To propose a hybrid approach for automatic identification of skin type and multi-class classification of skin diseases using dermatological images.
Approach:
Methodology: Combines convolutional neural networks (CNNs), cellular automata, and fractional-order derivatives to enhance skin type classification and disease categorization.
Key Findings:
The proposed method reduces misclassification of normal skin and increases classification accuracy by approximately 1.2 percentage points.
Achieved an accuracy of 92.8%, sensitivity of 91.4%, and F1-score of 91.7% for skin disease classification.
Achieved an accuracy of 92.4%, sensitivity of 91.1%, and F1-score of 91.4% for skin type classification.
Interpretation:
The integrated model offers a scalable approach for intelligent dermatological assessment and skin image analysis.
Limitations:
The study does not address the potential limitations of the individual techniques used.
Performance may vary based on the quality and diversity of the training datasets.
Conclusion:
The study presents an effective method for skin type and disease classification by integrating deep learning, cellular automata, and fractional derivatives.
