Pathology is in the midst of a radical transformation, driven by many factors. Foremost among these are the advancement of precision medicine, a pathologist shortage, and a need for more efficient and cost-effective workflows. Biocompare recently interviewed Mike Stewart, product manager at Philips Digital Pathology Solutions, to learn more about digital and computational pathology and where the field is headed. We also looked at some of the digital pathology solutions available now or in development.
Q: It seems we went from digital pathology to computational pathology pretty quickly. Can you talk about the evolution of the field and what computational pathology is specifically?
It is no secret that technological innovation in healthcare is growing at an increasingly fast pace across specialties, and pathology is no exception. Change is starting to infiltrate the industry as we know it with the deployment of digital pathology technology to aid workflows and increase pathologist collaboration.
Despite the pivotal role pathology plays in the diagnosis and treatment of disease, its practice is one of the last in medicine to benefit from digital workflows. In a setting where the largest expense has historically been the purchase of a microscope, it is easy to see how the transition to digital methods could seem financially daunting; but the benefits of “reinventing” the traditional laboratory are significant.
As healthcare shifts to focusing on delivering better care for patients at lower costs, ensuring that patients are diagnosed accurately and quickly is essential to enabling this transformation. Digitizing pathology breaks down the geographic barriers innately found in pathology’s manual review of glass slides. Enhanced collaboration allows pathologists to seek second opinions more quickly, collaborate with multidisciplinary care teams more effectively, and distribute workloads across more sites more evenly.
Once all of the information in pathology is digitized, applying artificial intelligence to it allows pathologists to derive deeper insights that weren’t previously possible. Linking image files to other valuable data sources—such as connected data ecosystems that include the patient’s history and unique risk factors—paints a fuller picture of the patient and can further enhance outcomes. This integration of data across pathology, radiology, clinical systems, and lab operations and applying artificial intelligence to derive insights is called computational pathology.
Q: What is the benefit to patients and healthcare in general of computational pathology?
Computational pathology tools enable pathologists to address the developing complex medical environment for cancer care. These tools can provide a comprehensive view and analysis of test results, treatment protocols, and behavior at a personalized level, allowing clinicians to obtain new insights across both individuals and various population subgroups.
As the industry drives toward precision medicine, computational pathology is a key factor in determining the speed in which this goal will become affordable and as accurate as possible. Computational pathology aims to improve diagnostic accuracy, optimize patient care, and reduce costs by improving laboratory efficiencies. As the connected data sets that define it become searchable, principal investigators who are researching new cancer treatments will be able to identify and recommend better, evidence-based treatment protocols for individual patients because they will be able to mine richer information from tissue than is possible with the naked eye alone.
Where digital pathology enables more efficient workflows, computational pathology will take the field one step further, allowing pathologists to use digital images in different and more efficient ways. In the future, smart image-recognition algorithms could help streamline pathologists’ workflows and help them focus on the things that matter most.
Two examples of the benefits of such streamlining are as follows:
Q: What is the role of computational pathology in personalized/precision medicine?
The use of whole-slide images, from the starting point of drug and biomarker discovery to preclinical biomarker teams and clinical researchers, may have significant benefits for the patient.
As we move toward an era of precision medicine, we will see greater emphasis on global collaborations between industry, academic institutions, and the experts in artificial intelligence. Driving this is the need to develop clinical algorithms that could more accurately determine the diagnosis, treatment, and life expectancy for each patient. Collaborations are already taking place between vendors, industry, and research institutions, which has the potential to translate their algorithms into a deep learning variant to accelerate performance and greater diagnostic and predictive accuracy.
Tissue diagnostics and biomarker analytics are the keystones of cancer discovery. However, delivering on the promise of personalized medicine requires multiple data sources to be integrated and analyzed. Management and analysis of large volumes of tissue samples are crucial to realizing the potential of personalized medicine.
To accomplish this, imaging and pathology informatics support will be essential to moving forward in an increasingly complex and multifaceted medical research environment. With the power of machine learning and the power of big data management tools, researchers can integrate data from multiple sources, including digital pathology and tissue imaging, and so enhance their ability to glean critical insights into disease and identify novel biomarkers.
