CADUCEO: A Platform to Support Federated Healthcare Facilities through Artificial Intelligence

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Clinical Decision Support Systems

Background:

• Data-driven algorithms excel in medical tasks like disease prediction and diagnostics, often matching clinician performance.
• Widespread clinical adoption of AI is hindered by needs for high-quality, representative data, lack of explainability, and regulatory concerns.

Purpose of the Study:

• To address the limitations of current AI algorithms in clinical settings.
• To propose a novel AI-based functional framework to facilitate wider AI application in healthcare.

Main Methods:

• Utilizing data augmentation and synthetic data generation techniques.
• Implementing federated learning for privacy-preserving data analysis.
• Employing explainable artificial intelligence (XAI) methods to enhance transparency.

Main Results:

• The proposed framework aims to overcome data quality and explainability challenges in clinical AI.
• Potential to reduce time for medical prevention, diagnosis, and prognosis by up to 70%.

Conclusions:

• Novel AI framework offers a viable solution for integrating AI into clinical practice.
• Federated learning and XAI are key to enabling broader AI adoption and improving patient outcomes.