Integration of radiomics ultrasound and TIRADS in diagnosis of thyroid nodules: a narrative review

BackgroundThe most popular technique for evaluating thyroid imaging is high-frequency ultrasonography; however, the TIRADS tool, intended to help with noninvasive risk assessment, has limitations in detecting thyroid cancerous nodules. The purpose of this article was to review the application of TIRADS in ultrasound radiomics and discuss its advantages and limitations.Main textA novel approach to medical picture processing called radiomics can help identify these nodules more precisely. Radiomics involves obtaining high-quality imaging for planning or diagnostic reasons, identifying a macroscopic tumor, extracting quantitative imaging features, and ranking the most informative findings according to prominence, independence, and reproducibility.ConclusionsRadiomics combined with TIRADS has demonstrated strong potential in enhancing the accuracy of thyroid nodule malignancy predictions, though challenges such as the need for larger, diverse datasets, and thorough validation persist. Incorporating clinical data, using deep learning models, and standardized imaging protocols could improve diagnostic precision, and further research will, therefore, be essential for its implementation in routine clinical practice.