Update on Thyroid Ultrasound: A Narrative Review

0
0
0

Update on Thyroid Ultrasound: A Narrative Review from Diagnostic Criteria to Artificial Intelligence Techniques

In recent years, the high incidence of thyroid nodules has become a significant concern. Ultrasound imaging plays a crucial role in detecting thyroid disease, but the management of thyroid ultrasound remains inconsistent. This review aims to explore the global guidelines of thyroid ultrasound and the application of artificial intelligence (AI) in this field.

1. Introduction

  • Thyroid Nodules Incidence: The increasing annual prevalence of thyroid nodules and the growing use of ultrasound have made it a major health issue.
  • Pathology and Treatment: Most thyroid nodules are benign, but a small percentage are malignant. Surgical resection is the primary treatment, but it has post-operative complications.
  • Importance of Standardization: Standardizing ultrasound diagnosis and management is crucial to improve accuracy and reduce unnecessary invasive tests.

2. Guidelines and Consensus on Thyroid Ultrasound

  • Common Diagnostic Standards: Various guidelines, such as the 2016 AACE/ACE/AME Guidelines, 2015 ATA Guidelines, and 2017 ACR TI-RADS, have been developed.
  • Suspicious Ultrasound Features: Solid nodule structure, hypoechogenicity, taller-than-wide shape, irregular margin, microcalcification, and invasion of surrounding tissue are associated with malignancy.
  • Assessment of Metastatic Cervical Lymph Nodes: Ultrasound is an important method for detecting lymphatic metastasis, with signs such as microcalcification, cystic degeneration, and peripheral blood flow.
  • Assessment of Nodule Blood Flow: The role of nodule blood flow in diagnosis is controversial, with different guidelines having different views.
  • Indications for FNA: FNA is widely used, but its application indications vary among guidelines.
  • Other Ultrasound Techniques: Ultrasound elastography and three-dimensional ultrasound are new technologies, but their acceptance among guidelines is limited.
  • Diagnostic Criteria for Special Populations: Special criteria for children and pregnant women have been developed due to their unique characteristics.

3. Image Classification Techniques in Thyroid Ultrasound

  • Conventional Ultrasound Limitations: Conventional ultrasound has limitations such as susceptibility to patient position and imaging artifacts, and subjectivity in diagnosis.
  • AI in Medical Imaging: AI, especially image classification, offers a new approach to improve diagnostic accuracy and efficiency.
  • Feature Extraction: Texture features from ultrasound images are important for diagnosis.
  • Pre-processing: Image pre-processing, such as segmentation, is essential for data consistency and accuracy.
  • Image Classification Algorithm Model: Various classifiers, such as artificial neural network (ANN) and support vector machine (SVM), are used for thyroid ultrasound image classification.

4. Limitations in Clinical Application

  • Data Collection and Validation: Sufficient data are needed for system validation, and the collection process is time-consuming.
  • Physician Training: Physicians need time to familiarize with complex formulas and algorithms.
  • Database Adaptation: One algorithm model may only be suitable for a specific database, and adjustments are needed for different databases.
  • Subjective Problems: Unavoidable subjective problems in clinical application, such as misdiagnosis due to atypical sections, need to be addressed.

5. Conclusions

  • Standardization and Technological Innovation: Standardization and technological innovation are key to improving thyroid ultrasound diagnosis.
  • AI as a Solution: AI can compensate for the limitations of traditional ultrasound and improve diagnostic accuracy.
  • Future Trends: AI will become the main trend in the development of thyroid ultrasound diagnosis in the future.

In conclusion, this review provides a comprehensive overview of thyroid ultrasound guidelines and the application of AI. It highlights the importance of standardization and technological innovation in improving thyroid ultrasound diagnosis. The use of AI techniques shows great potential in compensating for the limitations of traditional ultrasound and improving diagnostic accuracy and efficiency. However, further research and development are needed to overcome the limitations in clinical application and to fully realize the potential of AI in thyroid ultrasound diagnosis.

doi.org/10.1097/CM9.0000000000000346

Was this helpful?

0 / 0