Proposal of Local Automatic Weighing Attribute in CBIR

Area of Science:

• Medical imaging analysis
• Artificial intelligence in healthcare
• Oncology research

Background:

• Lung cancer is a leading cause of global cancer deaths.
• Computer-aided diagnosis (CAD) systems are crucial for early lung cancer detection.
• Content-based image retrieval (CBIR) shows potential for analyzing medical images but faces challenges.

Purpose of the Study:

• To address the semantic gap in content-based image retrieval (CBIR) for lung cancer diagnosis.
• To enhance the accuracy of CBIR systems in retrieving relevant lung cancer images.
• To develop an automated method for adjusting CBIR attribute weights based on user interpretation.

Main Methods:

• Development of a novel approach for automatic adjustment of CBIR attribute weights.
• Focus on aligning image retrieval with user semantic interpretation.
• Minimizing the semantic gap between low-level image features and high-level semantic concepts.

Main Results:

• Successfully minimized the semantic gap in CBIR for lung cancer image retrieval.
• Demonstrated significant improvement in retrieval accuracy.
• Validated the effectiveness of the automated weight adjustment method.

Conclusions:

• The proposed approach effectively bridges the semantic gap in CBIR for lung cancer.
• Automated adjustment of CBIR attribute weights enhances diagnostic accuracy.
• This method offers a promising solution for computer-aided lung cancer diagnosis.