Objective:

To develop a fusion model using B-mode ultrasound (BMUS) and superb microvascular imaging (SMI) to predict central lymph node metastasis (CLNM) in patients with papillary thyroid carcinoma (PTC), addressing limitations of current diagnostic methods.

Key Findings:

• The combined use of BMUS and SMI significantly improves the diagnostic accuracy for detecting CLNM in PTC, potentially impacting surgical decision-making.
• The integration of radiomics and ViT features outperforms standalone models in predicting CLNM, highlighting the importance of advanced imaging techniques.
• Utilizing 2.5D imaging mitigates overfitting while capturing more information than traditional 2D methods, enhancing diagnostic reliability.

Interpretation:

The study demonstrates that advanced imaging techniques and AI can enhance the preoperative assessment of CLNM in PTC, potentially leading to better surgical outcomes.

Limitations:

• The study is retrospective and may be subject to selection bias, particularly in patient selection and image quality.
• The reliance on high-quality ultrasound images may limit generalizability to other settings, necessitating further validation.
• The model's performance needs validation in larger, diverse populations to confirm its applicability.

Conclusion:

The innovative fusion model combining BMUS and SMI with advanced AI techniques shows promise in improving the prediction of CLNM in PTC, warranting further research and validation in diverse clinical settings.