From January to December 2013,
thyroid SWE was performed in 57 subjects using a 4-15 MHz linear transducer on commercially available clinical shear wave ultrasound system (Aixplorer,
SuperSonic Imagine,
Aix en Provence,
France).
SWE was performed on different thyroid tissue types as follows: 38 normal thyroid glands (Norm),
27 benign hyperplastic nodules (MNG) and 5 papillary carcinomas (PapCa).
Subjects were either normal volunteers or patients referred for thyroid US assessment,
and the final diagnosis of thyroid nodules was based on sonographic,
cytological ± histological results as per the patient’s routine clinical work-up.
Informed written consent was obtained from all subjects and local ethic committee approval had been obtained for this prospective study.
This study was supported by a grant from the research grants council of the Hong Kong Special Administrative Region,
China (grant Chinese University of Hong Kong.
Project ID 2140771).
SWE cineloops,
each lasting 10 seconds to allow for SWE image stabilization,
were acquired for normal parenchyma or nodules,
which were exposed to four levels of static compression.
Static compression was applied manually by an operator via the US linear transducer placed on the skin surface with an intervening US gel layer.
The 4 levels were based on the operator’s experience and by the following features on the corresponding grayscale images as follows:
1) MINIMAL,
a thick US coupling gel layer between the transducer and skin surface
2) MILD,
a thin US gel layer present
3) MODERATE,
light compression of the skin over the thyroid gland
4) HIGH,
substantial compression of the skin over the gland.
In addition,
the level of compression was quantified in terms of percentage axial (vertical) strain of the thyroid gland relative to the minimal compression level by referencing to the corresponding gray-scale images [Fig1].
All levels of compression were tolerated well by subjects.
No dynamic or cyclical compressions were applied during SWE.
Following SWE acquisition,
circular electronic regions-of-interest (ROIs) were placed on representative static SWE images; ROIs were placed either within the entire nodule or a 1cm diameter ROI was placed within normal parenchyma.
SWE tissue stiffness data were displayed automatically and the mean value of the ROI was recorded [Fig.2].
Up to four SWE readings were obtained at each compression level and an averaged SWE value was used for subsequent analysis.
SWE stiffness indices were compared between compression levels and between tissue types using Students t-tests,
with a p<0.05 indicating statistical significance.