EDM_Presentation_6222

1.

A System for Predicting
the Risk of Developing
Diabetes Based on Neural
Network Models
Gabitova Aigul
Kazan Federal University
gabit.aigul@gmail.com

2.

PROBLEM & MOTIVATION

3.

SYSTEM ARCHITECTURE

4.

DATASETS & NEURAL NETWORK

5.

EXPEREMENTAL RESULTS

6.

APPLICATION WORKFLOW

7.

DISCUSSION & LIMITATIONS
+
-
Multi-source risk analysis
No external clinical
validation
Accessible AI screening
Limited demographic
diversity
Full-stack architecture
Not a medical device

8.

CONCLUSION
CONCLUSION
FUTURE WORK
Multi-model AI platform
(Accuracy above 98%)
Integrated into real web
environment
Personalized
recommendations
Larger datasets
Supports accessible
diabetes screening
External validation

9.

Thank You for Your
Attention
Gabitova Aigul
Kazan Federal University
gabit.aigul@gmail.com