Prediction model based on MRI morphological features for distinguishing benign and malignant thyroid nodules

Motivation: The low specificity of many Thyroid Imaging Reporting and Data Systems (TI-RADSs) lead to a large number of unnecessary biopsies. Goal(s): This study developed and validated a predictive model based on MRI morphological features to improve the specificity. Approach: A retrospective analysis was conducted on 825 thyroid nodules pathologically confirmed postoperatively. Univariate and multivariate logistic regression was used to obtain β coefficients, construct predictive models and nomogram incorporating MRI morphological features in the training cohort, and validated in a validation cohort. Results: Compared with the TI-RADSs, predictive models have better specificity along with a high sensitivity and can reduce unnecessary biopsies. Impact: predictive models have better specificity along with a high sensitivity and may avoid numerous invasive needle biopsies