Resumen
Low specificity of ultrasound in detecting thyroid cancer warrants the development of new noninvasive modalities for the optimal characterization of thyroid nodules. Here, we present a new ultrasound-based technique, high-definition microvasculature imaging (HDMI) that provides quantitative measures of tumor microvasculature morphological features as new imaging biomarkers. This technique utilizes vessel enhancement filtering, morphological filtering, and vessel segmentation, which enable extraction of vessel morphological features including tortuosity, vessel density, diameter, Murray?s deviation, microvessel fractal dimension, bifurcation angle, number of branch points, and vessel segments. Without the help of contrast agents, through the utilization of HDMI on patients with suspicious thyroid nodules, we were able to resolve tumor microvessels at size scales of a few hundred microns. We further showed that analysis of tumor vessel morphological parameters could detect thyroid malignancy with high sensitivity and specificity. These findings provide a translational rationale for the clinical implementation of quantitative HDMI for thyroid cancer detection.