brede_mat_loganplot - Logan plot analysis
BP = brede_mat_loganplot(M, Mref, varargin)
Input: M 'Mat' structure or numerical matrix
Mref 'Mat' strutcure or numerical matrix with
reference value/plasma curve
Property: EndIndex [ {Inf} | positive integer ] Index where the
linear fitting of the Logan plot should end.
Info [ {0} | value ] Amount of feedback/debug
information
k2prime [ {Inf} | positive value ] k'2 paramere
StartIndex [ 1 | positive integer ] Index where the
linear fitting of the Logan plot should start.
TimingType [ end | {middle] ] Determines where the
integration (cumulation) of the values in the
matrix should stop when the 'Times' are given.
Times 'Mat' structure with times: the first column
should be the starting times of the scan
while the second column should be the end
time. There should be as many rows as there
are rows in the M and Mref matrices.
Output: BP Binding potential in a 'mat' structure
Logan plot analysis is a type of kinetic modeling for dynamic PET
scans. This function will estimate the binding potential.
The function expects an input where the activity curve for each
region/voxel is in a single column and each row correspond to a
time point/period. The outputted binding potentialwill be in a
single row.
The linear fit is computed with points from 'StartIndex' to
'EndIndex'.
The function may not estimate all voxels in a volume well. Those
where it is not possible to compute a sensible binding potential
value will have a NaN. A spatial smoothing with a median filter
as implemented in the brede_vol_smooth function may remove these
voxels.
Note that this function is not properly validated.
Reference:
PMOD, Logan's Reference Tissue Model based on Average k2',
http://pmod.com/technologies/doc/pkin/2322.htm
See also BREDE, BREDE_MAT, BREDE_MAT_CUMSUM, BREDE_MAT_MRTM,
BREDE_VOL_SMOOTH.
$Id: brede_mat_loganplot.m,v 1.8 2008/06/10 17:12:42 fn Exp $