Unique AI-Based Approach Automates Clinical Analysis of Blood Data

The clinical analysis of blood data, known as cytometry, is a labor-intensive process that is largely subjective, even for the most skilled laboratory staff. Current cytometry-based diagnostics for blood cancer and other immune diseases require doctors and analysts to evaluate complex, high-dimensional data sets. This analysis, which averages around 20 minutes per sample, is not only time-consuming but also faces a significant shortage of trained personnel. Moreover, the process is quite subjective, with approximately 30% variability in analysis between different operators. These challenges have limited the use of cytometry data for more personalized treatment. Now, a cloud-based machine learning platform can help labs manage their caseloads, provide an objective second opinion to every patient, and offer new insights to physicians for tailoring treatments to every patient's unique immune system.

hema.to (Munich, Germany) offers user-friendly software for clinical decision support in blood cancer cases using cytometry data. This artificial intelligence (AI)-powered tool, which is FDA registered and has CE-IVD approval, streamlines the diagnostic workflow, benefiting both diagnosticians and patients. Already implemented in leading hematology labs, the AI software is now being scaled up to support blood cancer diagnostics in laboratories across Europe and demonstrate significant improvements in diagnostic quality.

Image: AI-assisted analysis of single-cell blood data brings precision diagnostics to immune medicine (Photo courtesy of Freepik)

hema.to's proprietary algorithms, developed using its extensive and continuously growing database of diverse cytometry data sources, can predict disease biomarkers directly from the raw data generated by blood measurement devices. This addresses a hitherto unresolved issue caused by the lack of standardized measurement protocols, resulting in complex data variability that previously hampered automation. The company specializes in integrating data from various sources to identify predictive disease biomarkers. This technology has already been incorporated into the regular clinical workflow of two German labs for decision support. hema.to now plans to broaden its user base, expand the range of supported diseases, and enhance the quality of its AI models.

“Europe’s largest leukemia lab had the real need to speed-up their internal analysis workflows and worked with us to build a world-first AI prototype,” said Karsten Miermans, co-founder and CEO of hema.to. After the success of demonstration of AI-assisted clinical cytometry, we noticed that all labs have the same manual workflows and pain points. We founded hema.to two years ago to help labs across the world with their clinical cytometry workflows.”

Related Links:
hema.to