Your new experience awaits. Try the new design now and help us make it even better

ORIGINAL RESEARCH article

Front. Endocrinol.

Sec. Cancer Endocrinology

Volume 16 - 2025 | doi: 10.3389/fendo.2025.1551108

This article is part of the Research TopicClinical prediction models in cancer through bioinformaticsView all 9 articles

Construction and validation of a predictive model for lymph node metastasis in patients with papillary thyroid carcinoma

Provisionally accepted
  • 1First Hospital of Shanxi Medical University, Taiyuan, China
  • 2Shanxi Medical University, Taiyuan, Shanxi Province, China

The final, formatted version of the article will be published soon.

To study the occurrence of lymph node metastasis in patients with papillary thyroid carcinoma (PTC) and construct a predictive model to assess its predictive performance. Methods We retrospectively analyzed the data of 432 patients with PTC.The least absolute shrinkage and selection operator (LASSO) was used to select the features, and multiple logistic regression was used to analyze the predictive factors.Multiple machine learning (ML) classification models are integrated to analyze and identify the optimal model, while Shapley additive exPlanations (SHAPs) are used for personalized risk assessment. A total of 125 patients from Changzhi Heping Hospital were included in an external validation set to evaluate the generalizability of our model.Results: Predictors of central lymph node metastasis (CLNM) included age, sex, maximum nodule diameter, margin, morphology, number of nodules, relationship between the nodule and the thyroid envelope, and coarse calcification. A logistic classification model was identified as the optimal model, with a test set area under the curve (AUC) value of 0.798. The validation results using external data were consistent, demonstrating the stability and generalizability of our model. Conclusion We established a logistic model using the SHAP method, which provides evidence for the ability of the SHAP method to predict lymph node metastasis and serves as a basis for personalized healthcare.

Keywords: Papillary thyroid cancer, lymphonodi cervicales metastasis, machine learning, Prediction model, Shap

Received: 24 Dec 2024; Accepted: 26 May 2025.

Copyright: © 2025 Hao, Zhang, Su and Liu. 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) or licensor 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.

* Correspondence: Li-ping Liu, First Hospital of Shanxi Medical University, Taiyuan, China

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.