): index of cerebral blood volume determined by the area below the decreasing signal curve
  • rCBF (regional cerebral blood flow): index of cerebral blood flow corresponding to the ratio rCBV/MTT.

RCBV and rCBF are relative measurements, giving the calculation of ratios between pathological zone and healthy zone (which serves as a reference). The relative quantification of these measurements requires a deconvolution by the arterial input function, obtained from the signal of an arterial vessel of large dimension present in the image. This takes into account both the imperfect nature of the bolus, which is not instantaneous, and the patient’s cardiac function.

 

Optimization

The quality of perfusion imaging is affected by numerous factors:

  • Total acquisition time must include enough reference images before injection, and encompass the first pass of the bolus in the entire volume, without taking too long (recirculation).
  • The bolus must be injected as quickly as possible. A strong concentration of the agent will provoke a more significant drop in signal. The quality of the venous access, injection rate and patient hemodynamics must all be taken into account.
  • The patient’s movements must be avoided as these can induce calculation errors in the parametric images.
  • The choice of sequence type (GE-EPI / SE-EPI) will have an effect on the signal-to- noise ratio (better in GE-EPI) and microvascularisation sensitivity (better in ES-EPI, as TE increases, to the detriment of the signal-to-noise ratio).
  • Gadolinium chelates do not cross the normal blood-brain barrier. In pathological conditions where the latter is crossed, the contrast agent produces extravasation, reducing the T1 of the tissues. It contaminates the perfusion signal, leading to the risk of underestimating cerebral blood volume. More specifically, an increase in the signal above the baseline after the first pass of the contrast agent, will not show up in gamma curve modeling.