Thanks to pioneering work at Stony Brook Medicine, digital solutions are in development for pathology slide specimens — the last major frontier in digitizing medical images.
Dr. Joel Saltz, Chair of Biomedical Informatics at Stony Brook Medicine and a board-certified clinical pathologist with a PhD in Computer Science, is at the forefront of this major breakthrough.
For more than 20 years, Dr. Saltz and his team, which includes physicians and technologists from Johns Hopkins, Ohio State, Emory and Stony Brook, have been developing digital solutions for pathology slide specimens. His groundbreaking work in digital image viewing and archiving systems for pathology images set into motion technology developments that ultimately led to the FDA’s recent approval of using digital surgical pathology slides for interpretations.
What’s more, Dr. Saltz’s pioneering work in use of machine learning and artificial intelligence methods in pathology is at the forefront of clinical and research innovation in this field.
Dr. Saltz, along with his colleagues Fred Prior from the University of Arkansas and Ashish Sharma at Emory, received a five-year $8 million grant from the National Cancer Institute (NCI) to develop ways to use this new technology to enhance cancer treatments. This work is expected to dramatically improve the NCI’s Cancer Imaging Archive, a major resource for cancer researchers and clinicians, by enabling digital pathology imaging to enhance the existing radiology effort.
Pathology — the study of the nature of diseases and the structural and functional changes that they cause — has traditionally been a science that relies on human expertise for interpretation. A pathologist uses a microscope to examine biopsied specimens prepared on glass slides and makes a diagnosis based on keen observation.
There is increasing evidence that only a small fraction of the information locked in a glass slide is used in medical care, which can result in lost opportunities to steer treatment and missed opportunities to flag potentially aggressive cancers. In addition, it’s possible and even common for different yet equally skilled pathologists to have different readings of what they see on the same slide, which can result in different diagnoses, and potentially, even different treatments for the disease.
But, with reliance on the human factor of pathology being the standard mode, unique interpretations along with variations in both diagnoses and treatments have been an accepted part of the equation.
Digitizing pathology slides can help improve consistency in readings by allowing for machine analysis that can interpret more data than human counterparts. In addition, digitizing the slides allows for easier sharing of pathology data among physicians and leads to time-saving as well as more accurate diagnoses and treatments. Like radiology images, which started being digitized two decades ago, pathology images will be able to be shared among physicians via standard internet browsers.
While using this technology is not yet widespread, Dr. Saltz anticipates that it won’t be long before digital imaging and artificial intelligence based analytics in pathology becomes the new normal, bringing with it the promise of truer diagnoses, more uniformity in treatments, and the ultimate goal — better outcomes for patients at Stony Brook University Cancer Center.
About Dr. Joel Saltz
Dr. Saltz is a digital pathology pioneer, having worked for the past 20 years in the development of digital pathology whole slide image software, methods, tools and algorithms. He is a board certified clinical pathologist, holds an MD-PhD in Computer Science from Duke, completed a clinical pathology residency from Johns Hopkins and has founded Biomedical Informatics departments at Stony Brook, Emory and Ohio State. Dr. Saltz is Chair and Professor of Biomedical Informatics at Stony Brook University, is Associate Director for Informatics at Stony Brook Cancer Center and holds the Cherith endowed chair.
Digital Pathology Team Members
Colleagues from Stony Brook Medicine and Stony Brook University are leading the way in developing digital solutions for pathology slide specimens. They include:
Kenneth R. Shroyer, MD, PhD
The Marvin Kuschner Professor and Chair, Stony Brook Medicine Pathology Department
Tahsin M. Kurc, PhD
(worked together on this project initially at Johns Hopkins, and now at Stony Brook)
Vice Chair and Research Associate Professor, Department of Biomedical Informatics, Stony Brook Medicine
Fusheng Wang, PhD
(worked together on this project initially at Emory, and now at Stony Brook)
Assistant Professor, Department of Computer Science, Stony Brook University
Jonas Almeida, PhD, Professor and Chief Technology Officer
Department of Biomedical Informatics, Stony Brook Medicine
Dimitrios Samaras, PhD
Director, Computer Vision Lab, Computer Science Department, Stony Brook University