About the Systematic Review
Soft-tissue and bone tumours (STBT) are rare and diagnostically challenging lesions. This systematic review provides an overview of Artificial Intelligence (AI) methods in radiological imaging for diagnosing and prognosing these tumours. The review evaluates study adherence to the CLAIM and FUTURE-AI guidelines, aiming to enhance the development of trustworthy, clinically translatable AI tools.
Key Findings
- 325 articles reviewed from an initial pool of 15,015 abstracts.
- Moderate adherence to the CLAIM guidelines (28.9/53) and low adherence to FUTURE-AI guidelines (5.1/30).
- AI tools are at a proof-of-concept stage, with significant room for improvement in terms of clinical readiness.
Implications
- Define unmet clinical needs and clarify intended clinical settings for AI integration.
- Focus on training data that accurately reflect real-world clinical scenarios, and prioritize AI explainability.
- Address biases in AI models and ensure reproducibility by making both data and code publicly available.
Explore the Data
Access visualizations and datasets for CLAIM and FUTURE-AI evaluations.
CLAIM Visualization FUTURE-AI Visualization