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 Saha, K. Articles by Case, J. PubMed Articles by Saha, K. Articles by Case, J.

Recovery of temporal resolution of blood-pool time-activity curves in Rubidium-82 myocardial perfusion PET images

¹ CVIT, Kansas City, Missouri

857

Objectives: Arterial and myocardial time-activity curves (TAC) are needed to quantify myocardial blood flow (MBF) with Rb-82 PET. Fine temporal sampling for flow measurement are preferred, however, efficiency for computing MBF is burdened by processing required for the many temporal volumes. We developed an iterative deconvolution technique (IDT) to recover the temporal resolution (TR) from coarsely sampled TACs and improve processing efficiency for MBF calculations.

Methods: Eight patients with clinical rest/stress Rb-82 PET/CT on a Biograph-16 (Siemens, TN) were studied. The 2.5 min rest/stress listmode data was rebinned to two dynamic frames: a) 6s TR (20 frames)[1] and b) 12s TR (12 frames). IDT was applied to recover the 12s TR TAC for the blood-pool (BP) and myocardium to predict the 6s TR data. Gaussian kernel size was determined by blurring the 6s-TR for 1-30 sec to match the 12s TR data from 3 studies and applied to the other 5 studies. For the BP curves, time-to-peak (TTP1), peak values (PV1) and FWHM were analyzed. Time to 95% of plateau (TTP2) and plateau value (PV2) were estimated from the myocardial curves.

Results: Gaussian kernel size was 6.2s. TACs in the myocardium were equivalent for 6s and 12s TRs in normal and defect regions (TTP2=5.2%±7.3 and 6.7%±0.5 respectively, PV2=4.1%±5.7 for normal and (1.3%±0.3) for defects). 6s and 12s TR TACs had a significant difference in the BP region (TTP1=-1.7%±17.5, PV1=14.9%±2.6 and FWHM=18.5%±7.1). Applying IDT, the 6s TR and IDT 12s TR were highly correlated (r=0.997) with a minimal MD=-2.2%±6.86 and FWHM (r=0.97, MD=-1.8%±2.7). All p>0.05.

Conclusions: This iterative deconvolution technique successively recovers the temporal resolution of time-activity curves in the blood-pool region with a 40% improvement in processing efficiency for MBF quantitation.

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 Saha, K. Articles by Case, J. PubMed Articles by Saha, K. Articles by Case, J.