Research progresses of radiomics in thyroid nodules
- VernacularTitle:甲状腺结节的影像组学研究进展
- Author:
Yi-xin LIU
1
;
Peng-yu LI
1
;
Zi-liang GUO
1
;
Zhi-hui LI
1
;
Wan-jun ZHAO
1
Author Information
- Publication Type:Journal Article
- Keywords: radiomics; thyroid nodules; thyroid cancer; ultrasonic image; machine learning
- From: Journal of Regional Anatomy and Operative Surgery 2025;34(3):267-271
- CountryChina
- Language:Chinese
- Abstract: The detection rate of thyroid nodules has gradually increased in recent years.Comprehensive and accurate preoperative evaluation and early identification of risk factors help doctors to choose treatment options and improve prognosis.Radiomics extracts quantitative features from medical images for evaluation of thyroid nodules through high-throughput mining of invisible image features,which has been widely used and has excellent performance in the identification of benign and malignant thyroid nodules,lymph node metastasis,extrathyroidal extension,molecular biological changes,recurrence and prognosis of thyroid cancer.However,there are also shortcomings such as large differences in performance among models from different institutions.This article reviews the application value,limitations and future development prospects of radiomics in the thyroid nodules,so as to provide new ideas for clinical practice and research.
