Generalized aggregation index based collaborative fusion for medical diagnosis.

Area of Science:

• Data Science
• Artificial Intelligence
• Medical Informatics

Background:

• Defining similarity relationships is crucial for data fusion and decision-making.
• Traditional methods rely solely on data attributes, limiting relationship representation accuracy.

Purpose of the Study:

• To introduce a novel concept of generalized aggregation index (GAI) driven by both data and knowledge.
• To develop a GAI-based collaborative fusion method for multi-source medical data.

Main Methods:

• Proposed the concept of Generalized Aggregation Index (GAI) to represent cumulative generalized relationships.
• Developed a GAI-based collaborative fusion method where data units aggregate based on GAI-measured attraction capabilities.
• Explored both classical and quantum-inspired GAI methods.

Main Results:

• The GAI captures relationships beyond data attributes, incorporating decision effect knowledge.
• The GAI-based collaborative fusion method achieved high-quality fusion results.
• The proposed methods led to highly precise diagnosis conclusions in experimental analysis.

Conclusions:

• The GAI offers a more accurate representation of relationships between data units.
• GAI-based collaborative fusion is effective for multi-source medical data.
• The methods show promise for improving diagnostic accuracy in medical decision applications.