The clinical importance of thyroid nodules rests with the need to exclude thyroid cancer. In the present study, we developed a modified Thyroid Imaging Reporting and Data System (TI-RADS) score using gray-scale ultrasound, contrast-enhanced ultrasound (CEUS) and shear-wave elastography (SWE) images to predict malignancy of thyroid nodules and compared this modified score system with the subjective scoring criteria based on the Thyroid Imaging Reporting and Data System (TI-RADS, 2017 edition). The results revealed that by using SWE and CEUS (enhanced pattern) to downgrade TI-RADS category 4 and 5 nodules, the malignancy rate for TI-RADS category 4 and 5 nodules increased from 47.6% with American College of Radiology (ACR) TI-RADS assessment alone to 49.4% with ACR TI-RADS combined with shear wave elastography (SWE) and CEUS (enhanced pattern). Likewise, by using the modified TI-RADS to adjust TI-RADS category 3 nodules, the malignancy rate for TI-RADS category 3 nodules increased from 13.9%-20.0%. The discriminating power for detection of malignancy of the variable score 2 (ACR TI-RADS + SWE + CEUS), with an area under the curve (AUC) of 0.899 (95% confidence interval [CI]: 86.1%-93.6%), was higher than that of score 1 (ACR TI-RADS), with an AUC of 0.862 (95% CI: 81.9%-90.6%; p > 0.05). With a point 4.5 as the optimal cutoff value, a score of 1 predicted malignancy with an accuracy of 75.6%, sensitivity of 85.0% and specificity of 71.6%. However, with a point 5.5 as the optimal cutoff value, a score of 2 predicted malignancy with an accuracy of 84.9%, sensitivity of 81.0% and specificity of 86.6%. The modified TI-RADS based on ACR TI-RADS + SWE + CEUS (enhanced pattern) could contribute to a reduction in the number of biopsies performed on benign nodules and the implementation of consistent follow-up in clinical practice.