[EMPAIA-ecosystem for pathology diagnostics with AI assistance]

Area of Science:

• Pathology
• Artificial Intelligence
• Medical Diagnostics

Background:

• Deep learning and AI are increasingly used in pathology research.
• Clinical application of AI in pathology lags behind research due to several hurdles.
• Early certified AI solutions exist for specific analyses like prostate sections and breast cancer markers (ER, PR, Her2).

Purpose of the Study:

• To address the barriers hindering the widespread adoption of AI in clinical pathology.
• To develop an ecosystem (EMPAIA) that facilitates the integration of AI into routine diagnostics.
• To demonstrate the feasibility of broad AI application in image-based diagnostics through a moderated, collaborative approach.

Main Methods:

• The EMPAIA project focuses on building a supportive ecosystem.
• A joint, moderated process is employed to overcome existing hurdles.
• Technical solution approaches for AI integration are being developed and described.

Main Results:

• The EMPAIA ecosystem aims to remove hurdles for AI application in diagnostics.
• The project's strategy is designed to enable broad adoption of AI tools.
• The described components and strategy pave the way for AI in image-based diagnostics.

Conclusions:

• Overcoming current hurdles is crucial for the clinical implementation of AI in pathology.
• The EMPAIA ecosystem provides a framework for integrating AI into diagnostic workflows.
• A collaborative approach can accelerate the broad adoption of AI in image-based diagnostics.