Artificial intelligence - based ultrasound elastography for disease evaluation - a narrative review

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Biomedical Engineering

Background:

• Ultrasound elastography (USE) offers tissue stiffness information, complementing conventional ultrasound.
• USE is noninvasive and radiation-free, improving diagnostic capabilities.
• Operator dependence and observer variability can limit diagnostic accuracy in visual USE assessments.

Purpose of the Study:

• To review the fundamental concepts of USE and AI techniques for radiologists.
• To introduce AI applications in USE for lesion detection, segmentation, classification, and prognosis.
• To discuss current challenges and future directions for AI in USE.

Main Methods:

• Overview of AI techniques relevant to medical image analysis.
• Review of AI applications in USE across liver, breast, thyroid, and other organs.
• Focus on AI for lesion detection, segmentation, ML-assisted classification, and prognosis prediction.

Main Results:

• AI demonstrates enhanced diagnostic performance in applied USE for disease evaluation.
• AI integration aims to provide more objective and accurate diagnoses.
• AI in USE shows potential for improved lesion characterization and patient outcome prediction.

Conclusions:

• AI holds significant potential to overcome limitations of manual USE interpretation.
• AI-assisted USE can lead to more objective, accurate, and intelligent diagnostic outcomes.
• Further research into AI applications and challenges in USE is warranted for clinical advancement.