6 Practical Tips for Implementing Digital Pathology in Your Lab

After years of development, digital pathology is gaining momentum worldwide as an innovative technology associated with improved laboratory efficiency and productivity. Industry reports document an increasing number of labs worldwide deploying digital pathology. And there is much to learn from the labs at the forefront of this digital transformation.

Portugal-based IMP Diagnostics presents such an example. The high-volume reference laboratory began preparing for digital pathology almost a decade ago and initiated deployment in 2020. Now fully digital, IMP’s co-CEOs Isabel Macedo Pinto, MD, and João Sousa Pimentel, MBA, share the top six tips for going digital.

1. Create and empower a multidisciplinary project team

Long-term success requires an inclusive approach from the get-go. The team doesn’t need to be big, but it does need to be diverse. Include an overall project leader, a clinical lead to involve pathologists’ insights from the beginning, an IT representative to serve as a point person with tech vendors, and a business lead to oversee financial investments.

2. Visit pathology labs that have gone digital

Forget what you think you know about digital pathology; be open to learning more. Visit labs that use digital pathology to handle their routine workload. Doing so will help the project team identify and address key considerations, such as workflow modifications. The IMP team also found such visits provided critical feedback on vendor selection, particularly about their support in implementing digital pathology in a lab. 

3. Monitor digital pathology advances

Every lab has a distinct set of circumstances that will nudge them to go digital. Stay abreast of key milestones, such as hardware and software development, release of regulatory actions, and publication of clinical evidence to help identify the right time for your organization. IMP decided to shift from preparation to implementation based on clinical evidence; in this case, a peer-reviewed paper that documented the performance of a clinical grade-AI algorithm, affirmative regulatory actions for the software referenced in the paper, and the launch of a first-in-class high-throughput scanner.

     

4. Start small when you go live

While a one-time, universal transition across the organization may sound efficient, IMP underwent a limited roll-out to allow for a trial/test period, learning, and refinement. IMP’s structure organically facilitated this approach; they confined the initial roll-out to an R&D department test project and its subspecialty unit structure. Every organization can and should define a narrow scope for initial deployment.

5. Optimize pre-analytical steps

IMP’s phased roll-out revealed the importance of the pre-analytical steps. The organization now advocates a four-pronged approach to obtaining high-quality digital pathology images:

• Enforce adequate slide labeling. Mitigate skipped barcode cases in archived slides with labels by engraving slides with QR codes, which leads to fewer scanning errors. 
• Ensure slides have correctly aligned coverslips and are free of excess mounting media. Common challenges include air bubbles and excess mounting medium spread beyond the coverslip, which can interfere with automated slide handling. IMP addressed these issues by integrating automated equipment into their histology workflow to obtain consistent staining and coverslipping before digital processing.
• Make sure the slides are clean and placed correctly. Minimize duplication and failed scans by waiting until the slides are fully dry before loading them into the scanner and ensuring they are placed flat in the scanner rack.
• Choose the scanner location wisely. For a lean process, situate the scanners to allow for a seamless workflow in the histology laboratory. Additionally, ensure that the scanners are placed on a flat surface with as little vibration as possible.

6. Modify existing training guidelines to suit your needs 

The process of training pathologists to review cases digitally is critical, but it can be designed to facilitate easy and fast learning. IMP adapted the Royal College of Pathologists Guidelines to create a two-track training and validation process that respected the unique needs of its organizations and pathologists. For example, because their pathologists review cases in more than one subspecialty, IMP opted for a mixed initial archive validation set of 15–20 cases, representing the most common diagnostic challenges in each of those subspecialties.

In summary

Now, three years into deployment, IMP is digitizing 100 percent of its histology slides; their pathologists are validating the use of whole-slide images for primary diagnosis. IMP’s pathologists report that digital pathology provides tremendous value once fully implemented, including the ability to review cases remotely; obtain second opinions; access computational tools to assist with tasks, such as cell counting and immunohistochemistry scoring; and, in the near future, utilize AI-driven applications to unlock novel diagnostic insights. 

While we are still in the early days of digital pathology, IMP’s experience is encouraging and informative, providing pathologists and pathology labs real-world experience on how to implement this technology, reap significant benefits now, and lay the groundwork for a promising future.