Radiomics in Differentiated Thyroid Cancer and Nodules: Explorations, Application, and Limitations

Area of Science:

• Medical Imaging
• Radiology
• Oncology
• Machine Learning

Background:

• Radiomics extracts quantitative imaging features beyond visual interpretation.
• These features can be used with machine learning for classification and prediction.
• The role of radiomics in differentiated thyroid cancer (DTC) requires further clarification due to inconsistent findings.

Purpose of the Study:

• To review and summarize the literature on the classification and prediction performance of radiomics in DTC.
• To evaluate radiomics' capacity for predicting lymph node metastasis, distant metastasis, extrathyroidal extension, disease-free survival, and BRAF mutation.
• To assess radiomics' utility in differentiating benign from malignant nodules in DTC patients.

Main Methods:

• Systematic literature review of radiomics studies in DTC.
• Analysis of radiomics' application across various imaging techniques.
• Focus on prediction and classification tasks including metastasis, BRAF mutation, and nodule characterization.

Main Results:

• Radiomics shows potential for objective tumor evaluation in DTC.
• Inconsistent results exist regarding radiomics' classification and prediction value in DTC.
• Specific applications include predicting lymph node metastasis, distant metastasis, extrathyroidal extension, disease-free survival, and BRAF mutation status.

Conclusions:

• Radiomics offers a promising tool for improving clinical decision-making in DTC.
• Further research is needed to standardize radiomics applications and overcome limitations.
• The review highlights radiomics' potential utility for DTC patient management.