Differentiating Benign vs Malignant Thyroid Nodules Using DECT Radiomics

Overview

This study developed a dual-energy CT (DECT) radiomics model combining quantitative iodine parameters and traditional radiomic features to differentiate benign from malignant thyroid nodules preoperatively. The model demonstrated improved diagnostic accuracy over conventional imaging techniques, supporting enhanced clinical decision-making.

Background

Thyroid nodules are common, with a malignancy rate of 10-15%, necessitating accurate differentiation to guide treatment. Ultrasound is the first-line imaging but has limitations including operator dependency and overlapping features between benign and malignant nodules. MRI offers superior soft tissue contrast but is time-consuming and limited in detecting calcifications. DECT provides advanced imaging with virtual monoenergetic and material decomposition images, enabling quantitative assessment of iodine concentration and effective atomic number, which may improve lesion characterization. Radiomics further enhances diagnostic accuracy by extracting high-throughput features reflecting tumor heterogeneity.

Data Highlights

A total of 215 patients with thyroid nodules were retrospectively analyzed, divided into training (150) and test (65) cohorts. DECT images were acquired at arterial and venous phases with iodine-based material decomposition and virtual monoenergetic images at 40, 70, and 100 keV. Quantitative parameters included iodine concentration (IC) and effective atomic number (Zeff). Radiomic features were extracted from these images to develop predictive models.

Key Findings

• DECT quantitative iodine parameters (IC) and Zeff values significantly differed between benign and malignant thyroid nodules.
• Radiomic features extracted from DECT images captured tumor heterogeneity correlated with malignancy.
• Combining quantitative DECT parameters with radiomic features improved differentiation accuracy compared to either method alone.
• The developed DECT radiomics model showed higher diagnostic performance than conventional CT and ultrasound.
• Material decomposition images and virtual monoenergetic images provided complementary information enhancing lesion characterization.

Clinical Implications

Incorporating DECT quantitative iodine metrics and radiomics into preoperative evaluation can improve differentiation of benign and malignant thyroid nodules, potentially reducing unnecessary invasive procedures. This approach offers a noninvasive, objective tool to aid clinicians in treatment planning and risk stratification. DECT imaging may serve as a valuable adjunct to ultrasound, especially in complex cases.

Conclusion

The integration of DECT quantitative parameters with radiomic features provides a robust, noninvasive method for distinguishing benign from malignant thyroid nodules, enhancing diagnostic accuracy and supporting personalized clinical management.

Related Resources & Content

1. American Thyroid Association Guidelines -- Thyroid Nodule Management
2. Liu et al. 2020 -- DECT Quantitative Assessment of Lymph Nodes
3. Zhou et al. 2021 -- Radiomics Predicting Lymph Node Metastasis in Thyroid Cancer