Optimization of image quality and accuracy of low iodine concentration quantification as function of dose level and reconstruction algorithm for abdominal imaging using dual-source CT: A phantom study - 04/01/22
, Julien Frandon, Asmaa Belaouni, Philippe Akessoul, Takieddine Addala, Laure Berny, Jean-Paul Beregi, Joël Greffier
Bienvenue sur EM-consulte, la référence des professionnels de santé.
Article gratuit.
Connectez-vous pour en bénéficier!
Highlights |
• | Dose reduction has an impact on image quality in virtual monochromatic images (VMIs) and iodine quantification using dual-energy CT. |
• | Iterative reconstruction improves HU accuracy and image quality in VMIs and does not affect iodine accuracy. |
• | Using ADMIRE for VMIs could help reduce radiation dose and/or the amount of injected iodine. |
Abstract |
Purpose |
The purpose of this study was to assess the impact of advanced modeled iterative reconstruction (ADMIRE) algorithm and dose levels on the accuracy of Hounsfield unit (HU) measurement, image noise and contrast-to-noise ratio (CNR) in virtual monochromatic images (VMIs) with low iodine concentrations, and the accuracy of iodine quantification.
Materials and methods |
A CT phantom was scanned with dual-source CT using abdomen-pelvis examination parameters at four dose levels: 5, 8, 11 and 20 mGy. Images were reconstructed using filtered-back projection (FBP) and ADMIRE levels 3 and 5 (A3-A5). HU accuracy was assessed calculating the root-mean-square deviation (RMSDHU). Image noise and CNR were computed on VMIs at 40/50/60/70 keV for 4 iodine inserts with 0.5, 1, 2 and 5 mg/mL concentrations. Accuracy of iodine quantification was assessed by the RMSDiodine and iodine bias (IB).
Results |
The RMSDHU decreased significantly as the dose levels increased compared to 5 mGy, except for 8 mGy with A3 (P = 0.380) and with A5 level (P = 0.945). Noise increased by 63.0 ± 3.0 (standard deviation [SD])% from 20 mGy to 5 mGy. Noise decreased significantly by -53.8 ± 0.9 (SD) % with A5 compared to FBP. The CNR decreased by -43.1 ± 6.5 (SD)% from 20 mGy to 5 mGy. It increased using ADMIRE, and as the ADMIRE levels increased. The RMSDiodine and IB decreased as the dose level increased, and this was similar for all reconstruction types.
Conclusion |
ADMIRE strongly improves image quality in VMIs and slightly improves HU accuracy but does not affect the accuracy of iodine quantification.
Le texte complet de cet article est disponible en PDF.Keywords : Dual-energy CT, Iodine quantification, Monochromatic images, Iterative reconstruction
Abbreviations : ADMIRE, CNR, CT, CTDIvol, DECT, DLCT, DSCT, SECT, FBP, HU, IB, IR, MBIR, mAs, RMSD, SD, VMI
Plan
Vol 103 - N° 1
P. 31-40 - janvier 2022 Retour au numéro
Article précédent
- Comparison of two deep learning image reconstruction algorithms in chest CT images: A task-based image quality assessment on phantom data
- Joël Greffier, Julien Frandon, Salim Si-Mohamed, Djamel Dabli, Aymeric Hamard, Asmaa Belaouni, Philippe Akessoul, Francis Besse, Boris Guiu, Jean-Paul Beregi
- Abdominal metastases from Merkel cell carcinoma: Prevalence and presentation on CT examination in 111 patients
- Raphael Girard, Manel Djelouah, Maxime Barat, Paul Fornès, Sarah Guégan, Nicolas Dupin, Philippe Soyer, Christine Hoeffel
Bienvenue sur EM-consulte, la référence des professionnels de santé.