1. Optimisation of Chest Computed Tomography Using a Phantom: Impact of mAs and Reconstruction Techniques on Image Quality
C.S. Reis1; T. Faqir2; V. Harsaker3; P. Hogg 2; L. Kristoffersen3; I.L. van Rein4; K. Stancombe2; N.C. Warmerdam4;
C. Wergeland3

2. Outline Introduction Objectives Methods Results & Discussion
Conclusion
Further work

3. Introduction

4. Optimisation is needed
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Why CT
Optimisation is needed
Highest growth(1)
Used in screening programs(2)
High effective dose(3)
1.Commission E. EUROSTAT - Your key to European statistics [Internet]. Eurostat [cited 2015 Aug 5]. p. Needs for health care. Available from: 2. Schmidt CW. CT scans: balancing health risks and medical benefits. Environ Health Perspect. 2012;120(3):118– Team TNLSTR. Reduced Lung-Cancer Mortality with Low-Dose Computed Tomographic Screening. N Engl J Med [Internet] Aug 4;365(5):395–409. Available from:

5. How to reduce dose Exposure Parameters Reconstruction techniques
Introduction
Methods
Results/Discussion
Conclusion
Further Work
How to reduce dose
Exposure Parameters
Reconstruction techniques
Filtered back projection (FBP)
Iterative Reconstructions (IR)
1-Klink T, Obmann V, Heverhagen J, Stork A, Adam G, Begemann P. Reducing CT radiation dose with iterative reconstruction algorithms: The influence of scan and reconstruction parameters on image quality and CTDIvol. Eur J Radiol [Internet]. Elsevier Ireland Ltd; 2013;83(9):1645–54. Available from: 2-Willemink MJ, De Jong P a., Leiner T, De Heer LM, Nievelstein R a J, Budde RPJ, et al. Iterative reconstruction techniques for computed tomography Part 1: Technical principles. Eur Radiol. 2013;23(6):1623–31.

6. Introduction
Methods
Results/Discussion
Conclusion
Further Work
Objectives
To verify if the mAs variation affects detection of anatomical details
To verify if reconstruction techniques affects the observation of anatomical details

7. Methods

8. Phases of Research Image Acquisition Image Analysis Phantom – LUNGMAN
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Phases of Research
Image Acquisition
Phantom – LUNGMAN
Positioning: supine, head-first
Siemens Somatom Definition AS 128 slice CT scanner
Exposure parameters: 3 different fixed mAs values (10, 20, 30); 120 kVp, 1.2 pitch, 0.6 mm slice thickness and 512 x 512 matrix
4 reconstruction techniques: FBP or SAFIRE level 1, 3, and 5
Image Analysis

9. Phases of Research Image Acquisition VISUAL: Anatomical criteria(1)
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Phases of Research
The subjective anatomical visualisation for each structure was rated using a 4-point Likert scale as:
1 - It is not visible
2 - I can see it partially
3 - I can see it
4 - It is clearly defined
The subjective noise with a 5-point Likert scale:
1 - very poor: excessive noise or poor vessel wall definition
2 - poor: poor vessel wall definition and prominent image noise
3 - adequate: some image noise, vessel walls definition is minimal
4 - good: minimal image noise definition of vessel walls are visible
5 - very good: excellent definition of vessel walls, limited perceptual image noise.
The subjective overall image quality using a 5-point Likert scale:
1 - very poor: poor image quality due to artefacts, no definition between anatomical structures
2 - poor; prominent artefacts, minimal definition between anatomical structures
3 - adequate: minor artefacts present, definition between anatomical structures
4 - good: no perceptual artefacts present, clear definition between anatomical structures
5 - very good: no perceptual artefacts present, total definition between anatomical structures.
Image Acquisition
VISUAL:
Anatomical criteria(1)
Noise
Global IQ
Wilcoxon-test with Bonferroni correction - statistical differences between IQ for all images
Image Analysis
Commission E. European Guidelines on Quality Criteria for Computed Tomography European Guidelines on Quality Criteria [Internet]. Menzel H, Jessen K, Panzer W, Shripton P, Tosi G, editors. Europe. European Commission´s Radiation Protection Actions; p. Available from:

