| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
|
|||||||||
|
|
Instrumentation & Data Analysis: Image GenerationImage Generation Posters |
1 Nuclear Medicine, St Elisabeth Hospital, Zottegem, Belgium; 2 Nuclear Medicine, University of Louvain, Bruxelles, Belgium
1671
Objectives: We evaluate the quality of the quantification provided by the pinhole SPECT method. To do so we focus on the determination of the thyroid volume.
Methods: We performed more than 200 acquisitions of phantoms, with various sizes (from 20 to 150 ml), shapes ( 8 available) registered counts (varying in a range from 1 to 6, but still clinically relevant). All the acquisitions were made on a 180° anterior circular orbit. The dedicated software uses an OSEM reconstruction algorithm, adaptive filters and a threshold of 43% for ROI delineation. The number of iterations ranges from 1 to 20, with a subset number of 6. The pixel size was chosen between 2 and 3 mm. Various signal/background levels ranging from 1.1 to 2.0 were also evaluated. Finally, the reproducibility of the method was tested on 6 patients injected with I-123 and imaged twice the same day.
Results: The calculated mean volume of the classical thyroid phantom was 99,4 + 2.3% of the expected one. Accepting even an error level of 5% or, for the volumes less than 50 ml, a discrepancy of 3 ml, no phantom volumes was significantly affected by the studied parameters. There was a trend to overestimate the volume when it was smaller than 25 ml and/or when the phantom was larger than 7 cm along the longitudinal axis. For signal to background ratio above the value of 3 (a low value in clinical conditions), no significant volume variation was also detected.
Conclusions: We can conclude that the volume measurement by pinhole SPECT, at least for the thyroid in clinical conditions, seems precise and robust.
| ||||||||||||||||||||||||||||||||||||||
