Additional diagnostic value of ratio indices of quantitative contrast-enhanced ultrasound parameters in small solid C-TIRADS 4 thyroid nodules

TL;DR

This study shows that using ratio indices from contrast-enhanced ultrasound improves the accuracy of diagnosing small thyroid nodules classified as C-TIRADS 4.

Contribution

The study introduces ratio indices of quantitative CEUS parameters as a novel diagnostic tool for thyroid nodule classification.

Findings

The logistic model combining C-TIRADS and CEUS ratio indices achieved high diagnostic accuracy (AUC 0.935 in training, 0.910 in validation).

Ratio indices outperformed individual parameters in differentiating benign from malignant thyroid nodules.

The model demonstrated high sensitivity and specificity in both training and validation cohorts.

Abstract

To investigate the efficacy of contrast-enhanced ultrasound (CEUS) parameters, particularly ratio indices of quantitative CEUS parameters, for differentiation of small solid C-TIRADS 4 thyroid nodules.

235 small solid C-TIRADS 4 thyroid nodules with determinate pathological results, including 175 nodules in the training cohort and 60 nodules in the validation cohort were retrospectively evaluated. The ratio indices of the internal tissue to peripheral tissue and the internal tissue to healthy tissue of quantitative parameters were calculated. In the training cohort, the meaningful quantitative ratio indices with an AUC > 0.7 and qualitative parameters were further included in multivariate regression analysis. The diagnostic efficacy of the logistic model was evaluated.

In single-factor analysis, C-TIRADS, enhancement degree, mTTI ratio (L/P), TTP ratio (L/P), WiR ratio (L/P), WoR ratio (L/P) and TTP ratio (L/H) were significant parameters for differentiation of thyroid nodules (P < 0.05). The multifactor analysis showed that there was significant difference in C-TIRADS, mTTI ratio (L/P), TTP ratio (L/P), WiR ratio (L/P), WoR ratio (L/P) and TTP ratio (L/H) between two groups (P < 0.05). The logistic model was generated, and the AUC, sensitivity, specificity, and accuracy of the training cohort were 0.935(95% CI: 0.888–0.967),85.71%, 88.57%, and 86.86%, respectively. The logistic model demonstrated significantly higher diagnostic performance compared to individual parameters (P < 0.001). In the validation cohort, the diagnostic model had an AUC of 0.910,sensitivity of 87.9%, specificity of 92.6%, and accuracy of 90.25%.

Ratio indices of quantitative parameters have high diagnostic values in differentiating small solid thyroid nodules. Combining C-TIRADS with both qualitative and quantitative CEUS parameters enhances the diagnostic accuracy of malignant thyroid nodules.

Figures