The era of computational pathology is upon us. Driven by a growing base of evidence on the economic and clinical benefits of adoption, leading laboratories are increasingly deploying AI in routine practice to optimize operations and improve patient care. Detection solutions are now helping pathologists improve quality by identifying cancer foci and providing a second set of eyes, and workflow solutions are triaging and sorting cases to drive critical efficiency gains.
This accelerating momentum is creating a newfound sense of enthusiasm, inspiring some laboratories while increasing the pressure on others to keep up. It’s also fueling the hype around AI – and, sometimes, proliferating misconceptions that may make adoption seem overwhelming, especially for the organizations most at risk of falling behind.
AI is poised to make the biggest impact on pathology since the introduction of the light microscope over a century ago, so it’s time to dispel these myths. The right understanding can empower all laboratories to chart their path in computational pathology, setting everyone up to realize its potential today and in the future.
Myth: Without a clear vision, my laboratory isn’t ready to adopt AI.
It’s easy to understand why this misconception exists. After all, you know your laboratory’s needs better than anyone, so you might assume that you’re responsible for framing up your adoption plan – but you may not have a clear vision in mind. Perhaps you’re even wondering how you’re supposed to know all the answers when dealing with such a novel technology – and you’re not alone.
The reality is that you only need to come to the table with an interest in adopting AI. The right AI provider will jump at this enthusiasm and leverage their experience to educate and guide you toward unlocking new value with AI in a way that works for your organization. Importantly, this includes much more than just helping you select and deploy technology. As with all digital transformation, the shift to computational pathology also requires adapting your processes and empowering your team. You should count on your AI provider to serve as a guide throughout all stages of this journey, working with you to establish a business case, build internal support, identify use cases, and configure your workflows at the very least.
Myth: Deploying AI is too hard for my laboratory.
This misconception stems from a somewhat outdated truth. Implementing computational applications into the day-to-day workflow has historically been a challenge, especially when you consider that these solutions add the most value when incorporated into routine operations. As we’ve already noted, though, leading laboratories are increasingly running AI as part of their regular practice, and we can look to their approach to dispel this myth.
We know that computational applications can only be applied to whole-slide images, not glass microscope slides. It follows that any laboratory implementing AI in routine operations has already made the shift to digital pathology. Just as these laboratories require a digital platform to carry out their day-to-day work at scale, they also rely on it to deploy AI. This platform serves as a launchpad for all of their computational solutions so that they can seamlessly integrate them into daily practice.
By leveraging an open, interoperable platform, these laboratories – and your laboratory – can easily introduce new AI applications as needs change and technology evolves. You’ll gain the freedom to choose other solutions, from laboratory information systems to whole-slide scanners, and incorporate them into routine operations. In doing so, you can create a connected digital ecosystem that will scale with your laboratory and increasingly empower your team.
Myth: Our pathologists don’t want to use AI.
This myth probably doesn’t need much introduction. Ever since computational applications were nothing more than a pipe dream, pathologists have raised concerns that AI would replace them. More recently, they are also questioning whether AI will disrupt their familiar workflows.
You likely recognize that the growing body of evidence suggests otherwise, making this myth easy to dispel. What can be a bit more challenging, however, is determining how best to approach your pathologists to secure their buy-in and adopt AI in practice.
Keep in mind that your pathologists are the primary users of the computational solutions you’re looking to implement. You’ll want to put them at the center of the shift to AI, taking the time to hear their voices and educate them on the benefits that lie ahead. Let them weigh in as you determine your use cases and configure your workflows. You should take the same approach if your organization is also in the process of adopting digital pathology, because this, too, is a big transformation for your laboratory professionals. They should be included as primary decision makers when selecting a digital pathology platform. Because this platform powers all your operations, including your AI-powered workflows, it is critical that its users feel it is designed with them in mind.
Finally, it’s important to remember that you don’t have to navigate this journey alone. The right partner will be armed with best practices for winning over your team and provide the insight you need to overcome any objections.
As computational pathology quickly extends beyond academic medical centers and large laboratory networks, it’s time for your organization to take advantage of the unprecedented value ahead. With the right understanding, the right partner, and the right technology, you can go beyond the hype to adopt AI in a way that works for your laboratory – and chart your course as a leader in this new wave of diagnostic medicine.
Chief Product Officer at Proscia Inc., Philadelphia, Pennsylvania, USA.
