The Future of Digital Pathology

By Jordan Rosenfeld - June 18, 2024

19-240519-JP_Journals_Critical Values_CV

Healthcare institutions worldwide are increasingly adopting digital pathology to enhance accuracy, speed up diagnoses, and improve data storage. With significant advances in artificial intelligence and machine learning, digital pathology is changing fast, presenting both opportunities and challenges for pathologists. Experts explain the state of digital pathology today and explain why the time to embrace digital pathology is now, even in its early days. 

Taking the pulse of digital path 

For those who haven’t yet adopted much digital pathology technology, it may still feel like a strange new world. However, according to Jennifer Picarsic, MD, Director of the Cincinnati Children's Digital Pathology and Anatomic Pathology Informatics and Co-director of the Cincinnati Children’s Histiocytosis Center, “Digital pathology is not just the scanning of the slides,” she says, but rather “a whole process” that begins at electronic ordering of tissue exams in the OR with barcoding that follows that specimen through the lab throughout its journey to the pathologist’s review and into the electronic medical record. 

“The integration of the parts is what I see as full digital pathology—the ability to really increase efficiency in our everyday tasks,” Dr. Picarsic says. 

And even in the past five to 10 years, what seemed a futuristic concept is now becoming a reality, according to Dr. Anil Parwani, M.D., PhD, Director of Anatomic Pathology in the Pathology Department at Wexner Medical Center, The Ohio State University. “They now have high throughput robotic scanners that are approved by the FDA for primary diagnosis, even for specialized areas in pathology like cytopathology, renal pathology, hematopathology—things that were a barrier, and can now routinely digitize a slide in under a minute,” he says. 

Additionally, Dr. Parwani points out, there are CPT codes for billing around digital pathology, more image management systems in place that can connect the scanner to a pathologist workstation, and more integration with lab information systems.  

“So today, in 2024, we have more choices as a consumer in terms of hardware and software,” Dr. Parwani says. 

Rather than being nervous to adopt the technology, Dr. Picarsic suggests that investing in digital pathology, “[Is] really a strategic investment for how we look at the future.” 

While there isn’t “critical mass adoption” yet, Dr. Parwani says, “the fully adopted labs are now reaping the benefits of all this data that they accumulated.”  

He points out that in his laboratory they have already uploaded four million images. “That’s enough to start building your own model for predicting what type of a cancer is. It's without even doing any testing.” 

Dr. Picarsic understands that adopting new technology takes time, but she hopes that pathologists will soon see the benefits to their labs and their patients and come on board.  

Digital pathology improves efficiencies 

Digital pathology has the power to improve efficiency in everything from workflows to patient care, Dr. Picarsic explains. 

“If I have a complicated case and I need pathologist X to look at this case, but I also want to show pathologist Y and Z, now I don't have to wait all day until pathologist X gets to it or wait for a consensus conference. Rather, I can show these digital images in real time to more than one pathologist at the same time for enhanced quality assurance. There is also increased flexibility in asking specific slide level questions with image labeling and annotations; efficient second opinions and improved quality assurance in our everyday work environment is just the start.” 

Digital pathology improves patient care 

Additionally, by both granting greater access to a multitude of pathology opinions digitally, there is a reduction in the need for shipping glass slides, as digital pathology can speed up the time of an expert diagnostic review, from across the street to across the ocean. 

“For me, all of this is to improve patient care,” Dr. Picarsic says. “We’re getting more accurate reports to the patients via the clinicians who are making timely and accurate treatment decisions based on our critical interpretations. So, to me this is a no brainer.”  

With pathologists fighting increasingly greater workloads and workflow barriers with a shrinking workforce, Dr. Picarsic also sees digital pathology as a way to lighten that load. Which also means advocating for it within your institution or laboratory. 

“We have to really be at the table with the C-Suite/Executive team discussing why this is important for future patient care; if we don't do it as laboratory professionals someone else is going to take this from us. This is our opportunity to show patients how important pathology is to their care,” she says. 

The role of AI in digital pathology 

Another major advancement in digital pathology is the increasing sophistication of artificial intelligence (AI) and machine learning algorithms. 

“The first AI algorithm for prostate cancer detection was approved by the FDA,” says Dr. Parwani. While it’s not yet reimbursable, he says it’s, “a good tool to decision support, to screen your cases, to check your work, and to train residents.”  

AI can streamline tasks and improve diagnosis 

AI is especially helpful in the areas where pathologists do a lot of manual tasks, Dr. Parwani explains. “Especially where we… count cells, we count nuclei, we count mitosis, we look for mitosis, we look for microorganisms such as H. pylori.”  

Pathologists sometimes need special assays and special tests to do these counts, and it is time-consuming work. “But with AI apps, those possibilities have become available,” he says. 

Additionally, he finds that using AI for such tasks gives him more objectivity in his diagnoses. “Instead of saying 20 percent to 40 percent of the cells are positive for this biomarker, I can say 23.6 percent of the biomarkers are positive.” 

Being able to give an even more specific number can make treatments much more specific, too, he says. “That estimate can vary between the pathologists, between labs, and oftentimes these are expensive drugs that you're giving a patient and it's dependent on that range. Sometimes the patient may not even need that drug or may need higher doses of it.” 

The next decade of pathology 

If digital pathology is already doing amazing things now, the next decade to 20 years could be nothing short of astonishing. At the least, Dr. Parwani says, “I think we are going to have more imaging in pathology, not just digital whole slide imaging, which is converting a microscopic slide to a whole slide image. We will see a spectrum of images all the way from the gross room where a specimen is examined all the way to histology. You may not even need to make glass slides anymore. You'll go directly from the tissue to an H&E image.” 

Additionally, AI may be able to generate images itself based on patient samples and data, what Dr. Parwani calls “artificially related images,” such as virtual staining and predicting biomarkers.  

He also thinks there will be more overall adoption of the technology and tighter integration with EMRs and laboratory information systems. “We'll have almost dashboards of information widgets on your smartphone, so you can see the patient's information in a nutshell.” 

“It is going to help experts become more attuned to what the most important aspects of disease are for that particular patient,” Dr. Picarsic says. 

She’s not worried about it rendering pathologists less necessary, either. Instead, she says, “We're going to have sophisticated pathologists that have tools to help them predict what cancer is going to respond to this treatment, what cancer needs additional testing, and how to apply this testing in the best manner for patient benefit.” 

Dr. Parwani sees digital pathology as having an ultimate trajectory of more “patient-centric diagnoses, not case-centric” and adds, “I think it'll become a standard of care.” 

Why pathologists should embrace digital pathology 

“Those at the centers that understand this technology will really help drive the new forefront of innovation in pathology,” Dr. Picarsic insists. 

Dr. Parwani agrees. “Don't be afraid of the technology. This is a technology that will help you, will be good for patient safety, and patient care. It'll be good for potentially more revenue, more collaborations in the future when we do have the CPT codes and the reimbursement.” 

Those who don’t prepare for it now will be “lagging,” Dr. Parwani says.  

Those who understand its limitations and its uses, and overcome the barriers to implementation, will be well poised to deliver the best possible patient care now and in the future. 

 

 

Jordan Rosenfeld

Contributing Writer