The Role of AI in Pathology
Article Jul 25, 2017 | By Anna MacDonald, Editor for Technology Networks
The role of the pathologist in the diagnosis of cancer is undergoing a digital transformation, with developments in artificial intelligence (AI) and machine learning set to change the way pathology labs work.
We spoke to David West Jr., CEO of Proscia, to learn more about the role that digital pathology software can play in the lab, the benefits it can bring, and what the future may hold for pathology.
Can you tell us how digital pathology software is currently being used in pathology labs?
For the past 150 years, pathologists have made cancer diagnoses by looking at tissue under a microscope. Today, digital whole slide imaging (WSI) allows the capture of the entire tissue sample on a slide, as opposed to a narrow field of view the microscope provides. The image files that are produced from whole slide imaging systems are very large and are usually at a magnification of 20x or 40x. This allows pathologists to capture a great deal of data that requires specialized digital pathology software to view and store such large amounts of information.
Once in the “digital realm” where pathologists are dealing with pixels rather than glass, two key functions are made possible. Firstly, distance no longer matters. Sending slides between two institutions on different sides of the world becomes much easier, opening up new collaboration possibilities, helping pathology departments and private labs to grow, and improving access to subspecialty experts, especially for patients in remote areas.
Secondly, image analysis algorithms will be able to operate on the images. These are already used for automated or semi-automated immunohistochemistry quantification, helping to drive standardization and speed of analysis – and new methods are in development to augment hematoxylin and eosin stain. The area could play a profound role in reducing inter-observer variability for many cancers and in driving faster, more precise quantitative workflows.
Gone are the days of having cabinets full of slides that are difficult to share and only viewable by squinting down a microscope. Pathologists can now examine tissue on their computers from anywhere in the world via WSI – and, with the help of computational pathology software powered by intelligent machine learning, will have new, precise information at their fingertips. The digital revolution is already underway.
How do you see this changing over the next few years?
Computational pathology is the next era of digital pathology with major advances in automation and augmentation.
As pathologists adopt digital pathology software, new realms of possibilities will open up for applications that enhance how pathologists review cases. One thing that is very clear is there are a decreasing number of pathologists globally while there are an increasing number of cases that require pathology attention, leading many pathologist to take on more cases than ever before. The implementation of digital pathology software has shown to improve productivity of pathologists by 13 percent. Over the next few years, we will see the imbalance between the amount of data and the number of pathologists even out as deep learning neural networks analyze the invaluable, untapped data amassed by pathology departments and translational research centers around the world. This new class of machine learning algorithms will be used to more accurately diagnose and predict disease.
What benefits can AI and machine learning bring to pathology?
The digitization of pathology is generating a tremendous amount of data, new methods are becoming available to technologists, and enormous amounts of computing resources are accessible via the cloud at low cost. All of this means that we can start implementing AI technology in the lab by training intelligent algorithms to recognize broad or specific patterns on a whole slide image to translate features evident in the tissue into prediction (such as metastasis and recurrence) and classification (staging, grading, and differential diagnosis). This enables the creation of predictive biomarkers based on precise measurements of histological patterns, providing pathologists with new tools to answer questions about a given patient’s disease. This is especially useful where molecular tests fall short – image-based assays could be part of a portfolio of tests in a pathologist’s precision medicine arsenal. Scanned slides can be used to train deep neural networks to learn how the cellular morphology reveals genetic and epigenetic changes in the tissue. In some ways, it’s still early days, but this is already starting to be used in cutting-edge laboratories, and deep learning is likely to have a broad range of applications within pathology.
Will there still be a need for pathologists?
The misconception when it comes to artificial intelligence in healthcare is if the machine can work faster and reach the same conclusions as a human then humans become irrelevant in the diagnostic process. In reality AI does not replace the expertise and experience of pathologists, rather the software and deep learning applications provide pathologists with data and insights they would not have had using traditional microscopy. Software like ours is designed to facilitate the work of the pathologist by automating monotonous, but necessary, tasks like identifying mitotic cells or screening benign tissue, and augment their capabilities to perform tasks like the identification of “cancer hotspots.” Pathologists can then focus their time on making diagnoses that are informed by their years of training and insights derived from digital pathology software.
In the next 10 years, we anticipate that much of the professional component of pathology will be divorced from the physical laboratory, with human pathologists working in software driven “labs.”
David West Jr. was speaking to Anna MacDonald, Editor for Technology Networks.
You can find out more about Proscia here.
Thanks to advances in therapeutics, chronic myeloid leukemia has been transformed from a deadly cancer into a manageable, chronic condition. However, patients have to stay on these drugs their whole lives, as there isn't a diagnostic test sensitive enough to detect traces of the disease that threaten a relapse. This article delves into the unmet need for such a test, and explains why droplet digital PCR (ddPCR) is a strong candidate that may fill this gap.READ MORE