AI-driven laboratory diagnostics with medicalvalues

Lab results are often complex and not easy to interpret. For many diseases, a medical diagnosis requires the analysis and combination of different values. That's why one of the themes at the MEDICA LABMED FORUM at MEDICA 2021 highlighted "Integrative and AI-driven diagnostics" - and illustrated how AI can help interpret laboratory results and values.

In his forum presentation, Florian Stumpe showcased the medicalvalues AI-based software platform, which is designed to support decision-making for physicians. In this MEDICA-tradefair.com interview, he talks about the possible applications of AI in laboratory diagnostics.

Mr. Stumpe, which areas or tasks benefit from AI-driven laboratory diagnostics?

Florian Stumpe: Our medicalvalues platform factors in two dimensions that are relevant for physicians: First, the solution aims to provide clinical decision support based on laboratory values. However, the doctor still makes the ultimate medical diagnosis. Second, medicalvalues supports physicians when they request additional laboratory parameters.

Do you focus on specific diseases?

Stumpe: Unlike other companies or start-ups, we cover a wide range of diseases. Of course, there are also special cases that are more intricate though we generally try to provide broad support for the medical and related sectors.

We have had great success as it pertains to rheumatoid arthritis, oncological diagnosis and staging, thyroid disorders, and kidney failure.

What role does medicalvalues play in laboratories?

Stumpe: medicalvalues provides diagnostic knowledge based on clinical guidelines and research literature. Unlike other machine learning algorithms, our solution does not need to be trained first, but can already support the physician in this initial setting. Apart from the decision support system, it also offers integration options for data transfer from a laboratory information system. This solution can be installed on-site, meaning on the servers of the laboratory or hospital facility.

What are the latest trends to watch pertaining to AI in laboratory diagnostics?

Stumpe: In the long run, integrative diagnostics will become increasingly important, whether it relates to the integration of laboratory data or clinical symptoms, or diagnostic imaging. It was one topic that was addressed by MEDICA.

In the short term, AI can help monitor laboratory values that have already been requested in the past. That being said, we must still lay the foundation in Germany when it comes to the application of standards such as LOINC (Logical Observation Identifiers Names and Codes), thus providing a standardized and comparable method to use AI algorithms across different facilities.

What were your key takeaways from MEDICA?

Stumpe: The trade fair facilitated a great exchange with medical specialists and laboratory physicians, which allowed you to network with experts from all aspects of the industry. I also enjoyed listening to presentations from other areas. As a software developer, it enables you to gain valuable insights. Last but not least, it was amazing to visit MEDICA in-person, learn about the latest trends and get market updates.