Artificial Intelligence and Machine Learning in Bone Metastasis Management: A Narrative Review.

Area of Science:

• Oncology
• Radiology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI) and machine learning (ML) are increasingly utilized in diagnosing and managing bone metastases.
• Applications span lesion detection, segmentation, prognostic modeling, fracture risk assessment, and surgical decision support.
• The rapidly evolving and heterogeneous literature presents challenges for clinical contextualization.

Purpose of the Study:

• To conduct a narrative review of AI/ML applications in bone metastasis management.
• To focus on clinically relevant problems like lesion detection, segmentation, and skeletal-related event prediction.
• To prioritize recent studies with reported model performance and clinical translation potential.

Main Methods:

• Narrative literature review.
• Focus on AI/ML applications in bone metastasis detection, segmentation, prognostication, and surgical decision support.
• Prioritization of recent, peer-reviewed studies reporting model performance and clinical translation.

Main Results:

• Most studies focus on imaging-based diagnosis and lesion segmentation using radiomics and deep learning, with high internal but limited external validation.
• Emerging research on prognostic models and fracture risk estimation is in early proof-of-concept stages.
• Few AI frameworks are integrated into routine workflows; explainability, bias, and cost-effectiveness are rarely evaluated.

Conclusions:

• AI and ML tools offer significant potential for standardizing imaging assessment and personalizing bone metastasis management.
• Future research should emphasize externally validated, multimodal models and AI-augmented clinical decision tools.
• Development of AI-augmented alternatives to the Mirels score and incorporation of explainability and cost-effectiveness analyses are crucial.