Evaluating the role of microvascular flow imaging in benign–malignant differentiation of solid thyroid nodules

Aim: This study aimed to investigate the diagnostic value of microvascular findings detected by Superb Microvascular Imaging (SMI) in differentiating benign from malignant solid thyroid nodules, and to assess the potential advantages of SMI over conventional Doppler techniques as well as its added value when combined with grayscale ultrasound. Material and Method: Examinations performed between 2023 and 2025 were retrospectively reviewed in the imaging archive of Balıkesir University Hospital. Ninety nodules were included (23 malignant, 67 benign). Grayscale features and TI RADS scores were recorded. On RDUS/PDUS and SMI, vascular morphology, distribution, patterns, and Adler grade were evaluated. Cut-off values were calculated for the vascular index (VI) on SMI and for shear wave elastography (SWE). Diagnostic performance was assessed using sensitivity, specificity, PPV, and AUC. Results: On SMI, "crab claw–like," "root hair–like," and "single penetrating vessel" patterns were observed in 86.9% of malignant nodules (p<0.001) and yielded 87% sensitivity and 97% specificity for benign–malignant discrimination (p<0.001). In contrast, "wheel-like," "tree-like," "diffuse punctate," and "peripheral vascularity" patterns were present in 97% of benign nodules (p<0.001). Combining SMI pattern assessment with B-mode provided additional diagnostic benefit (specificity 97.0%, PPV 90.9%, AUC 0.948; p<0.001). Flow morphology, distribution, and Adler grades did not differ significantly between RDUS/PDUS and SMI for malignancy detection (p>0.05). For VI, the cut-off was ≥10.95 (sensitivity 69.6%, specificity 53.7; p=0.285). SWE showed a cut-off ≥2.85 m/s (25–26 kPa) with 95.5% sensitivity and 55.2% specificity (p<0.001). Conclusion: SMI depicts microvascular architecture more effectively than conventional Doppler and is more useful when integrated with B-mode. Keywords: Thyroid nodule, Superb Microvascular Imaging, crab-claw–like pattern, root-hair like pattern, microflow characteristics