Traditionally, when patients undergo a biopsy, their tissue or cell samples have been sent to a pathology lab, put on blocks, cut into sections, stained with a stain, and then analyzed under a microscope.
Even in the era of digitalization, this process remains relatively unchanged.
“Pathology still lives on glass,” said Brian Rubin, president of the Cleveland Clinic’s Institute of Pathology and Laboratory Medicine.
“One of the really amazing things about pathology today is that it’s pathologists with microscopes and glass slides,” said Dr. Andy Beck, CEO of the platform developer. IA PathAI. “People are amazed that it’s not digital.”
But PathAI and Cleveland Clinic aim to change that. The Northeastern Ohio Medical System and the Boston-based developer of AI and deep learning medical pathology tools have embarked on a five-year research collaboration. This partnership will involve digitizing pathology specimens and linking clinical data with digitized pathology data. PathAI will also develop algorithms based on analysis of laboratory pathology workflows and usage.
Using AI to Heal
The end goal is to help the Cleveland Clinic leverage AI to more quickly identify disease and match patients with the best therapies unique to their condition, Beck said.
“There is enormous potential to create new algorithms in AI-powered pathology that will be useful for research and patient care,” he said.
PathAI’s tools use deep convolutional neural networks (CNNs), which consist of millions of layered, densely interconnected processing nodes. Using tens of millions of parameters, these complex networks can identify patterns in images and videos.
These tools will be applied at the Cleveland Clinic as it executes a five-year plan to digitize 1.5 million slides. The medical system has one of the largest glass databases in the world – what Rubin described as a “fantastic, well-organized library of well-annotated, well-diagnosed slides.” As he explained, the operator of 10 medical facilities in northeast Ohio has been using digital blade technology in a limited way for years.
As part of the effort, the clinic will purchase several new scanners and add a dozen specialists — including analysts to develop methods for transferring anonymized data — to its team of 100 subspecialty pathologists.
“Partnering is really about launching large-scale development that we couldn’t do alone,” Rubin said.
“Our commitment is to provide the best possible care to our patients and it is increasingly clear that AI-powered pathology can radically improve diagnostic accuracy and treatment choice,” he added. “By doing this work, we are able to maximize the value of machine learning for our patients and fuel deeper innovation that can lead to better outcomes.”
Providing information at deeper molecular levels can lead to more accurate diagnoses, Beck agreed.
“In a research setting, it can predict things that you couldn’t predict with the naked eye at all,” he said. “Ultimately, deep learning-based pathology can have a very broad impact on research as well as clinical care.”
Large-scale improved diagnostics
It looks like a win-win, Beck and Rubin agreed. Yet widespread digitization has been limited. The biggest hurdle, Rubin said: additional debt. Pathology is a low-margin business, and creating digital slides requires scanners with an average value of around $500,000, plus upfront software and staff costs. The benefits – at least the perceived ones – have not yet outweighed these costs.
“Whether you’re looking at something under a microscope or on a desktop computer, unless there’s an added benefit and it costs nothing, most people won’t try to get into it big time. scale,” Rubin said. “They must have an incentive.”
The Cleveland Clinic sees a major problem in data portability and the ability to share images. There are also immense opportunities in terms of storage – physical slides are large and the intensive physical space they require could be reduced with digital slides.
And of course, there are far-reaching implications, Rubin said: more accurate predictions, better patient outcomes, better understanding of the molecular underpinnings of disease, information for research and education.
“Pathology is a very subjective practice,” Rubin said. Although highly skilled, pathologists can make mistakes or miss things – that’s why patients are always told to get a second opinion. “It’s nice to be able to just add layers of security, add extra layers of quality,” he said.
In just six years, PathAI has excelled in its niche: the company closed a $165 million Series C funding round in 2021 and also recently acquired Poplar Healthcare Management, one of the largest pathology labs in the country. .
The company is focused on the drug development cycle, with the goal of providing new insights into how pathology relates to drug response, Beck said. PathAI is deploying its systems in clinical trials with various healthcare providers and is also developing new types of tests entirely based on digital images.
The partnership with the Cleveland Clinic promises many new and different opportunities.
For example, Rubin pointed out, pathology requires robust cataloging, annotation, and retrieval methods. Pathologists pull slides from the physical archive all the time – when a patient has a cancer recurrence, for example, the new and old slides are compared. Samples must also be stored safely for varying lengths of time depending on different regulations.
As Rubin said: What are the steps needed to achieve all of this in the digital realm?
The partnership aims to answer these questions, he said, while establishing diagnostic and workflow algorithms and models to take pathology to the next level.
“It’s really new territory,” Rubin said. “It’s exciting for us as academic pathologists to now work with AI/ML developers.”
Beck estimated that only about 5% of labs are fully digital. But he said he was confident the field will be transformed over the next five to 10 years, “Labs will be very different.”
However, achieving this will only happen through partnerships, which requires a significant commitment of time and money for vendors and technology companies. Breaking tradition takes a good reason and a motivating factor, and medical institutions need help recognizing the value of increased quality and reproducibility, Beck said. Tech industry players also need to help healthcare support digitization efforts and make them less cumbersome.
“With scanning alone, the convenience hasn’t been enough to switch between glass slides and microscopes,” Beck said. “But once you’ve applied AI, the value that these digital images can provide increases dramatically.”
“We see an incredible opportunity to accelerate innovation in precision pathology and use our strengths to connect communities across the healthcare ecosystem, including patients, biopharma and academic research,” said said Beck.
VentureBeat’s mission is to be a digital public square for technical decision makers to learn about transformative enterprise technology and conduct transactions. Learn more about membership.