Artificial intelligence in rare diseases: toward clinical impact.

Area of Science:

• Medical Informatics
• Genomics
• Artificial Intelligence

Background:

• Rare diseases (RDs) impact over 400 million globally, with many facing diagnostic delays and limited treatment options.
• Existing Artificial Intelligence (AI) tools for RDs often remain in proof-of-concept stages, hindering clinical translation.
• Key challenges include data scarcity, privacy concerns, and algorithmic biases, limiting AI's impact on patient care.

Purpose of the Study:

• To review recent advancements in AI, including generative models, federated learning (FL), and explainable AI (XAI), for addressing RD challenges.
• To identify critical technical, ethical, and infrastructural priorities for translating AI innovations into clinical benefits for rare disease patients.
• To highlight the importance of equity, FAIR data principles, and global coordination in the development and deployment of AI for RDs.

Main Methods:

• Literature review of recent developments in AI for rare diseases.
• Analysis of emerging AI techniques such as generative models, federated learning, and explainable AI.
• Synthesis of technical, ethical, and infrastructural requirements for clinical translation.

Main Results:

• Advances in generative models, FL, and XAI are beginning to overcome data scarcity, privacy, and bias issues in AI for RDs.
• The review outlines specific priorities including equity, FAIR data principles, and global collaboration.
• These priorities are essential for bridging the gap between AI innovation and tangible patient benefits.

Conclusions:

• AI holds significant promise for accelerating rare disease diagnosis and therapeutic discovery.
• Translating AI's potential into clinical reality requires a concerted focus on ethical considerations, data accessibility (FAIR principles), and international cooperation.
• Strategic prioritization of technical, ethical, and infrastructural factors is essential for realizing the full clinical impact of AI in rare disease management.