Keywords:
CAD, Physics, CT, Computer applications, Computer Applications-Detection, diagnosis, Tissue characterisation
Authors:
G. Ficarra1, E. Barabino2, M. Verda3, S. Casella4, S. Caprioli5, G. Cittadini1; 1Genoa/IT, 2Genova/IT, 3Imperia/IT, 4Savona/IT, 5Arenzano/IT
DOI:
10.26044/ecr2019/C-2626
Results
According to our results the acquisition parameters that mainly influence the variability of CTTA features are milliamperage and section thickness.
By keeping milliamperage constant we obtained the lowest variability (Table 2), with an average QCD of 0,065±0,006 for scanner A and 0,067±0,010 for scanner B.
Conversely our results showed a somewhat lower advantage in keeping the voltage constant (mean QCD 0,072± 0,0119 for scanner A; 0,087± 0,006 for scanner B). When comparing the scanners independently from acquisition parameters,
only a moderate difference emerged between machines (mean QCD for scanner A: 0,080± 0,100; for scanner B 0,095± 0,160; fig.2 and fig.
3).
However,
a difference emerged by comparing the devices at the same voltage; if at 80 kV the variability of the features does not significantly differ (0,087±0,124 for scanner A; 0,094±0,168 for scanner B) increasing the voltage lead to a greater difference (respectively mean QCD 0,071±0,112 vs 0,083±0,163 at 100kV; 0,067±0,183 vs 0,081±0,155 at 120 kV; 0,063±0,146 at 140kV; fig.4).Among CTTA features, co-occurence matrices-derived features showed the lowest variability (fig.5).
Not surprisingly, a difference emerged when comparing smaller and larger sized co-occurence matrix-derived features, the latter showing lower variability (fig. 6 and fig.7). According to our results,
the optimal acquisition parameters to reduce variability in CTTA features can vary for each scanner; a compromise that allows to minimize variability can be achieved by maintaining the section thickness between 2.5mm and 5mm,
milliamperage between 150 mA and 350 mA and voltage around 140kV (fig. 8 and fig.9).