2011 OSU Molecular Life Sciences
Interdisciplinary Graduate Programs Symposium

 

Home

Registration

Agenda

Abstracts

Poster abstracts

Poster number 26 submitted by Katherine Binzel

How does the acquisition time impact PET quantification?

Katherine M. Binzel (Department of Radiology, The Ohio State University), Jun Zhang (Department of Radiology, The Ohio State University), Nathan C. Hall (Department of Radiology, The Ohio State University), Michael V. Knopp (Department of Radiology, The Ohio State University)

Abstract:
Image quality and accurate quantification are critical factors in the evaluation of positron emission tomography (PET) studies for initial staging and therapy response assessment, as is scan time for the patient and overall work flow. We investigate the impact of varying the bed acquisition time on the standardized uptake values (SUV). Using the listmode data of 31 patients acquired on a Philips GEMINI TF 64 PET/CT, emission scans were reconstructed with 3 iterations, 30 subsets, 14.1 cm kernel width using 15, 30, 60, 90, 120, and 180 seconds per acquisition volume from skull to thighs. Total counts per frame were recorded and regions of interest (ROI) were placed on 47 tumor lesions as well as reference areas in the heart, liver, kidneys and bladder. PET images were quantitatively evaluated using SUVmax from the 90s per frame standard local protocol as a reference value, calculating ΔSUVmax for all other durations. Total counts for longer frame durations increased, as expected, linearly. However variations in SUVmax were not consistent, doubling the acquisition length, and thus number of counts, did not reproducibly decrease ΔSUVmax. Accepting a 10% ΔSUVmax variation, 15s per frame in the region of the bladder, 30s at the level of the heart and kidneys, and 60s in the liver revealed quantitative data within the limit. The average SUVmax of all tumor lesions measured was within 12% of the reference value, though there were considerable individual variations. While SUVmax generally increased with shorter acquisition, lower intensity SUVmax (<5) lesions did not follow this trend. Differences due to background within the different anatomic regions were observed and suggest that a count density approach should be considered when optimizing PET workflow.

Keywords: positron emission tomography, quantification