Development of a clinical decision model for thyroid nodules

Area of Science:

• Endocrinology
• Oncology
• Medical Informatics

Background:

• Thyroid nodules are common, with most being benign.
• Fine needle aspiration (FNA) has limitations, yielding indeterminate results in over 20% of cases.
• Many patients undergo diagnostic thyroid surgery for benign nodules.

Purpose of the Study:

• To develop and validate a predictive model for thyroid nodule malignancy risk.
• To reduce unnecessary diagnostic surgeries for benign thyroid nodules.

Main Methods:

• Prospective cohort study of 216 patients with thyroid nodules.
• Utilized ultrasound (US), electrical impedance scanning (EIS), and FNA cytology.
• Developed a Bayesian Network Analysis model with ten-fold cross-validation.

Main Results:

• The Bayesian model demonstrated high predictive accuracy for malignancy (AUC = 0.88).
• Positive predictive value was 83% and negative predictive value was 79%.
• Thyroid nodule size, FNA cytology, US, and EIS characteristics were significant predictors.

Conclusions:

• An integrated predictive model using Bayesian inference is clinically relevant for thyroid nodules.
• The model effectively predicts malignancy and warrants further validation in prospective trials.