Machine learning (ML) techniques as effective methods for evaluating hair and skin assessments: A systematic review

Area of Science:

• Dermatology and Computational Science

Background:

• Machine Learning (ML) and Artificial Intelligence (AI) offer advanced capabilities for evaluating human skin and hair.
• A systematic review was conducted to assess the efficacy of ML/AI in skin and hair analysis.

Approach:

• Searched PubMed, Web of Science, IEEE Xplore, and Science Direct for publications from January 2010 to March 2020.
• Included 20 peer-reviewed articles focusing on "hair and skin analysis" after rigorous screening.
• Analyzed prevalent ML methods including Support Vector Machine (SVM), k-nearest Neighbor, and Artificial Neural Networks (ANN).

Key Points:

• Artificial Neural Networks (ANNs) achieved the highest accuracy (95%), followed by Support Vector Machines (SVM) at 90%.
• ANNs demonstrated superior sensitivity (82.30%) and specificity (96.90%).
• These ML techniques are frequently applied in diagnostic frameworks, notably for Melanoma assessment.

Conclusions:

• ML techniques, particularly ANNs, are highly effective for analyzing and evaluating skin and hair assessments.
• The review highlights the potential of AI and ML in improving diagnostic accuracy and efficiency in dermatology.
• Further research can explore advanced ML applications for complex skin and hair conditions.