A Year of Excellence: Celebrating Prof. Dr. Lena Maier-Hein’s Contributions to Medical Imaging & AI in Oncology

We are proud to celebrate an outstanding year of achievements by Prof. Dr. Lena Maier-Hein, one of the Helmholtz Imaging Center Coordinators and a distinguished leader in oncological research and imaging at the DKFZ German Cancer Research Center. In 2024, Maier-Hein was distinguished with the German Cancer Award as well as the Baden-Württemberg State Research Prize for excellence in applied research with a grant of 100,000 euros. Both awards acknowledge Lena’s pioneering work in developing and validating innovative AI-based imaging methods. Further affirming her contributions, the European Research Council awarded her both a Starting Grant, which supports promising young researchers, and a Consolidator Grant, to support her ambitious project aimed at neural spectral image decoding. These major achievements will surely reinforce her status as a role model for young scientists. We sat down with Lena to discuss her recent achievements, and her future plans.

Congratulations on winning the German Cancer Award 2024 and the Baden-Württemberg State Research Prize, Lena! Can you share what these recognitions mean to you and your team at the DKFZ and NCT Heidelberg?

Thank you! This recognition means a lot to my team and me, as it demonstrates we have been on the right track all along. I entered the field of AI research more than 20 years ago during my computer science studies at KIT and Imperial College London, back when skepticism prevailed – now we are finally reaping the rewards.

Receiving the ERC Consolidator Grant is a significant milestone. Could you elaborate on how your approach differs from current practices in surgical imaging?

Current practice in surgical and interventional imaging heavily relies on ionic radiation which comes with disadvantages for patients and clinical staff alike, and harbors a high potential for complications. Moreover, current imaging modalities do not allow surgeons to obtain functional information, such as oxygenation parameters, on the tissue they handle. Yet, especially in minimally-invasive surgery, accessing this information in real time is crucial to ensure optimum performance. In close collaboration with clinical partners, we have developed novel imaging techniques that, using special spectral cameras and machine learning, enable real-time, completely radiation-free visualization inside the human body—making information invisible to the human eye visible. I hope that this way we contributed to paving the way for next-generation surgical imaging.

How do you balance the technical aspects of your work with its clinical applications?

Ensuring that technical advances ultimately have practice-changing impact is a goal of the highest priority for our work. To this end, we always work in close collaboration with our clinical partners. Technical and clinical aspects also synergize in our Helmholtz Imaging research on validation strategies of AI algorithms for image analysis, where, for example, clinical functions need to be reflected in the selection of technical validation metrics.

With your ERC Grant, you aim to advance multispectral optical and photoacoustic imaging, and integrate AI in medical procedures. How do you envision these technologies, particularly AI-based imaging, impacting patient outcomes in oncology and improving cancer surgeries in the near future?

Future surgeons will have unprecedented insights into functional tissue parameters, such as the oxygenation level, of the tissue they are operating on – reliably and in real time. Globally, this will hopefully enable a new generation of safer and more effective interventional oncology, with reduced post-and intraoperative complication levels that are less dependent on an individual surgeon’s expertise.

How do you envision the future of AI in medical diagnostics and treatment?

Medical AI is developing at a rapid pace. The recent rise of artificial general intelligence (AGI) – that is, singular intelligent systems that can solve many different tasks independently and flexibly, rather than narrow AI models trained only for one specific task – will enable the integration of multiple different data sources such as imaging, genetic and clinical patient data to establish a personalized treatment plan. Medical AGIs will be able to autonomously develop hypotheses and interpret data, allowing not only for more effective treatment but also promising new and efficient prevention strategies.

What is the impact of large well-curated datasets in this context, and are these datasets sufficiently accessible? If not, what strategies should be implemented to improve their availability?

Large, well-curated datasets are transformative for machine learning, enabling more accurate, robust, and generalizable results while fostering reproducibility and trust in research. However, access to such datasets is often limited by costs, privacy concerns, proprietary restrictions, and infrastructural challenges. To address these barriers, strategies like promoting open data initiatives and research on data-centric AI are needed. Helmholtz Imaging, for example, will soon launch a dedicated Data Curation Unit to address some of these issues. We are also actively setting up high-quality benchmarks in various socioeconomically relevant domains.

What role do your team members play in your projects, and how do you foster collaboration?

I am fortunate enough to be able to rely on a team of outstanding, highly motivated researchers who share my passion for medical AI research and without whom I would be entirely lost. I would like to specifically highlight Dr. Alex Seitel who has been working with me for almost two decades now and who is the best deputy I could imagine.

What advice would you give to young researchers aspiring to contribute to medical technology?

My advice is: Don’t be afraid to challenge – literally “re-search” common practice in your field. Question everything! Don’t lose sight of the ultimate goal – advancing medical technology should always be focused on achieving patient benefit. To generate tangible clinical success, make sure to involve all technical, clinical, societal and administrative stakeholders – transdisciplinarity is the way to go.

Thank you so much for your time, Lena!

Lena’s work in AI for imaging analysis is enhancing surgical precision and driving advancements in medical technology. Her innovative approach is improving the accuracy of cancer surgeries and opening new possibilities for treating other diseases. A dedicated scientist and role model, Lena’s contributions inspire the next generation of researchers and promise a brighter future for medical imaging and oncology. We look forward to the ongoing impact of Lena’s research and the innovations her team will contribute to oncology and medical imaging.