Ultrasound-based diagnostic classification for solid and partially cystic thyroid nodules

Area of Science:

• Radiology
• Oncology
• Endocrinology

Background:

• Differentiating benign from malignant thyroid nodules using ultrasound (US) remains challenging.
• A standardized US-based classification system is needed for improved diagnostic accuracy.

Purpose of the Study:

• To evaluate the diagnostic efficacy of a prospectively designed US classification system.
• To assess the system's performance in classifying solid thyroid nodules and partially encapsulated tumors (PCTNs).

Main Methods:

• A prospective study of 1289 thyroid nodules in 1036 patients.
• Nodules were classified in real-time using a 5-category US system (benign to malignant).
• Diagnostic efficacy was determined by comparing US classifications with histopathologic results from 505 surgically resected nodules.

Main Results:

• The US classification system demonstrated high diagnostic performance for both solid nodules and PCTNs.
• Sensitivity, specificity, and accuracy rates were reported for solid nodules and PCTNs, with specific values provided for key metrics.
• The system showed good differentiation capabilities, with notable performance metrics for solid nodules (e.g., 87.5% accuracy) and PCTNs (e.g., 81.5% accuracy).

Conclusions:

• The developed US-based classification system offers valuable guidance for managing thyroid nodules.
• The system aids clinicians in distinguishing between benign and malignant thyroid lesions.
• This classification approach can potentially improve patient management strategies for thyroid nodules.