Thyroid Nodule Malignancy Risk Prediction Model from Multicenter Study

631 thyroid nodules from 478 patients at Yueyang People’s Hospital and 193 nodules from 144 patients at Hunan Cancer Hospital were used to develop a malignancy risk prediction model. Wei Liu published a logistic regression model achieving an AUC of 0.929 on external validation. The study aims to provide a clinically interpretable tool for assessing C-TIRADS 3–4 thyroid nodules.

Use Cases

• Training machine learning models for thyroid nodule malignancy risk prediction based on ultrasonographic features and laboratory indicators.
• Benchmarking logistic regression against random forest, SVM, XGBoost, and LightGBM models for clinical prediction tasks.
• Building individualized risk assessment tools, such as nomograms, for clinical decision support.
• Evaluating the stability and interpretability of predictive models in external validation cohorts.

Strengths

• The dataset is derived from a multicenter study involving 824 total nodules, providing a foundation for model development.
• The logistic regression model demonstrated strong external validation performance with an AUC of 0.929 (95% CI: 0.887–0.965).
• The study includes an independent external validation set from a different hospital (Hunan Cancer Hospital), enhancing robustness.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is 798.5 KB, indicating a limited scope of data.

Provenance

Source

Yueyang People’s Hospital and Hunan Cancer Hospital.

Collection Method

Multicenter retrospective study.

Freshness

Last updated 2026-05-18 04:17:02; freshness should be verified.

Geography

China (Hunan province).