Summary: | <p>Abstract</p> <p>Background</p> <p>The purpose of this study was to assess the advantages and disadvantages of 3D gray-scale and power Doppler ultrasound, including thin slice volume rendering (TSVR), applied for evaluation of thyroid nodules.</p> <p>Methods</p> <p>The retrospective evaluation by two observers of volumes of 71 thyroid nodules (55 benign, 16 cancers) was performed using a new TSVR technique. Dedicated 4D ultrasound scanner with an automatic 6-12 MHz 4D probe was used. Statistical analysis was performed with Stata v. 8.2.</p> <p>Results</p> <p>Multiple logistic regression analysis demonstrated that independent risk factors of thyroid cancers identified by 3D ultrasound include: (a) ill-defined borders of the nodule on MPR presentation, (b) a lobulated shape of the nodule in the c-plane and (c) a density of central vessels in the nodule within the minimal or maximal ranges. Combination of features provided sensitivity 100% and specificity 60-69% for thyroid cancer.</p> <p>Calcification/microcalcification-like echogenic foci on 3D ultrasound proved not to be a risk factor of thyroid cancer.</p> <p>Storage of the 3D data of the whole nodules enabled subsequent evaluation of new parameters and with new rendering algorithms.</p> <p>Conclusions</p> <p>Our results indicate that 3D ultrasound is a practical and reproducible method for the evaluation of thyroid nodules. 3D ultrasound stores volumes comprising the whole lesion or organ. Future detailed evaluations of the data are possible, looking for features that were not fully appreciated at the time of collection or applying new algorithms for volume rendering in order to gain important information. Three-dimensional ultrasound data could be included in thyroid cancer databases. Further multicenter large scale studies are warranted.</p>
|