[Diagnosis value of thyroid imaging reporting and data system in thyroid nodules]

Area of Science:

• Radiology
• Oncology
• Endocrinology

Background:

• Thyroid nodules are common, requiring accurate diagnostic tools.
• The Thyroid Imaging Reporting and Data System (TI-RADS) is a standardized classification system for ultrasound-based assessment.
• Evaluating the diagnostic performance and interobserver reliability of TI-RADS is crucial for clinical application.

Purpose of the Study:

• To assess the diagnostic value of the proposed TI-RADS classification for thyroid nodules.
• To investigate the interobserver variability among sonographers using TI-RADS for ultrasound analysis.

Main Methods:

• Retrospective review of 667 thyroid nodules from 532 patients with pathological confirmation.
• Independent assessment of ultrasound images by 4 experienced sonographers using TI-RADS criteria.
• Calculation of diagnostic sensitivity, specificity, PPV, NPV, accuracy, and interobserver agreement using kappa statistics.

Main Results:

• Overall diagnostic performance: sensitivity 94%, specificity 79%, NPV 97%, accuracy 81%.
• Positive Predictive Values (PPVs) for TI-RADS categories 4 and 5 were 55% and 92%, respectively.
• Substantial interobserver agreement was found for categories 3, 4B, and 5 (Kappa=0.62), with moderate agreement for category 4A (Kappa=0.57).

Conclusions:

• TI-RADS classification exhibits high diagnostic value in differentiating benign from malignant thyroid nodules.
• Substantial interobserver agreement among experienced sonographers supports its reliability.
• TI-RADS criteria are valuable for guiding appropriate patient management and treatment planning.