Welcome (back) to the PET journal Club. Participation is open.PET Journal Club Evan's Home Page
Friday Nov 11
Dynamic PET Data Analysis, Maxoyer, Huesman, Budinger,
Knittel. JCAT 1986 10(4) 645-653
Thurs Sept 15 - 10am
Endres
et al. 2003 Carfentanil modeling by various reference region methods
Rubin
et al . 2001 Rat Head Holder Evaluation
Thurs Aug 25 - 10 am.
(we will try to discuss 3 short papers.)
Serotonin
1a is reduced in Panic Disorder
Serotonin
1a is not changed in PTSD
Serotonin
1a in depression
Ancient history of the PET
journal club
Wed Aug 21 2002, 4 pm
This recent NeuroImage paper by Pappata et
al. in Orsay is a very exciting twist on a new application of PET for
detection
of NEUROTRANSMITTER
activation in a single subject in a single
scanning session. "In
Vivo detection of striatal dopamine release during reward ..."
Briefly: Their model assumes that time-activity
curves measured by PET during the performance of a task -purported to
raise
dopamine levels-
can be described by a linear combination of a
few temporal copmonents (think of them as basis functions). These
temporal
components
are chosen based on simulations of 11C-raclopride
uptake with and without dopamine and/or blood flow changes at the time
of the task.
If they estimate the coefficient of the basis
function describing deflections in the time curve due to DA changes to
be nonzero, they consider
that to be an indication of DA activation.
The strength of the activation is related to the significance of the
t-test
comparing the coefficient
and zero (null hypothesis.)
Interesting points to watch for
1. Is it plausible that dopamine activates only
for unexpected (R-) reward and not expected reward?
2. Can we really describe the response of the
PET curve to DA as a LINEAR model? After all, the
model that they (we all) use to describe
raclopride
competition with dopamine is NONLINEAR.
3. Note their comment that they are HIGHLY
sensitive to motion artifacts.
4. For members of the Summer Image
Registration
Course - note their rejection of the mutual information
algorithm for motion correction.
5. For those of you who liked Sweeney Todd,
note the reference to the paper by Fried,
2001:
Microdialysis in people!!!!
Thurs May 2, 2002
I will give a brief introduction to multiple-injection ligand
studies
with PET for the purpose of getting precise estimates of B'max, the
number
of available receptors.
Lets use two of my papers as introductory reading
Synapse, 1996
Synapse,
1999
Note. This method was pioneered by Jacques Delforge of the PET
group in Orsay. (But with short notice, I have my papers digitized and
NOT
his - so we can start here at least.)
The 1996 paper is an investigation with your favorite DA ligand,
CFT,
to estimate number of Dopamine transporters in the striatum of monkeys
before and after MPTP as a model of Parkinson's.
The 1999 paper is a letter defending the 1996 paper AND giving a
succinct
explanation of why the multiple-injection scheme is valid - and
preferable
to other methods (complexity notwithstanding.)
These papers - coa-authored by myself and Brad Christian, PhD, among others, will serve as a good intro to Dr Christian's talk the next day in the same room
Wed December 19, 2001
Lets look into Partial Volume Effects and PV Correction methods.
Muller-Gartner HW, Links JM, Prince JL, Bryan RN, McVeigh E, Leal JP, Davatzikos C, Frost JJ. Measurement of radiotracer concentration in brain gray matter using positron emission tomography: MRI-based correction for partial volume effects.J Cereb Blood Flow Metab 1992 Jul;12(4):571-83
This is a classic PET paper which was the first (?) to offer a pixel-by-pixel method for correcting images for the partial volume effect.The method uses information from MRI images. It segments the MR into different tissue types; creates simulated PET images for a given 'true' concentration for each tissue type assuming uniformity within a tissue type and then 'overlays' the simulated PET for each type to build up a simulated PET image of the brain. Doing this multiple times establishes a relationship between true concentration and measured (in the simulated) concentration. This relationship is the basis for correcting real images back to their 'true' concentrations.and then lets see how the theory affects some data: perhaps a blood flow application...
Wed. Nov 14, 2001
Koeppe RA, Holthoff VA, Frey KA, Kilbourn
MR, Kuhl DE (1991): Compartmental analysis of [11C]flumazenil kinetics
for the estimation of ligand transport rate and receptor
distribution
using positron emission tomography. J Cereb Blood Flow Metab 11:735-744.
This paper goes to what are we measuring when we measure DV'' -
that's the so called Vdi image!
Ilgin et a.l Neurology
(1999), 52:1221-1226.. PET imaging of Dopamine Transporter in
progressive
supranuclear palsy and Parkinson's disease.
This paper is from competitors to our own Allison Brashear, MD - We
must understand our enemy to vanquish them!
Wed. Oct 17:
Simplified Reference
Tissue Model for PET Receptor Studies. A. Lammerstma and S Hume.
NeuroImage (1996), 4:153-158.
This is one popular way to make Binding Potential estimates/images
in the ABSENCE of blood data. But it comes with its share
of assumptions.
The above paper is based on this earlier reference region
paper:
I will try to scan it for on-line access, otherwise, you'll need to
come
to my
office to copy it. But its a really good discussion of what it
means when the reference region is not perfect (ie devoid of receptors)
and what assumptions must be satisfied to use a reference region model.
Compartmental Analysis of Diprenorphine Binding to Opiate Receptors
in the Rat In Vivo and Its Comparison with Equilibrium data.
Cunningham, Hume et al., J
Cereb Blood Flow and Metab, (1991), 11: 1-9.
Preview: This paper has microdialysis, PET, raclopride, Parkinsons, DOPA, Dopamine, small animals --- it has it al! Come hear why QUANTITATIVE imaging in small animal PET is not straightforward.
Evan's Cliff notes:Figure 1. demonstrates a crucial bias in BP measurements from PET data when the dose of cold carrier (i.e., unlabeled) ligand is too high. The question becomes, Can it be corrected? If not, we can't compare results across studies if the mass dose is greater than the threshold value at which measured BP drops off. (see below)
....................................|
...................................V
Equation 1: BP = [appBmax/{Mass dose + appKD}] + NS describes Figures 1 and 2. But is it correct? Where does it come from? Does it make sense that Binding Potential should be non-zero at very large mass doses? I don't think so! I think the form of the model should be simply: BP = B / [K + M] . Although, I know that such a functional form would have difficulty fititng the points at 5000 nmol/kg in the Fig, above.
Figure 3. demonstrates the time course of dopamine response to a pharmacological stimulation. This one presumably involves synthesis is thus would be expected to be slower than simply "release". Nevertheless, it is relevant to anyone hoping to image DA changes.
Reminder:
1 Curie = 37 GBq
so a specific activity of 37GBq/umol = 1Ci/umol = 1000mCi/umol (a decent SA for 11C-raclopride)Interesting: Here's what happens if we merge data from 2 different studies of BP vs mass by the Hammersmith group.
Possile Future Papers:
Perhaps we will look at blood flow models in late Dec or early
Jan.
Stay tuned!