I. Introduction
Even though nodules are the most common type of benign lesions in thyroid, and the most frequently studied [1], their malign incidence varies from 5 to 15% [2]. Malignancy risk is initially established by high resolution thyroid ultrasound (TU), a non-ionizing image procedure with a reliable visualization power [3]. Common TU examination includes description of the nodule in terms of composition (cystic, solid, or mixed), echogenicity, shape, margins and presence of echogen focuses [4], as defined by the TIRADS evaluation system [5]. This protocol establishes a risk of malignancy according to a pre-defined scale, from TI-RADS 1 (benign) to TI-RADS 5 (probably malignant) [6]. Nevertheless, final thyroid nodule (TN) recommendation (e.g., repeat procedure vs. surgery) depends on the cytology of needle aspiration, since TI-RADS features may overlap both benign and malignant nodules [7], while inter-and intra-observer variability may lead to low thyroid cancer diagnosis specificity, i.e., from 27% to 78.0% [4]. Yet Fine Needle Aspiration Biopsy (FNAB) is currently the most reliable procedure to confirm or reject TU estimations [1], this is still susceptible to error by the limited specimen collection or observer experience [8]. Inadequate FNAB is a frequent limitation of the cytopathological analysis [9], with highly variable reported error rate, i.e., between 1% to 20% [10]. Therefore, the unique TN malignancy confirmation strategy relies on the evaluation of complete or partially extracted gland, which is a drawback on clinical patient management.