- Human Biospecimens
- For Researchers
- For Biospecimen Contributors
- For Patients
April 11, 2019
Every human biospecimen contains a wealth of data – about patients, their conditions and the biological processes that the sample has undergone. Data is also the currency of artificial intelligence, which is streamlining discovery by crunching data as biospecimens are presented to computers.
For example, AI can “digitally stain” human tissue samples for viewing by histopathologists – professionals who examine tissue under microscopes to detect disease. The conventional sample staining process involves a technician using dye to provide the necessary contrast for viewing thin slices of tissue on slides. Without this contrast, the samples would be inscrutable.
At UCLA, AI is automatically converting low- or no-contrast microscopic images of tissue into higher-contrast images, thus enabling histopathologists to detect the tissue’s biological features. According to the UCLA Engineering Institute for Technology advancement:
Researchers at UCLA have developed a deep learning-based method to take a microscopic image of naturally present fluorescent compounds in unstained tissue sections and digitally transform this “auto-fluorescence” image into an equivalent image of the same tissue, as if it was taken after the standard tissue staining process. Stated differently, this deep learning-based method virtually stains unlabeled tissue samples, replacing the manual and laborious processing and staining steps that are normally performed by histotechnologists or medical personnel, saving labor, cost and time by substituting most of the tasks performed a histotechnologist with a trained neural network.
This digital staining process also standardizes images that would otherwise vary by the technician doing the staining. The “technician” is now a computer. A panel of board-certified pathologists found no clinically significant difference in staining quality between traditionally stained and virtually stained images.
AI detects brain tissue pathology
In another example of AI innovation, researchers at the Icahn School of Medicine at Mount Sinaihave trained computers to detect neurofibrillary tangles—biological structures associated with diseases like Alzheimer’s and chronic traumatic encephalopathy—in digitized microscopic slides of brain tissue samples.
“Applying deep learning, these images were used to create a convolutional neural network capable of identifying neurofibrillary tangles with a high degree of accuracy directly from digitized images,” Mount Sinai explained.
“Utilizing artificial intelligence has great potential to improve our ability to detect and quantify neurodegenerative diseases, representing a major advance over existing labor-intensive and poorly reproducible approaches,” said lead investigator John Crary, MD, PhD, Professor of Pathology and Neuroscience at the Icahn School of Medicine at Mount Sinai. “Ultimately, this project will lead to more efficient and accurate diagnosis of neurodegenerative diseases.”
Valuable biospecimen data continues to accumulate
Human biospecimens, though valuable in themselves, yield data in volumes that are valuable in teaching computers to make scientists and clinicians faster, smarter and more productive, and to assist us in diagnosis, biomarker discovery and other important tasks that advance healthcare.
That’s why we at iSpecimen, in addition to providing a platform for biospecimen exchange, have also been managing a database of de-identified clinical information, including results from more than 300 million laboratory tests from more than 8 million patients.
Every specimen is a font of data, and when it comes to specimens and data, there’s big power in numbers.
Learn more about the iSpecimen Marketplace where you can browse millions of richly annotated, de-identified tissue samples and biofluids or to request a quote or custom collection. You can join for free and creating a login is easy.