Nomogram Combining Radiomics With the American College of Radiology Thyroid Imaging Reporting and Data System Can...

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Thyroid nodules require accurate malignancy prediction.
• Current guidelines like ACR TI-RADS have limitations in diagnostic performance.

Purpose of the Study:

• To develop and validate a nomogram integrating B-mode ultrasound (BMUS) radiomics and ACR TI-RADS scores.
• To enhance the prediction of malignant thyroid nodules and improve guideline performance.

Main Methods:

• Retrospective collection of 451 pathologically proven thyroid nodules.
• Development of an ACR-Rad nomogram combining BMUS radiomics score (Rad-Score) and ACR TI-RADS score (ACR-Score).
• Validation of the nomogram's discrimination, calibration, and clinical utility.

Main Results:

• The ACR-Rad nomogram demonstrated superior predictive performance (AUC 0.877 training, 0.864 validation) compared to ACR-Score alone.
• Significant improvements in AUC, NRI, and IDI were observed for both senior and junior radiologists.
• The nomogram showed increased specificity, accuracy, PPV, and reduced unnecessary FNA rates.

Conclusions:

• The ACR-Rad nomogram is a valuable tool for predicting malignant thyroid nodules.
• It effectively improves the performance of ACR TI-RADS for radiologists of all experience levels.