BioIVT’s ASTERAND® Human Tissue Repository contains clinical data from more than one million biospecimens and can be mined to reveal patterns and relationships in disease etiology and pathology.
BioIVT, a leading provider of research models and services for drug and diagnostic development, is enabling customers to mine its extensive biospecimen repository and related clinical database using artificial intelligence (AI) and machine learning (ML) applications to advance scientific knowledge of different diseases. AI allows huge volumes of data to be analyzed to reveal previously undiscovered patterns and relationships.
“We want to use BioIVT’s extensive, well-characterized, high quality tissue repository and accompanying clinical database to help progress customers’ drug discovery and diagnostics research. We already have hematoxylin and eosin (H&E) images and clinical data from more than one million biospecimens, collected over 30 years, in our ASTERAND Human Tissue Repository,” said Cathie G. Miller, PhD, Senior Director of Product Marketing, Personalized Medicine at BioIVT. “In addition, our biobank consistently contains hundreds of thousands of disease-state and normal biospecimens that are collected by our global network of more than 350 donor institutions. Every one of our biospecimens is supplied with more than 200 data points, including patient outcomes,” added Dr. Miller.
Also featured in each clinical dataset are the donor’s diagnosis and status, co-morbidities, treatments (past and present), medications, and cause of death if available. Biospecimens from donors with cancer include the AJCC/UICC stage, and histological data, such as the percentage of tumor, normal and necrotic tissue.
All BioIVT donor centers follow institutional review board-approved protocols, ensuring that they meet high standards of ethics, informed patient consent and sample integrity. To maintain donor confidentiality and prevent sampling bias, each BioIVT biospecimen is identifiable only by a barcode.
BioIVT’s repository is particularly well suited for AI research because the company was an early adopter of high-quality digital images and has been building a rich database in many therapeutic areas since 2000. The company incorporated Aperio Digital Pathology Slide Scanners into its tissue pathology review process in 2016 and its team has also scanned biospecimens from many historical cases. As a result, BioIVT has an extensive collection of full-slide Aperio H&E images that are ready to use with ML-based image analysis platforms. These digital images can be viewed on BioIVT’s website when customers log in to assess available inventory.
To ensure accuracy and consistency, all BioIVT biospecimens are reviewed by board-certified pathologists who confirm that every tissue block, and hence every image, is representative of the donor’s diagnosis. All the related clinical data are also independently reviewed by BioIVT specialists and verified by board-certified pathologists.
Additional information about BioIVT’s AI-related capabilities can be found at https://info.bioivt.com/ai-solutions.