MRI-based model to predict preoperative extrathyroidal extension in papillary thyroid carcinoma

Area of Science:

• Radiology and Oncologic Imaging
• Medical Imaging Analysis
• Thyroid Cancer Research

Background:

• Accurate preoperative assessment of extrathyroidal extension (ETE) in papillary thyroid carcinoma (PTC) is crucial for surgical planning and patient outcomes.
• Existing imaging modalities may have limitations in precisely predicting the extent of PTC invasion.
• Developing novel predictive models can enhance preoperative staging and guide treatment strategies.

Purpose of the Study:

• To develop and validate a predictive model for preoperative ETE in PTC using MRI features.
• To identify key MRI and clinical predictors of ETE in PTC.
• To assess the clinical utility and performance of the developed predictive nomogram.

Main Methods:

• Retrospective analysis of 140 PTC cases, divided into training and validation cohorts.
• Evaluation of MRI features including T2-weighted imaging, contrast-enhanced MRI, and diffusion-weighted imaging (DWI).
• Logistic regression analysis to identify independent predictors and develop a nomogram; performance assessed using ROC curves and calibration tests.

Main Results:

• Age, protrusion value, and apparent diffusion coefficient_Brightest_rate (ADC_Best_rate) were identified as independent predictors of ETE.
• The nomogram demonstrated strong discrimination and calibration in both training (AUC=0.826) and validation (AUC=0.805) cohorts.
• ADC_Best_rate showed superior predictive performance compared to other ADC metrics and varied in accuracy based on gender.

Conclusions:

• A nomogram incorporating age, protrusion value, and ADC_Best_rate effectively predicts preoperative ETE in PTC.
• This model aids surgeons in optimizing therapeutic decision-making for PTC patients.
• ADC_Best_rate shows promise as a functional imaging biomarker for ETE assessment in PTC.