Determination of optimal CT exposure factors for lung lesions using an anthropomorphic chest phantom for SPECT-CT

Thompson, J, Szczepura, K ORCID: https://orcid.org/0000-0002-2566-3308, Tootell, AK ORCID: https://orcid.org/0000-0002-8567-8659, Sil, J, Manning, D and Hogg, P ORCID: https://orcid.org/0000-0002-6506-0827 2010, Determination of optimal CT exposure factors for lung lesions using an anthropomorphic chest phantom for SPECT-CT , in: European Association of Nuclear Medicine (EANM) 2010, 9th-13th October 2010, Vienna, Austria.

Full text not available from this repository.

Abstract

PURPOSE: SPECT‐CT is a valuable modality for diagnosis and staging of primary and secondary malignant disease and also diagnosing benign conditions. CT incurs a significant radiation burden; consequently CT acquisition parameters require optimisation to conform to the ALARP principle. METHOD: An anthropomorphic chest phantom was scanned on the Toshiba Aquilion 16‐ slice MDCT at 135kV, 1mmx16 detector configuration (reconstructed at 3mm slices and 3mm intervals), 320mm FOV with variation of mA (10‐25) and pitch (0.94 and 1.44). Simulated lesions (5, 8 and 10mm at 100HU) were placed in clinically relevant positions within the phantom. 5 image readers of varying CT experience completed the study; data was analysed via Jackknife Free‐Response Receiver Operating Characteristic (JAFROC) and Dorfman‐Berbaum‐Metz Multi‐ Reader Multi‐Case (DBM‐MRMC) analysis. Signal to noise ratio (SNR) was also measured for all simulated lesions. RESULTS: JAFROC method revealed no significant difference between acquisition options (variation of mA and pitch) or image readers (ANOVA). Image readers with experience in producing reports on investigations made a greater number of non‐lesion localisations (NL) (false positives) compared to those who had no reporting responsibility or experience. Reader performance was generally better at lower pitch protocols for the same mA. Increased partial voluming could account for this. CONCLUSION: Results suggest equal performance for acquisition parameters in terms of the localisation information provided. This suggests no significant deterioration in image quality at lower resolutions. Despite this it is thought that low pitch (0.94) techniques should be favoured due to general improved reader performance at this level. No correlation was seen between SNR and JAFROC confidence ratings. Further study would attempt to investigate an even lower mA range for further dose saving potential.

Item Type: Conference or Workshop Item (Paper)
Schools: Schools > School of Health and Society
Journal or Publication Title: European Journal of Nuclear Medicine and Molecular Imaging
Publisher: Springer
ISSN: 1619-7070
Related URLs:
Depositing User: K Szczepura
Date Deposited: 26 Feb 2021 13:20
Last Modified: 26 Feb 2021 13:20
URI: http://usir.salford.ac.uk/id/eprint/59684

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)