How should pathology labs embrace digital transformation?

October 23, 2024

“Data, data, data is everywhere. This is all we’re talking about. Generating data is the real key – how do labs go from specimen collection and processing to being known for data generation and information and knowledge information?”

– David McClintock, Chair of Computational Pathology and Artificial Intelligence, Department of Laboratory Medicine and Pathology, Mayo Clinic, at the 2024 Pathology Informatics conference

More than ever before, pathologists are trying to get a handle on all of their lab’s data and harness it in a way that is more beneficial for their patients, employees and laboratories. In the past, punch cards, hard drives and tapes kept the information stored but did not always make for easy or fast retrieval of information. And today, these forms of storage are no longer enough for the vast amount of information that needs to be accessible.

“We were doing punch cards for a long time and those stored about 33 bytes per card,” said McClintock. “If you were to do a whole slide image today on punch cards, you would need a humpback whale’s worth of punch cards to store that one whole slide image.”

Global healthcare data is projected to increase from 2,300 to 10,800 exabytes between 2020 and 2025.¹  This represents an annual growth rate of 36%² – faster than data from other industries – which could explain why more healthcare companies are looking to the cloud to help them better organize and retrieve data.

What is digital pathology?

Digital pathology can enable laboratories to image large numbers of glass pathology slides rapidly at high resolution through a digital scanner and assist in their interpretation through artificial intelligence (AI) derived algorithms for clinical decision support. The digital pathology ecosystem is composed of three parts – information systems; whole slide-imaging systems, which can include scanners, image management software and monitors; and system tools, such as native applications, third-party applications and image analysis. Whole-slide images can be stored in servers or remotely via the cloud, which can allow pathologists to access them remotely.
 

How does digital pathology benefit healthcare?

Pathologists are in high demand as the average number of pathologists hired per practice is increasing.³ Remote access to whole-slide images could allow pathologists to serve more labs or allow labs to employ pathologists who are not physically close to the lab.

Digitization of slides has the potential to transform how pathology services are delivered as labs face higher volumes and complexity in histopathology.⁴  To leverage the full power of digital pathology, the image management system would be integrated with either hospital information systems picture archiving and communication systems, laboratory information systems, electronic medical record and/or radiology information systems. Ultimately, digital pathology may promote patient access to healthcare.

Get started in the transition to digital pathology and AI

McClintock recommends first taking the time to consider the big picture on how healthcare laboratories and clinical laboratories can improve through better data management. He said his team within the Computational Pathology and AI Division at Mayo Clinic has been looking into this in four different ways when analyzing how to use artificial intelligence and its data. They include:

1. Operational Efficiencies - How do we do more with less and improve our operations?
2. Employee Experience - How do we improve our employees’ experience, better retain them and ensure they are working at their maximum efficiency and training level?
3. Patient Experience - How do we improve the quality and accuracy of patient care? 
4. Beyond the Status Quo - How do we go beyond our usual day-to-day operations and innovate using our data in exciting new ways?

Once those questions are answered, there are more operational questions to consider before making decisions. McClintock suggests understanding the data capabilities of the laboratory by getting the answers to key questions about data systems, people and data processes.

McClintock said it’s important for each organization to lay out its data vision and define its own digital transformation.

“I truly believe none of us are going to do it all by ourselves,” McClintock said. “There’s too much to do. None of us are going to control all those petabytes of data and exabytes of data we’re generating. I would encourage you to ask your vendor, ‘How are you helping us create those visions?’”

Sources