A systematic literature review on transparency and interpretability of AI models in healthcare: taxonomies, tools, techniques, datasets, open research challenges, and future trends

Key Points

• Findings reveal that transparency in AI models is crucial for reliable healthcare outcomes, enhancing trust among stakeholders.
• Key evidence indicates that interpretability tools are varied, but only a small percentage meet the needs of healthcare applications.
• Systematic literature review analyzes datasets and techniques across multiple studies, pointing to the importance of shared benchmarks.
• This assessment highlights the need for improved frameworks to address open research challenges in AI transparency and interpretability.