Ultrasound is a common and non-invasive tool,
used to observe the morphological features of breast lesions,
often used as a complementary technique for mammography.
US has two major disadvantages: many solid breast lesions are indeterminate at last and required biopsy,
because of US was not adequately specific in differentiating benign from malignant solid breast masses,
and it is a technique which depends on the operator experience.
Elastography is a new advance in US technology developed in recent years,
with the aim to evaluate nodule elastic features and relative elastic distribution between lesion and surrounding normal tissue.
[1,2] This advanced technique was used to aid differential diagnosis because it can yield details not only about morphologic features but also about lesion hardness.
Strain is affected by both physiological and pathological processes that cause structural changes [3],
with harder tissues (e.g.
breast cancer) causing decreased strain compared to softer tissues [4].
Free-hand elastography images are obtained through gradual manual compression of examined tissue that allows a real-time elastogram as a color image superimposed on B-mode image.
There has been concern about the reproducibility of these techniques in clinical practice,
especially in strain,
in which the compressive force applied needs to be appropriate.
Elastographic free-hand method,
based on slight manual compression/decompression with a conventional transducer to displace the tissue under investigation,
allows to calculate tissue elasticity in real time,
monitoring the movement determined in the tissue.
Tissue elasticity can be assessed using qualitative (color map) or quantitative (strain ratio) parameters and several clinical studies have been reported in the literature. [5,6,7,8]
Color map,
a qualitative variable based on the 5-point scoring method introduced by Ueno-Itoh,
was the first elastographic parameter used.
Nodules can be categorized from 1 to 5,
based on the degree and distribution of strain in the lesion and in the surrounding normal breast tissue.
[4]
The strain ratio was subsequently introduced as a semi-quantitative measurements and several clinical studies showed that it was one of the most useful tool to differentiate benign and malignant breast lesions.
[9] Until now strain ratio is calculated placing manually two ROIs (Regions Of Interest) in the target lesion and in the surrounding normal fat tissue (considered as reference) respectively,
resulting in significant inter-observer variability.
[10,11]
In recent years,
different kinds of semi-automatic algorithms were introduced to reduce operator dependence.
The purpose of the present study is to compare manual and a semi-automatic algorithm (Assist Strain Ratio,
ASR) to place target and reference ROIs,
considering inter-observer variability.
All breast lesions were subjected to core biopsy or fine needle aspiration cytology (FNAC) for histological diagnosis as reference standard.