Machine learning and Fuzzy logic fusion approach for osteoporosis risk prediction

Area of Science:

• Biomedical Engineering
• Data Science
• Public Health

Background:

• Osteoporosis is a global metabolic disorder with significant public health impact.
• Progression is influenced by modifiable lifestyle risk factors, but binary data presents challenges for traditional machine learning.
• Existing machine learning models struggle with accuracy and computation time due to the nature of binary risk factors.

Purpose of the Study:

• To develop an optimized model for predicting osteoporosis risk by integrating fuzzy logic with machine learning.
• To address the limitations of binary data in identifying modifiable risk factors for osteoporosis.
• To enhance prediction accuracy and reduce computational demands in osteoporosis risk assessment.

Main Methods:

• A novel approach fusing machine learning with fuzzy logic was developed.
• Three binary lifestyle risk factors (e.g., diet, smoking, exercise) were converted into a single fuzzy input, incorporating uncertainty.
• The model's efficiency and accuracy were optimized by reducing the number of input features.

Main Results:

• The fusion model demonstrated improved accuracy and reduced computation time compared to traditional machine learning methods.
• Condensing multiple binary variables into a single fuzzy input effectively handled data uncertainty.
• The model provides guidance for lifestyle modifications to mitigate osteoporosis risk.

Conclusions:

• The proposed fuzzy logic and machine learning fusion offers a more accurate and efficient method for osteoporosis risk prediction.
• This approach effectively manages uncertainty in binary lifestyle data, leading to better predictive performance.
• The validated model can aid in personalized risk assessment and lifestyle intervention strategies for osteoporosis prevention.