An artificial immune network based algorithm for diabetes diagnosis

Area of Science:

• Biomedical Engineering
• Computational Biology
• Artificial Intelligence in Medicine

Background:

• Diabetes mellitus is a global health concern requiring accurate and efficient diagnostic tools.
• Traditional diagnostic methods can be time-consuming and may require specialized equipment.
• Developing intelligent systems for early disease detection is crucial for effective management.

Purpose of the Study:

• To present a novel artificial immune network (AIN) based algorithm for diabetes diagnosis.
• To evaluate the efficacy of the proposed AIN algorithm in identifying diabetic patients.

Main Methods:

• Implementation of an AIN algorithm involving the creation of an initial immune antibody network.
• Evolution of the network through learning from foreign antigens, simulating immune system responses.
• Diagnosis of diabetes using a majority vote from k-nearest neighbor antibodies within the network.

Main Results:

• The developed AIN algorithm demonstrates a potential for accurate diabetes diagnosis.
• The network's ability to learn and adapt from antigen data contributes to diagnostic performance.
• The k-nearest neighbor voting mechanism provides a robust method for classification.

Conclusions:

• The proposed artificial immune network algorithm offers a promising, intelligent approach to diabetes diagnosis.
• This computational method could complement existing diagnostic strategies for diabetes.
• Further research can explore the optimization and clinical validation of this AIN-based system.