Artificial intelligence in dermatopathology: Updates, strengths, and challenges

Area of Science:

• Medical Artificial Intelligence
• Digital Pathology
• Dermatopathology

Background:

• Artificial intelligence (AI) is increasingly influential in medicine.
• AI, including machine learning and deep learning, offers potential in pathology.
• Focus is on AI's application in dermatopathology and neoplastic dermatopathology.

Purpose of the Study:

• To explore the role and impact of AI in dermatopathology.
• To assess AI's diagnostic and prognostic capabilities in skin pathology.
• To review AI's performance in classifying skin lesions and predicting outcomes.

Main Methods:

• Review of studies on AI applications in dermatopathology.
• Analysis of AI's performance in diagnostic tasks using whole slide imaging.
• Evaluation of AI in classifying melanocytic lesions and differentiating skin conditions.

Main Results:

• AI, particularly convolutional neural networks, can exceed human pathologist sensitivity and specificity.
• AI aids in identifying skin lesions, classifying melanomas, and predicting disease outcomes.
• AI demonstrates effectiveness in neoplastic dermatopathology, assisting in accurate diagnoses.

Conclusions:

• AI integration in dermatopathology shows significant promise as a diagnostic aid.
• Collaboration between medical and technical experts is crucial for responsible AI implementation.
• AI is positioned as a supportive tool for pathologists, not a replacement for human expertise.