From Pixels to Precision 001: Making Sense of the Digital Pathology Ecosystem- Part 1

Digital pathology is a complex field. There are a multitude of vendors across the entire ecosystem that span many different verticals and it’s quite easy to mix up who is doing what. In this two-part series on the digital pathology market ecosystem, we will break down the market and provide an overview of how different companies, including PathAI, fit into this ecosystem. This week, we will dive into setting up a basic framework for understanding digital pathology.



A digital pathology workflow consists of five major components: a laboratory information system, a whole slide scanner, an image management system, artificial intelligence applications, and data storage. For a highly functioning digital pathology workflow, all five of these components must be deeply interoperable and integrated to drive seamless use and review of cases by pathologists daily without compromising safety. We’ll do a deeper dive into interoperability and integration in a later post.

• • The laboratory information system is the centerpiece of the digital laboratory. Think of it as the central nervous system that is dictating and informing what the other components of the digital workflow are doing. A laboratory information system is a sophisticated software system that orchestrates the management of critical laboratory data, encompassing patient information, testing requests and results, and enabling data integrity and operational efficiencies.
  • The whole slide scanner is the cornerstone for digitalization. It’s the hardware that is the first step in the digitization workflow, generating a high-resolution image within seconds to minutes from an unstained or stained-glass slide. Whole slide scanners can be subdivided into high-throughput scanners or low-throughput scanners. High-throughput scanners are more expensive but are often a necessary purchase for any laboratory considering at-scale digital deployment. On the other hand, low-throughput scanners are cheaper and are often used for select use cases in digital pathology such as remote consults, frozen sections, and teaching/education.
  • Once the glass slide is scanned, the whole slide image will be uploaded to an image management system, which serves as a temporary digital repository for pathology images that pathologists can use to retrieve and search cases, view whole slide images for analysis and run embedded artificial intelligence-based applications.
  • With the creation of a digital image, algorithms developed using artificial intelligence can be applied to the images to augment and enrich pathologist decision making. Use cases can include case triage and workload balancing, quality control, tumor detection, biomarker quantification, disease severity assessment, biomarker prediction, and prognostication. Artificial intelligence embedded directly into the image management system can also help drive more value upstream in a laboratory workflow and help to build the business case for digitalization.
  • After the images are reviewed and artificial intelligence is run on the images, the slides must be stored in either on-premise storage or cloud storage. This robust storage infrastructure accommodates the safe retention of patient images and associated data. More and more laboratories are moving to cloud-based tools and storage so that they do not have to invest in data centers and infrastructure on their own.

We’ll use this framework to do a deep dive into each component over the next few weeks. We hope that this framework is useful to both old and new entrants evaluating digital pathology solutions and considering wide-scale adoption and deployment in anatomic pathology laboratories.