Stefanie Speidel, PhD

Professor Stefanie Speidel was named a MICCAI Fellow in 2023 for her outstanding contributions to surgical data science and to the MICCAI community. She was one of the first researchers in the world to develop innovative approaches in machine learning-based surgical vision in 2005, almost a decade prior to the rest of the field.  

Prof. Speidel is a full professor for “Translational Surgical Oncology” and director at the National Center for Tumor Diseases (NCT/UCC) Dresden since 2017 as well as one of the speakers of the DFG Cluster of Excellence CeTI and the Konrad Zuse AI school SECAI. She received her PhD from Karlsruhe Institute of Technology (KIT) with distinction in 2009. Her current research interests include machine learning for image- and robot-assisted surgery, bridging the gap between medicine, computer science and engineering.

Prof. Speidel is the co-founder and co-chair of the Endoscopic Vision Challenge (EndoVis), the first computer-assisted intervention (CAI) challenge at a MICCAI conference, which was introduced in 2014 and has been held annually ever since. She is also the co-chair of the Surgical Data Science Initiative, a broad international collaboration aiming to improve the quality of interventional healthcare and its value through harmonizing the methods and practices of capturing, organization, analysis and modeling of surgical data. She has also co-authored more than 150 publications.

Prof. Speidel has contributed to MICCAI conferences for over 14 years and her work has been recognized by several best paper awards. She was an Area Chair in 2017, 2018 and 2020, and a MICCAI Program Chair in 2021 and 2022. She is also a member of the MICCAI Society’s board of directors, where she leads the Diversity Working Group. Dr. Speidel has also made fundamental contributions to the Information Processing in Computer-Assisted Interventions (IPCAI) conference having been General Chair in 2020, 2021 and 2022 and Area Chair in 2016 and 2017.

We are very proud to call Prof. Stefanie Speidel a MICCAI Fellow and pleased that she could share more about her research and MICCAI experiences with us in the following interview.

Q. Congratulations on become a MICCAI Fellow in 2023 for your pioneering work in Surgical Data Science! You are one of the first researchers worldwide who developed innovative approaches in machine learning-based surgical vision in 2005. What were some of the key challenges you faced and how did you overcome them?

A.  When I started my PhD years ago, there were no public datasets available for surgical vision, and recording data in the OR was challenging. In the first months, I reached out to companies, hoping to borrow an endoscope, and luckily, one vendor provided me with a stereoendoscope that had been discontinued due to low commercial demand. With limited resources, I designed a phantom using basic materials, taking a hands-on and pragmatic approach. Additionally, I attended surgeries to capture video recordings from the da Vinci system. I had very motivated clinical partners, so we could also do phantom and ex-vivo data capture sessions during the weekends and nights. At that time, machine learning methods were far less advanced than today, relying heavily on handcrafted feature engineering without end-to-end learning, struggling with domain shifts, rather than the powerful deep learning techniques we now have.

Q. What have you observed in the field of computer-assisted intervention as it has changed and progressed over the years?

A. CAI has evolved over the years, driven by advances in AI, robotics, novel sensors or intraoperative imaging. Additionally, data availability for CAI applications has improved, though challenges remain—large-scale, multi-center datasets are still sparse. However, compared to two decades ago, initiatives like EndoVis have fostered data sharing within the community. Despite these advancements, translational success stories in surgery remain limited, and democratizing access to surgical skills is still an ongoing effort. That said, I’m optimistic that we will see embodied AI innovations making a real impact in the operating room in the near future since models are becoming widely accessible, we have a growing data availability and regulatory support, clinicians often act as driving forces for translation and are involved in the development from the beginning and industry investment that accelerates the integration.

Q. With your extensive involvement in the MICCAI conferences from author, reviewer, Area Chair and Program Chair and now as a MICCAI Society board member, what have you enjoyed the most about your experience in the MICCAI community?

A. I've always appreciated the openness and support of the MICCAI community, where researchers from diverse fields come together. Over the years, I've not only collaborated but also built friendships at the conferences—it truly feels like a family.

Q. What do you think are the most exciting opportunities for the future of the MICCAI Society?

A. There are many exciting opportunities already taking shape in various directions, ranging from translation of MICCAI research towards real-world clinical impact, enabling access to MICCAI and AI resources for researchers in LMICs, ensuring that models are trustworthy and generalize well beyond their training data to ensure correct functionality for a wide range of patients, or tackling sustainable computing challenges for healthcare applications.