Fuzzy segmentation and black widow-based optimal SVM for skin disease classification

Area of Science:

• Dermatology
• Medical Imaging
• Computational Biology

Background:

• Skin cancer is a significant global health concern, ranking as the fourth leading cause of non-fatal disease risk.
• Accurate skin disease classification is challenging due to issues like inaccurate outputs, overfitting, and high computational costs.

Purpose of the Study:

• To present a novel Support Vector Machine-based Black Widow Optimization (SVM-BWO) algorithm for improved skin disease classification.
• To address the limitations of existing methods in terms of accuracy and computational efficiency.

Main Methods:

• Utilized a dataset of five skin diseases (psoriasis, paederus, herpes, melanoma, benign) and healthy skin images from ISIC-2018.
• Implemented image pre-processing to remove noise, followed by fuzzy set segmentation to isolate skin lesions.
• Extracted color, texture (Gray-Level Co-occurrence Matrix), and shape features for classification.
• Employed the SVM-BWO algorithm for the classification task, implemented in MATLAB 2018a.

Main Results:

• The proposed SVM-BWO methodology achieved a classification accuracy of 92%.
• Experimental performance analysis demonstrated that the SVM-BWO approach outperformed other state-of-the-art techniques.
• The method effectively handles feature extraction and classification for diverse skin conditions.

Conclusions:

• The SVM-BWO algorithm offers a robust and accurate solution for skin disease classification.
• This novel approach shows significant potential for improving diagnostic accuracy in dermatology.
• The study highlights the effectiveness of metaheuristic optimization combined with machine learning for medical image analysis.