This page is obsolete: please read the new version.

Reference tissue input compartment models for regional analysis of binding potential

Background

Two reference tissue compartment models are available for quantitation of binding potential (BP_ND): the (original) full reference tissue compartment model (FRTM) and the simplified reference tissue model (SRTM). The advantage of all reference tissue models is that blood sampling and plasma metabolite analysis are not needed. Thus the errors caused by the uncertainties in the measured plasma metabolite fractions are avoided. The plasma input is replaced with reference tissue showing non-existent (or very low) specific binding.

FRTM includes four parameters: R₁ (ratio of the K₁ values of regions of interest and reference tissue), k₂, k₃, and BP_ND (k₃/k₄) can be estimated using nonlinear fitting. SRTM includes these same parameters except k₃; nonlinear fitting is usually applied to estimate these parameters, but also linear methods could be used.

One assumption common to all of reference tissue models is that K₁/k₂ is similar in all studies regions. To provide unbiased results, SRTM also requires that the kinetics in region of interest is fairly simple, and could be fitted reasonably well with two-compartment model.

If kinetics of the tracer are simple enough to fulfill the requirements of SRTM, then the more complicated RTCM would produce results with high variance. If the reference region has specific binding, the BP_ND will be overestimated:

BP_ND^Apparent = (BP_ND^ROI + 1) / (BP_ND^Reference + 1) - 1

The parameters R₁ and BP_ND are unitless. The unit of k₂ and k₃ is min^-1.

Model fitting software

The parameters of RTCM and SRTM can be estimated using programs fit_frtm and fit_srtm or regbfbp. BP can also be estimated from the multilinear form of SRTM using program lhsrtm.

Steps of calculation:

Any or all of of the following steps can be done in Windows XP command prompt window or Solaris terminal window.

1. Preparation of regional tissue TAC data

This is explained in detail elsewhere. In short: draw ROIs and calculate regional TACs from dynamic images, and calculate averages over planes and regions if required.

2. Adding weights to regional tissue TAC data

Add weights to regional tissue TAC data.

3. Computing the BP

Depending on the suitable model, run one of the programs fit_frtm and fit_srtm with at least the following command line parameters:

regional tissue TAC file (DFT datafile)
name of the reference region in the TAC file
length of study to be used in the fitting (usually a large number to allow fitting to the end of data)
name for result file; note that existing file is overwritten

In addition, the fitted TACs can be saved for later use.

For example, the command could be like this:

fit_srtm ua268.dft 'cer avg' 999 ua268.res

The initial values and constraints for the model parameters can be entered to the programs fit_frtm and fit_srtm with option -i. You should use the same parameter settings file for your whole study material.

As an alternative, regional BP can be estimated also using the basis function approach, which is also used in the Roger Gunns RPM program for producing BP_ND images. This can be done using program regbfbp with at least the following command line arguments:

regional tissue TAC file (DFT datafile)
name of the reference region in the TAC file, or name of file containing only the TAC of reference region
isotope (C-11 or F-18)
name for result file; note that existing file is overwritten