NEW YORK – Digital pathology software startup Aiosyn said that it plans to double its head count and create new product-specific teams as the firm ramps up the commercialization in Europe of its tool to aid the treatment of breast cancer patients.
The Nijmegen, Netherlands-based firm recently announced that its Aiosyn Mitosis Breast application was the first product to receive CE marking under Europe's In Vitro Diagnostic Regulation (IVDR) for the assistance of pathologists in identifying mitotic figures in whole-slide images. The firm said that recent clinical trial results show that the tool can help pathologists to more quickly and consistently identify cells undergoing mitosis for the grading of breast cancer tumors.
David Tellez, Aiosyn's chief technology officer, said that pathologists typically perform a visual assessment of slides or slide images to count cells that are undergoing division, the results of which can help them to determine tumor aggression. However, the work is tedious and prone to variation in counts among different pathologists, while various features in slides are similar in appearance to mitotic figures, he said.
"The product that we developed helps the pathologist to find these mitotic figures," he said. "These are very small patterns that are hard to find and easy to miss in the images."
Tellez said that Aiosyn's initial focus is on the commercialization of Aiosyn Mitosis Breast in the EU and other countries that accept the CE mark. Its commercialization plans include leveraging its relationships in the research community, finding opportunities through existing customers, and identifying interest among other healthcare institutions, especially among providers that tend to be early adopters of cutting-edge technologies.
Later in 2025, though, the firm plans to begin the process of securing US Food and Drug Administration marketing clearance for the test. Tellez said that most of the inquiries that the company is receiving about its software is coming from potential US customers, so the firm has moved up from 2026 its plans to begin the work necessary for a US launch. He noted that the company does not yet know what studies or data will be required by the FDA, but he said that the company has regulatory experience from its work in the EU.
Meanwhile, he said that the company sees opportunities to begin working with early adopters in the US by offering Aiosyn Mitosis Breast as a research-use-only tool. That could help the firm to build trust in its algorithm and overcome any technical hurdles to integration with software platforms that are used by those customers.
As a startup with 15 employees, Aiosyn recently started a funding round that could support doubling the size of that team, according to Tellez. An expanded roster would include separate teams for its three product lines with staff for regulatory affairs, sales, and business development on each team.
In addition to Aiosyn Mitosis Breast, the firm has developed an automated quality control application that is used to detect scanning artifacts in digital images of histology slides. It also offers a research-use tool to analyze renal biopsies and a research-use tool to analyze the quality of those renal biopsy images.
Tellez said that those expansion plans will depend on the amount of funding that the company secures. He declined to provide additional details on the funding round.
Aiosyn isn't the only firm in the EU that has developed software to aid the mitosis detection. Paige has also secured CE marking under the old In Vitro Diagnostic and UK Conformity Assessed marking for its expanded Paige Breast Suite, which includes an AI-powered tool for mitotic counting. The company said that it is also in the process of securing CE marking under IVDR.
Matthew Hanna, chair of the College of American Pathologists' Artificial Intelligence Committee, said that a tool that can increase inter-pathologist agreement and draw the eyes to areas with mitotic figures can be helpful for mitotic figure quantification and tumor grading, although how useful each tool is depends on how well it is trained on good data. He was not aware of any products that are FDA cleared for clinical use in the US to identify mitotic figures, although he noted that several firms have developed research-use-only applications that include similar highlighting functions and mitotic figure counts.
Hanna said that some healthcare providers in the US have validated those RUO products for clinical use as laboratory-developed tests.
The Aiosyn Mitosis Breast software is used to overlay green boxes onto slide images to highlight suspected mitotic figures. Tellez said that the company instructs pathologists to treat the results as the findings from another pathologist who also reviewed the slides.
While he noted that pathologists may find that the algorithm can mistakenly label some cells as mitotic figures or miss others, in one study researchers found that the test's performance was on par with the average among a group of practicing pathologists.
In that study, Aiosyn evaluated its performance in comparison with a panel of 10 pathologists who ranged from a few years of experience to 20 years of mitosis counting. Tellez said that the software and pathologists identified similar numbers of mitotic figures in slide images.
In a more recent clinical study in 2023 by researchers from Radboud University Medical Center, Aiosyn, and Linköping University in Sweden, 28 pathologists from nine countries were asked to perform mitosis counting on slides. The pathologists read the slides in phases so that they assessed each image unassisted and separately evaluated them with the help of Aiosyn Mitosis Breast, Tellez said. The pathologists had about 33 percent higher consistency when evaluating reported mitosis counts with the help of the AI-based software, and they had a 16 percent average increase in productivity when examining resections, he said.
He noted that researchers on that study have been working on a manuscript for peer-reviewed publication.
Pathology labs are under pressure to deliver results more quickly, and the efficiency improvements shown in the clinical study will provide a return on investment for labs, Tellez said.
Aiosyn is also offering its software for integration with various digital pathology platforms. Unlike some other developers that have dedicated viewers or image management systems, he said, Aiosyn's team is working with customers to ensure that the software works across the fragmented software ecosystem in labs.
The firm has developed plug-and-play modules for certain platforms such as Sectra Medical's digital pathology system and is working with other platform developers and individual customer labs to ensure that the software works with patient slide images.
"What we pledge is that, if you have a system that is compatible with AI, we're able to deliver on integration and we're able to connect it," Tellez said.