The Promise of Explainable AI in Digital Health for Precision Medicine: A Systematic Review

Area of Science:

• Digital Health
• Precision Medicine
• Artificial Intelligence
• Machine Learning

Background:

• Healthcare is increasingly adopting personalized treatments tailored to individual patient characteristics.
• The integration of artificial intelligence (AI) with digital health data is becoming essential for precision medicine.
• Explainable AI (XAI) is critical for fostering transparency, accountability, and trust in AI-driven healthcare applications.

Purpose of the Study:

• To synthesize the literature on explainable machine learning models for digital health data in precision medicine.
• To identify key themes and applications of explainable AI in optimizing patient healthcare.
• To underscore the need for further development and validation of AI explanation methods in healthcare.

Main Methods:

• A comprehensive literature review of 27 peer-reviewed journal articles.
• A Google Scholar search was conducted up to September 19, 2023, with no time constraints.
• Topic-modeling approach used to distill key themes from the selected articles.

Main Results:

• Identified key themes include optimizing patient healthcare via data-driven medicine, predictive modeling, disease prediction using deep learning on biomedical data, and machine learning applications in medicine.
• Specific applications of explainable AI were examined, focusing on its contribution to transparency, accountability, and trust.
• The review synthesized current research on machine learning explainability within the precision medicine context.

Conclusions:

• Explainable AI is vital for the effective and trustworthy implementation of digital health data in precision medicine.
• Further research and validation of explanation methods are necessary to advance the delivery of precision healthcare.
• The findings emphasize the growing importance of interpretable AI models in clinical decision-making and patient care.