Alzheimer's Imaging Consortium.

Area of Science:

• Digital pathology
• Neurodegenerative disease research
• Machine learning in diagnostics

Background:

• Neuropathological evaluation is the gold standard for diagnosing neurodegenerative diseases but faces challenges in data sharing and sample preparation variability.
• Digital pathology and machine learning offer solutions for standardizing approaches and improving diagnostic accuracy.

Purpose of the Study:

• To develop the Brain Digital Slide Archive (BDSA), a federated, open-source platform for neurodegenerative disease research and diagnostics.
• To support whole slide images (WSIs) for Alzheimer's disease (AD) and related dementias (ADRDs) adhering to FAIR data principles.
• To enable standardized sharing of WSIs, annotations, and metadata for machine learning algorithm development.

Main Methods:

• Building on the Digital Slide Archive (DSA), the BDSA is a NIH U24 funded initiative.
• The platform integrates WSIs from nine AD/ADRD research centers, harmonizing data and implementing data-sharing agreements.
• Includes an anonymization tool for donor confidentiality and adheres to administrative best practices for a secure platform.

Main Results:

• Initial development integrates WSIs from nine research centers with harmonized data and universal data-sharing agreements.
• An anonymization tool is incorporated to ensure donor confidentiality.
• Data use and material transfer agreements are completed for all contributing U.S. sites, with ongoing software testing and standardized protocols.

Conclusions:

• The BDSA democratizes access to neuropathological data via a standardized, user-friendly digital repository.
• Integrated tools for privacy, data sharing, and machine learning will enhance understanding of neurodegenerative diseases and foster collaboration.
• As a federated open-source resource, the BDSA has the potential to transform digital neuropathology for AD/ADRD research and diagnostics.