Gehe zu: Thieme
Connecting through Technology

Medical Knowledge meets AI

MORE PRECISE, MORE PERSONALIZED AND MORE TARGETED

Katrin Siems, Senior Executive Vice President, Global Head of Marketing & Sales at Thieme
digital hand and anatomic hand point at Thieme logo

Thieme is one of the leading providers of medical specialist information. We spoke to Katrin Siems, Senior Executive Vice President Marketing and Sales at the Thieme Group, and AI expert Alexander Thamm about the opportunities and prospects of combining quality-assured specialist content and standardized patient data with artificial intelligence.

Alexander Thamm, founder and CEO der Alexander Thamm GmbH

About the person

"We are firmly convinced that with AI we can relieve the burden on those involved in patient care and improve the quality of treatment!"

Katrin Siems heads the AI @Thieme task force. As Senior Executive Vice President, Global Head of Marketing & Sales, her focus is on bringing future-proof digital solutions with concrete customer benefits to the market.

Ms. Siems, how can AI help to bring together Thieme's wealth of knowledge and the specific information needs of various stakeholders in the healthcare sector even better?

Katrin Siems: AI will revolutionize the way in which relevant data and specialist information is collected, processed and made available in the healthcare sector. Artificial intelligence makes it possible to provide quality-assured specialist information on specific patient data. The content can be tailored in a more precise, more personalized and more targeted way with respect to the specific needs of professionals from different healthcare sectors – from doctors, nurses and therapists to administrative staff and management – and made directly available in the care process.

This starts with data collection, for example. In situations, in which there is time pressure or when users prefer to speak rather than type, AI, together with speech recognition, can support the documentation of the patient consultation and medical history. As a result, we can use this information to provide users with exactly the specialist information they need at the relevant point in the treatment process.

Training, further education and knowledge transfer and acquisition can also take on a new quality - by using AI to tailor the learning experience based on users' previous individual learning progress and experience. We also want to achieve a high degree of personalization in specialist research. For a cardiologist researching the topic of heart insuffiency, for example, we can use self-learning algorithms to proactively compile the latest research findings and place them in the current state of research.

Personalization with AI is also crucial for the future provision of relevant specialist information – whether for learning or at the point of care. Users should be able to choose exactly the format that suits them best. For example, a specialist article could become a summarized short text for a quick read in between, a video for the sofa at home or an audio contribution for one`s car ride. Here, AI enables a high degree of automation and personalization. Thieme's offerings can also be made more accessible with AI-supported translations. In the foreseeable future, our customized content will be multilingual and therefore usable worldwide.

Mr. Thamm, you have helped many companies from different industries to integrate AI into products and processes. What challenges do you see for companies in the healthcare sector and what potential do you see at Thieme?

The biggest challenge when using AI is that generic AI language models are trained with unfiltered knowledge from the internet and occasionally hallucinate, i. e. compile information according to probability principles. However, the reliability of information is crucial, especially in the medical environment.

The wealth of knowledge that Thieme has at its disposal gives it a long-term competitive advantage over such generic AI models. The medical information and content from Thieme, which AI accesses and learns from, has been checked and reviewed by experts and prepared in a didactically meaningful way. The second major advantage that Thieme as an information service provider has over other providers of AI-based information is its large network of authors and editorial expertise. Only humans can currently create original new knowledge. AI is not able to do so, as it only learns from what already exists.  

Working with the Thieme employees, I have experienced a great spirit of optimism, the will to change into a tech company that is shaping the future – and at the same time there is also pride in the company`s own history as a leading medical publishing house.

About the person

"The wealth of knowledge that Thieme has at its disposal gives it a long-term competitive advantage over generic AI models."

Alexander Thamm is the founder and CEO of Alexander Thamm GmbH [at], one of the leading providers of data science and artificial intelligence in German-speaking countries.

Where in the healthcare system can AI provide the greatest benefit?

Alexander Thamm: I see a huge opportunity in the fact that an AI that is trained on reliable information can serve as a kind of "medical co-pilot". The AI sorts, filters and personalizes the reliable knowledge and provides the doctor with exactly the information he or she needs in the specific treatment situation. This allows people working in the healthcare sector to focus more on patients and on what sets humans apart from AI: intuition, creativity, interpersonal perception and care.

Anyone who develops digital solutions for the healthcare sector has to meet specific requirements in terms of data protection and the traceability of results. Doesn't this contradict the use of AI – keyword: "black box"?

Katrin Siems: One of Thieme's strengths is the provision of high-quality medical information and specific knowledge for the healthcare industry for a wide range of usage scenarios. We therefore offer the perfect basis for the use of AI. Sensitive patient data have the highest priority for Thieme. We must consider and guarantee their confidentiality and security when designing AI systems. Users' trust in AI models can be built up by not creating a "black box" at all, but by presenting the AI results transparently and comprehensibly.

Alexander Thamm: In principle, every answer from an AI is based on statistical probability. So you will never get the same answer 100 percent of the time. Depending on how an AI is structured, however, it is possible to minimize the deviations. So-called explainable AI, which makes its decision-making processes comprehensible, is one such approach in which we are well ahead in Europe. If Thieme provides its reliable, qualified and yet easy-to-use information to such a comprehensible AI system, this can also be a great advantage.

Will there still be a Thieme product or Thieme application in three or five years' time that can manage without AI?

Katrin Siems: AI will become an integral part of our internal processes as well as our digital solutions, whether through its use in the editorial process or as an integral part of the products themselves. We are firmly convinced that we can use AI to relieve the burden on those involved in patient care and improve the quality of treatment! However, we are not integrating AI as an end in itself, but because we really can use it to achieve better medicine and a healthier life.

The interview was conducted by Dr. Andreas Mehdorn, Consultant Digital Health & Public Affairs, Thieme Communications