Size-Specific Predictors for Malignancy Risk in Follicular Thyroid Neoplasms: Machine Learning Analysis
Xin Li1, Wen-Yu Yang2, Fan Zhang3
1Department of General Surgery, Peking University Third Hospital, Beijing, China.
JMIR Cancer
|July 11, 2025
Summary
Predicting follicular thyroid neoplasm (FTN) malignancy risk is challenging. This study identified size-specific predictors, including calcification and nodule appearance, to aid preoperative diagnosis of FTNs.
Area of Science:
- Endocrinology
- Oncology
- Radiology
Background:
- Distinguishing benign from malignant follicular thyroid neoplasms (FTNs) poses a surgical challenge, especially for small tumors.
- Preoperative risk stratification for FTNs is crucial for surgical planning and patient management.
Purpose of the Study:
- To identify preoperative predictors for malignancy risk in follicular thyroid neoplasms (FTNs).
- To investigate size-specific predictors for malignancy risk in FTNs, differentiating between small and large tumors.
Main Methods:
- A retrospective cohort study included 1494 patients with follicular thyroid adenoma (FTA) or carcinoma (FTC).
- Follicular thyroid neoplasms (FTNs) were categorized as small (<3.0 cm) or large (≥3.0 cm) based on diameter.
- Machine learning identified key predictors from demographic, sonographic, and hormonal variables, with odds ratios calculated for malignancy risk.
Main Results:
- Small FTNs with macrocalcification, peripheral calcification, or in younger patients had higher malignancy risk.
- Large FTNs with a nodule-in-nodule appearance showed increased malignancy risk.
- Lower thyroid-stimulating hormone levels and larger mean diameter were associated with malignancy risk in both small and large FTNs.
Conclusions:
- Size-specific predictors for FTN malignancy risk were identified.
- Stratified prediction based on tumor size is essential for improving the accuracy of preoperative diagnosis of follicular thyroid neoplasms.
