EFSUMB EUROPEAN FEDERATION OF SOCIETIES FOR ULTRASOUND IN MEDICINE AND BIOLOGY ‘Building a European Ultrasound Community’ Young Investigator’s Award EUROSON 2016 Leipzig, Germany 28 October 14.00 – 15.30
7 minutes each for presentation plus 2 minutes discussion
In vitro quantification of tissue elasticity using three shear wave elastography platforms on liver fibrosis phantoms Anesa Mulabecirovic, MD 1,2 Anders Batman Mjelle, MD 1,2 Odd Helge Gilja, MD, PhD 1,2 Roald Flesland Havre, MD, PhD 1 1. National Centre for Ultrasound in Gastroenterology, Department of Medicine, Haukeland University Hospital, Bergen, Norway 2. Department of Clinical Medicine, University of Bergen, Bergen Norway
ABSTRACT Introduction To assess and validate the reproducibility of quantitative elastography measurements, using shear-wave methods on four individual tissue-mimicking liver fibrosis phantoms with known Young’s modulus. Methods
We used three different shear wave elastography platforms: GE Logiq E9 SWE, Philips iu22 XM ARFI and Samsung RS80A. Both linear (high frequency) and curvilinear (low-frequency) probes were applied. The objects were four individual tissue mimicking liver fibrosis phantoms with different Young’s modulus within the range of soft biological tissue (2.7kPa, 11.5kPa, 24.8kPa, 46.3kPa). Two individual investigators performed all measurements in parallel. Each investigator made ten non-continued measurements of each phantom. The platforms were evaluated for inter- and intraobserver variability, coefficient of variation, ICC and Bland-Altman using the median value. Statistical analysis was performed with SPSS. Results
All three elastography platforms showed excellent intra-and interobserver agreement (interclass correlation 0.981-1.000 and intraclass correlation 0.987-1.000). All four liver fibrosis phantoms could be differentiated by quantitative elastography, by all platforms (p<0.001). In the Bland-Altman analysis the differences in measurements were larger for the phantoms with higher Young’s modulus. All platforms had a coefficient of variation in the range 0.00-0.21 for all four phantoms, equivalent to low variance and high reproducibility. (see attached table 1) Conclusion
All systems used in this study provided a high reproducibility in quantitative measurements in a liver fibrosis phantom and excellent inter- and intraclass correlations. The GE Logiq E9 SWE had the best inter-and intraclass correlation, whilst Philips iu22XM ARFI and Samsung R08A provided elasticity measurements closest to the elasticity values provided by the manufacturer of the phantom.
Table 1. Median (min-max) elasticity measurements performed with three elastography platforms on liver fibrosis phantoms
Elastography system
Philips iu22 XM ARFI
Samsung RS80A SWE
Probe shape
curvilinear
curvilinear
linear
Observer
Phantom 1 2,7 kPa +/- 5%
A
1,93kPa (1,76-2,23)
8,29 kPa (7,359,66)
20,46 kPa (18,234,88)
47,09 kPa (38,2950,75)
B
1,89 kPa (1,8-2,01)
7,73 kPa (7,237,92)
19,68 kPa (18,8823,9)
44,74 kPa (36,7346,85)
A+B
1,9 kPa (1,76-2,23)
7,82 kPa(7,23-9,66)
20,19 kPa (18,234,88)
45,5 kPa (36,73-50,75)
A
2,1 kPa (2-2,3)
7, 6 kPa (7,2-8,9)
18,40 kPa (17-20)
43,85 kPa (40,9-45,8)
B
2,1 kPa (2-2,2)
7,65 kPa (7-8,1)
18,15 kPa (17,6-20,7)
43,65 kPa (40,9-47,8)
A+B
2,1 kPa (2-2,3)
7,6 kPa (7-8,9)
18,30 kPa (17-20,7)
43,7 kPa (40,9-47,8)
A
1,9 kPa (1,85-2,98)
7,02 kPa (6,987,09)
17,06 kPa (16,917,16)
35,93 kPa (35,5236,47)
B
1,97 kPa (1,932,05)
6,88 kPa (6,796,91)
17,05 kPa (16,8317,27)
36,47 kPa (36,3-36,82)
A+B
1,89 kPa (1,852,98)
6,94 kPa (6,797,09)
17,06 kPa (16,83-1727)
36,37 kPa ( 35,5236,82)
A
1,86 kPa (1,831,94)
7,24 kPa (7,157,28)
18,92 kPa (18,2519,52)
39,44 kPa (38,5-40,44)
B
1,92 kPa (1,851,94)
7,21 kPa (7,157,23)
18,47 kPa (17,9618,97)
38,67 kPa (37,8439,54)
A+B
1,95 kPa (1,831,94)
7,23 kPa (7,157,28)
18,75 kPa (17,9619,52)
38,9 kPa (37,84-40,44)
GE E9 Logiq SWE
curvilinear
Phantom 2 11,5kPa +/- 5%
Phantom 3 24,8 kPa +/- 5%
Phantom 4 46,3 kPa +/- 5%