Medical image retrieval: past and present

Area of Science:

• Medical Informatics
• Radiology Informatics
• Digital Imaging

Background:

• The proliferation of Picture Archiving and Communication Systems (PACS) in healthcare settings has led to a significant increase in medical imaging data volume.
• Managing these large and complex imaging databases necessitates the development of effective image retrieval solutions.

Purpose of the Study:

• To review the historical evolution and current status of medical image retrieval systems.
• To identify the limitations of existing text-based and content-based image retrieval methods.
• To propose a direction for future advancements in medical image retrieval.

Main Methods:

• A comprehensive literature review was conducted on medical image retrieval systems.
• Analysis of past developments and the present state of both text-based and content-based retrieval approaches.
• Evaluation of the effectiveness of current systems in handling large-scale medical image data.

Main Results:

• Traditional text-based retrieval systems have limitations in managing the complexity of medical images.
• Content-based image retrieval (CBIR) offers improvements but still faces challenges in accuracy and efficiency.
• Existing systems struggle to fully leverage the rich information contained within medical images.

Conclusions:

• Intelligent content-based retrieval systems are crucial for enhancing medical image management.
• The integration of semantic systems with content-based retrieval is essential for more effective and accurate image retrieval.
• Future research should focus on developing sophisticated semantic-enhanced CBIR systems for clinical applications.