10. Phases of Research Image Acquisition OBJECTIVE Measurements on ImageJ
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Phases of Research
Manning DJ, Ethell SC, Donovan T. Detection or decision errors? Missed lung cancer from the posteroanterior chest radiograph. Br J Radiol [Internet] Mar;77(915):231–5. Available from:
Image Acquisition
OBJECTIVE
Measurements on ImageJ
Edges of anatomical structures
Pixel values in a plot profile(6).
Trend line was added to the linear points of the plot profile
Edge gradient was calculated
Image Analysis

11. Results & Discussion

12. Visualisation of Anatomical Structures
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Visualisation of Anatomical Structures
RT and mAs: variation in visibility within the acceptance level
> variation for 10 mAs and FBP in axial images
Not significant (Wilcoxon Test p = 0.491; p = 0.124; p = 0.384)
To analyse the anatomical structures globally we added the scores to achieve a global analysis of the anatomical structures and as you can see (line on the graph) the acceptable score to have total visualization of the structures is 12. Almost all images have this score except at 10 mAs w/ FBP in axial images. The best results were observed at 20 mAs w/ SAFIRE5.
However, these variations and these differences are not statistically significant.

13. Introduction
Methods
Results/Discussion
Conclusion
Further Work
Noise and Overall IQ
No significant difference Wilcoxon Signed Rank Test (FBP with SAFIRE 1, 3 and 5 respectively: p = 0.405; p = 0.251; p = 0.083; SAFIRE 1,3 and 5: p = 0.046; p = 0.926).
The same results were obtain for the other two parameters that we ask to the observers to analyse: Noise and overall IQ.

14. Edge Gradient - Sharpness
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Edge Gradient - Sharpness
These results were also confirmed by the analysis of edge gradient. This gradient allows to see how sharp is the edge. So the measure should be around 90º to guarantee that the edge is well visible. With our results It is possible verify that the mAs and RT promote differences in images and the best results were obtained for 20 mAs. BUT (move to the table) when we look to the this table that shows how much different are these differences in edge we can see that the differences are very small.
Higher sharpness for 20 mAs – SAFIRE 5
Lower noise, easier detection of edges

15. Clinical Relevance mAs
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Clinical Relevance
mAs
Important to study mAs further and observe impact on image quality
Clinical indication
Real anatomical structures and lesions can be affected in different ways depending on clinical indication
This study showed that visualisation of anatomical structures was possible even at a low mAs value of 20, partial visibility was made at 10 mAs, therefore future research needs to consider mAs and other clicical indications because the diferent characteristics of lesions and anatomy can affect the visualisation

16. Dose (DLP) Comparison with the European Guidelines:
Introduction
Methods
Results/Discussion
Conclusion
Further Work
Dose (DLP)
Comparison with the European Guidelines:
The dose that was used to produce the
Dose reduction of 91.1% can be achieved whilst maintaining visualisation of anatomical structures

17. Introduction
Methods
Results/Discussion
Conclusion
Further Work
Limitations
Phantom
Artefacts cannot be simulated
Lesion variety
Anatomical variations
Patient size
Observers with different clinical experience – impacts analysis (1)
Kakinuma R, Ashizawa K, Kobayashi T, Fukushima A, Hayashi H, Kondo T, et al. Comparison of sensitivity of lung nodule detection between radiologists and technologists on low-dose CT lung cancer screening images. Br J Radiol [Internet] Sep;85(1017):e603–8. Available from:

18. Conclusion

19. Introduction
Methods
Results/Discussion
Conclusion
Further Work
How does the mAs and reconstruction techniques affect the detection of anatomical structures in lung CT
Detection of anatomical structures was largely unaffected by reconstruction techniques and mAs variations

20. Further work

21. Introduction
Methods
Results/Discussion
Conclusion
Further Work
Next step
To verify if there is a relationship between visual and physical measures of image quality
Gain a better understanding of visual and clinical value of reconstruction techniques at lower doses
Bigger variety in clinical indications, patient size and exposure parameters.

22. Acknowledgments Ruurd Visser for support - statistical analysis
5 observers
Staff of Hanze University and UMCG that helped us acquire the images for this research

23. Thank you for your attention!