A machine learning model for predicting bone and/or lung metastasis in differentiated thyroid carcinoma - enhancing precision in risk stratiﬁcation

Aug 5, 2026

Differentiated thyroid cancer has an overall favorable prognosis, but a subset develops aggressive disease with distant metastases, especially to bone and lung, which markedly reduces survival. Existing prediction approaches often depend on limited clinical variables and show suboptimal accuracy and generalizability. This study develops a machine learning model using SEER data and external validation to improve risk stratification and enable earlier intervention in high-risk patients.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Background

Methods

Results

Conclusion

Key predictive variables

Model performance and validation

TYPE Original Research PUBLISHED 08 September 2025 DOI 10.3389/fendo.2025.1528392 OPEN ACCESS EDITED BY Natarajan Bhaskaran, Saveetha Medical College & Hospital, India REVIEWED BY Malgorzata Troﬁmiuk-Muldner, Jagiellonian University Medical College, Poland Wencai Liu, Shanghai Jiao Tong University, China *CORRESPONDENCE Shengyin Liao [jxmu_2005@163.com](jxmu_2005@163.com) RECEIVED 14 November 2024 ACCEPTED 25 August 2025 PUBLISHED 08 September 2025 CITATION Huang L, He L, Chen R and Liao S (2025) A machine learning model for predicting bone and/or lung metastasis in differentiated thyroid carcinoma: enhancing precision in risk stratiﬁcation. Front. Endocrinol. 16:1528392 . doi: 10.3389/fendo.2025.1528392 COPYRIGHT © 2025 Huang, He, Chen and Liao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. A machine learning model for predicting bone and/or lung metastasis in differentiated thyroid carcinoma: enhancing precision in risk stratiﬁcation Libin Huang 1,2, Limei He 1,2, Ru Chen 3 and Shengyin Liao 1,2* 1 Department of Medical Oncology, The First Hospital of Putian, Teaching Hospital, Fujian Medical University, Putian, Fujian, China, 2The School of Clinical Medicine, Fujian Medical University, Fuzhou, Fujian, China, 3 Department of General Surgery, The First Afﬁliated Hospital of Nanchang University, Nanchang, China Background: Differentiated thyroid cancer (DTC) incidence is rapidly rising worldwide. While most cases have a favorable prognosis, a subset of patients develop aggressive disease with distant metastases, particularly to the bone and lung, which signiﬁcantly worsens outcomes. Current prediction models are limited in accuracy, often relying on basic clinical factors. This study aims to develop a machine learning model to improve prediction of bone and lung metastasis in DTC, enhancing risk stratiﬁcation and early intervention. Methods: Using the SEER database, we developed several machine learning models—including XGBoost, Random Forest, Gradient Boosting Machine, Logistic Regression, Naive Bayes, and Classiﬁcation and Regression Trees (CART)—to predict bone and lung metastasis risk in DTC patients. LASSO regression was applied to select key predictive variables, and SMOTE was used to address data imbalance. The model’s generalizability was evaluated using an external validation cohort from China. Results: The XGBoost model demonstrated the highest performance, achieving an AUC of 0 .988. Key predictive variables identiﬁed and included in the model were tumor size, radiation therapy, surgical interventions, histologic types, T and N stages, laterality, race, and household income. SHAP analysis conﬁrmed the importance of these variables, with tumor size, radiation, and surgery emerging as primary predictors. In the external validation cohort, the model achieved an AUC of 0 . 866, indicating reliable predictive capability across clinical settings. Conclusion: This model accurately predicts bone and lung metastasis risk in DTC, offering valuable clinical utility for risk stratiﬁcation and supporting early intervention strategies to improve outcomes in high-risk patients. KEYWORDS thyroid cancer, bone metastasis, lung metastasis, SEER, machine learning Frontiers in Endocrinology 01 [frontiersin.org](frontiersin.org) 1 Induction Thyroid cancer (TC) is one of the most rapidly increasing malignancies globally, with a notable rise in incidence over the past few decades (1, 2).Differentiated thyroid cancer (DTC) is the most common type of malignant thyroid tumor, origina

FAQs

What clinical problem does the study address?

The study targets the need for more accurate prediction of bone and lung metastasis risk in differentiated thyroid cancer to improve early identification of high-risk patients.

Which machine learning methods are used to build the model?

Models include XGBoost, Random Forest, Gradient Boosting Machine, Logistic Regression, Naive Bayes, and CART, with LASSO for variable selection and SMOTE for handling imbalance.

How well does the best model perform in training and external validation?

The XGBoost model achieved an AUC of 0.988, and in an external validation cohort it achieved an AUC of 0.866, supporting reliable predictive capability.

Aug 5, 2026

Upload by Evangeline