Accelerating pharmaceutical R&D with a user-friendly AI system for histopathology image analysis

Area of Science:

• Digital pathology
• Computational biology
• Pharmaceutical research

Background:

• Histopathology analysis is crucial in pharmaceutical R&D but often requires specialized expertise.
• Integrating advanced computational tools into existing workflows can be challenging.

Purpose of the Study:

• To develop an AI-driven system for accessible histopathology data analysis.
• To foster interdisciplinary collaboration between data scientists and clinicians.
• To streamline quantitative analysis of whole slide images.

Main Methods:

• Deployment of state-of-the-art AI tools as self-service modules on an open-source whole slide image platform.
• Integration of segmentation, feature extraction, and multi-instance learning models.
• Implementation of a CI/CD pipeline for robust model deployment and versioning.
• Automated cataloging of analysis outputs for data provenance tracking.

Main Results:

• Enabled non-data scientists (e.g., clinicians) to utilize and evaluate AI algorithms.
• Provided interactive visualization of results as annotations and heatmaps.
• Successfully applied to diverse pharmaceutical development tasks including glomeruli segmentation, podocyte counting, target engagement measurement, and PD-L1 score prediction.
• Integrated the Segment Anything model to accelerate annotation processes.

Conclusions:

• The developed system effectively democratizes AI-driven histopathology analysis in pharmaceutical R&D.
• Facilitates interdisciplinary collaboration and accelerates drug development timelines.
• Offers a flexible and scalable platform for integrating novel AI models.