Q: How is the role of the pathologist shifting with the advent of computational pathology?
With the growing number and complexity of cancer cases, the demands on pathologists are ever increasing. At the same time, the industry is facing a severe shortage of pathologists and is also shifting to value-based care. As a result, fewer people must review a larger number of more complex cases with greater accuracy and at faster speeds. Digital and computational pathology can help the pathologist address this seemingly impossible situation.
By allowing pathologists to work efficiently, collaborate across borders, speed up consults, and uncover new insights, digital pathology can help deliver a more efficient means to support targeted, patient-specific therapy and accurate, first-time-right decision-making. Computational pathology could further enhance these core capabilities with greater accuracy and consistency.
Platforms Advancing the Field
There are a number of digital pathology solutions currently available and many more in development. Many of the available products target one or more of three distinct areas—clinical diagnostics, applications that aid the pathologist in routine diagnosis; molecular pathology; and drug and biomarker discovery (often an image and data management research platform for whole slides and tissue microarrays).
Among the solution providers is Philips, which provides digital pathology software technologies to support the pharma industry and academic researchers in biomarker discovery and drug development. With its Xplore software, an open web-based platform, Philips says it provides a flexible solution for the storage, management, and exploration of whole-slide pathology images and tissue microarrays (TMAs), with the ability to associate additional multimodality data for research purposes. “With powerful data management, scoring, search, and data visualization tools, Xplore will help researchers and pathologists to quantify data more easily, identify trends, outliers, and clinically significant cohorts, and enrich datasets,” Stewart explains.
The open platform structure of Xplore offers integration with third-party image analysis vendors, such as the Definiens Tissue Studio® and Indica Halo®. Additionally, researchers have the flexibility to use images from multiple digital pathology scanner vendors, including the high-quality images from the Philips IntelliSite Pathology Solution, as well as providing support for other common brightfield and fluorescent scanners.
Image: Lab technician reviews slides for scanning. Image courtesy of Philips Digital Pathology Solutions.
Another company in the digital pathology arena is Proscia, which was founded in 2014 by students from Johns Hopkins Whiting School of Engineering. It seeks to improve the efficiency, speed, and quality of pathology diagnostics and research. CEO David West says that the company’s software “enables much of the functionality expected from digital pathology, providing a backbone for digital pathology adoption while leveraging artificial intelligence to power these transformations through automation and augmentation.”
Agilent recently released an online version of its Dako Atlas of Stains that is powered by Proscia’s digital pathology platform. Proscia’s cloud-based platform allows users to view and manipulate the slide images online, as well as annotate them and assess the stains, West explains.
Earlier this month at the European Association of Urology Congress, Chinese scientists from Nanjing Drum Tower Hospital showcased an artificial intelligence system that they said could diagnose and identify cancerous prostate samples as accurately as any pathologist. "This is not going to replace a human pathologist," said research leader Hongqian Guo. "We still need an experienced pathologist to take responsibility for the final diagnosis. What it will do is help pathologists make better, faster diagnosis, as well as eliminating the day-to-day variation in judgment that can creep into human evaluations." Guo also said the software could help overcome any local shortage of trained pathologists.
Guo's group examined 918 prostate whole mount pathology section samples from 283 patients. The pathology images were subdivided into 40,000 smaller samples; 30,000 of these samples were used to “train” the software, the remaining 10,000 were used to test accuracy—the results showed an accurate diagnosis in 99.38% of cases (using a human pathologist as a gold standard), Guo explained.
Additional Resources: A New Frontier in the Fight Against Cancer (Video)
Lung cancer specialist Roy Herbst from Yale Cancer Center joins pathologists from University Health Network, Toronto, and Queen Elizabeth Hospital, U.K., as well as Agilent thought leader Carlos Cardo-Cordon, from Mount Sinai School of Medicine, to speak about the increasing importance of the pathologist in the multidisciplinary team of specialists that collaborate around each cancer patient.