Lyme rashes disease classification using deep feature fusion technique.

Area of Science:

• Dermatology
• Artificial Intelligence
• Medical Imaging

Background:

• Accurate classification of Lyme disease rashes is crucial for effective clinical management.
• Current diagnostic methods may benefit from enhanced computational approaches.

Purpose of the Study:

• To develop and evaluate an in-depth feature fusion system for automatic classification of Lyme disease rashes.
• To improve the accuracy and efficiency of identifying erythema migrans (EM) images.

Main Methods:

• Utilized three deep learning models (Densenet201, InceptionV3, Exception) for initial feature extraction from EM images.
• Developed a meta-classifier (deep convolutional neural network) to fuse features from the base models.
• Integrated extracted deep features with original images for final classification.

Main Results:

• The proposed deep feature fusion system achieved a classification accuracy of 98.97%.
• This accuracy significantly outperforms existing state-of-the-art models for Lyme disease rash classification.
• The system demonstrates high efficacy in distinguishing erythema migrans.

Conclusions:

• The novel in-depth feature fusion system offers a highly accurate method for automatic Lyme disease rash classification.
• This AI-driven approach can significantly aid clinicians and dermatologists in diagnosing Lyme disease.
• The developed system represents a substantial advancement in computational dermatology and medical diagnostics.