Journal of Pathology Informatics Editorial calls Corista-led study on use of AI and machine learning for renal biopsy analysis “Sound” and Results “Encouraging”

CONCORD, Mass. (PRWEB)
June 09, 2021

A Corista-led study on the impact of applying artificial intelligence to the workflow of a Renal Pathologist and how this might ease and improve the Pathologist’s workflow in reviewing renal biopsy slides has received an editorial stamp of approval from the Journal of Pathology Informatics.

“Although being a limited proof-of-concept study on a small number of cases, the concept appears sound, and the results were encouraging,” a commentary on “The Digital Fate of Glomeruli in Renal Biopsy” noted in the March 22 edition of the publication. “The conclusions of the authors . . . reinforce the belief that these machine learning tools are going to become a routine tool for the practicing pathologist, earlier than expected.”

The pathologists conducting the study¹ developed a prototype system of digitized, synchronized section display and machine learning-based feature detection in medical renal biopsies.

Today, a renal pathologist must review four slides separately to analyze a biopsy featuring glomeruli, tiny capillaries in a unique formation that act as filters throughout the kidney and can show evidence of a variety of diseases. This involves switching slides on a microscope and locating the same fields of view and the same cells or structures within those fields, which can be painstaking and time-consuming. By digitizing the biopsy slides used in the study, the sections were available in one four-panel simultaneous viewing display for synchronized navigation and magnification.

The prototype also involved applying a convolutional neural network (CNN) to analyze the digitized slides. A CNN is an algorithm that can take in images, assign importance to various aspects of them, and learn to differentiate one from the other. Using information from the digitized slides, the CNN was able to accurately locate glomeruli in renal biopsies.

Taken together, the simultaneous viewing display and machine learning algorithms showed that they have the potential to improve the efficiency and accuracy of medical renal tissue biopsies. “We would need to conduct much more rigorous clinical trials to confirm those improvements,” said David Wilbur, chief medical scientist and pathologist for Corista. “Several renal pathologists have expressed interest in participating, so we’re hopeful that we’ll be able to move beyond proof of concept to actual use in the relatively near future.”

Corista, a leader in integrated digital pathology solutions, has a history of innovation in the industry. In mid-2020, the company received a USPTO patent grant for its new Virtual Slide Stage (VSS). This interface device greatly improves the ergonomic efficiency of digital slide viewing for pathologists and could ease the path to greater adoption of digital pathology platforms.

About Corista

Corista delivers the industry’s most extensive array of workflow, analytical and collaborative tools for pathology. Medical centers can seamlessly integrate with LIS/EHRs in a unified digital environment of whole-slide, gross and static pathology images. Physicians have 21st-century tools to collaborate, communicate, teach and report with access to ‘best of breed’ image analytics. Specialists can receive digital consults from remote physicians and patients from across the globe, and investigator-initiated researchers have a research and development platform to develop and apply their algorithms. Corista provides for a new level of interoperability for pathology, integrating whole slide image scanning systems, image analytics and LIS/EHR platforms with a rich, collaborative environment for physicians, patients, bio-tech and pharmaceutical scientists. This is Integrated Pathology™ only from Corista. For more information, visit https://corista.com/.

¹ “Using Image Registration and Machine Learning to Develop a Workstation Tool for Rapid Analysis of Glomeruli in Medical Renal Biopsies.” Published November 7, 2020, Journal of Pathology Informatics. David C. Wilbur, Alexander Andryushkin, Richard Y. Wingard II, Eric Wirch – Corista. Jason R. Pettus – Department of Pathology, Dartmouth-Hitchcock Medical Center. Maxwell L. Smith and Lynn D. Cornell – Department of Pathology, Mayo Clinic.

Share article on social media or email: