HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Gallezot, J.-D. Articles by Carson, R. PubMed Articles by Gallezot, J.-D. Articles by Carson, R.

A new reference region kinetic analysis method: Comparison with SRTM and MRTM

¹ Diagnostic Radiology, Yale University, New Haven, Connecticut

1602

Objectives: SRTM and MRTM are among the simplest methods to quantify PET data using a reference region (RR). SRTM tends however to be biased (when the underlying hypotheses are not fulfilled), and MRTM tends to be noisy (due to the use of late data only). This study presents a new non-linear t* method, which aims at providing a trade-off between MRTM accuracy and SRTM precision.

Methods: Like MRTM, the new method is based on the Graphical Analysis and use three parameters: V_R, the distribution volume ratio, , the slowest eigenvalue of the target region, and R₁, the ratio of K₁ values in the target and reference tissues. The operational equation is: C_T(t) = R₁C_ref(t) + (V_R-R₁) C_ref(t) \#8857; e^-(t-t*) + (V_R C_ref(u)du-C_T(u)du) x e^-(t-t*), t > t*, with integrals from 0 to t*. The 3 parameters are estimated with weighted least squares. The new method was tested on simulated time-activity curves of [C-11]P943, a 5-HT_1B antagonist radioligand, using a typical input function. We tested 2 versions, without (V1) and with (V2) a coupled fit of the early data (t<5 min) with the model C_T(t) = R₁C_ref(t) to help estimate R₁. The methods were assessed using the bias (B) and standard deviation (SD) of BP_ND (=V_R-1) estimates from noisy replicates.

Results: MRTM showed little bias (1%) and a SD of 22% with t* of 20 min. SRTM showed lower noise (6%) but larger bias (-12%). With t* of 20 min, V1 showed little bias (-2%) and a SD of 24%, similar to MRTM. For V2, as expected, SD was reduced (11%) but t* must be increased (50 min) to keep the bias low (-5%).

Conclusions: The new RR method shows promise for noise reduction of BP estimates. Additional studies are required to define which tracer kinetics are most appropriate for the new method.

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Gallezot, J.-D. Articles by Carson, R. PubMed Articles by Gallezot, J.-D. Articles by Carson, R.