Biomedical Engineering Reference
In-Depth Information
exchanges of the tracer within the tissue compartment (the impulse response
function of the tissue).
Ideally, the input function should be measured in the capillaries of the tis-
sue of interest but this is not possible. Instead, samples are obtained by fre-
quent blood sampling at the peripheral artery such as radial artery or femoral
artery, under local anesthesia. The arterial concentration of the tracer has to
be measured because the concentration of the tracer in venous blood may dif-
fer markedly from that in arterial blood. This is particularly true for tracer with
high extraction fractions. Unless the tracer can freely diffuse between red blood
cells and plasma, the concentration time course in plasma is required for the
input function. Another consideration is the in vivo formation of radiolabeled
metabolites that can contribute to the radioactivity counts in total plasma or
whole-blood, necessitating a time-dependent estimation of the fraction of radio-
labeled metabolites present in plasma. Estimation of radiolabeled metabolites in
plasma is often possible by means of chromatographic analysis (e.g. using HPLC,
TLC, or octanol extraction) for the measured samples. Modeling of metabolites
formation is also possible [65,66], but the statistical quality of the measurements
may not support reliable estimation of the additional parameters in the more
complicated compartment model.
Arterial blood sampling is currently regarded as the gold standard method of
measuring the input function for PET study in spite of many possible sources of
error such as insufficient sampling rates and counting errors in the blood sam-
ples. Particularly for radiotracers with rapid kinetics and short measurement
times, delay and dispersion effects between the blood sampling site and the ac-
tual arterial blood that drives the tissue of interest need to be taken into account
during the compartmental model fitting. The procedure of arterial catheteriza-
tion under local anaesthesia, however, is very invasive in nature and would
involve potential risks of arterial sclerosis and ischemia to the distal extrem-
ity. An alternative, yet less invasive, approach is to obtain blood samples from
arterialized-vein (or “hot” vein). This method is referred to as the arterialized-
venous (a-v) sampling method. The hand is heated in a water bath which then
promotes arterio-venous shunting to avoid the discomfort and risks associated
with arterial cannulation [67, 68]. Since it only requires the placement of venous
cannulas, it is less invasive and is better tolerated than arterial cannulation by
the subjects. However, it requires prolonged hand warming to ensure adequate
shunting, and it is very dependent upon the site chosen for the placement of the
Search WWH ::




Custom Search