Programme Thursday 6 July 2023

Dr. Paul Elbers, intensivist and Associate Professor of Intensive Care Medicine, Amsterdam UMC, location VUMC

How to get advanced models from bytes to the bedside of critically ill patients.

Paul Elbers is intensivist and associate professor of intensive care medicine at the Amsterdam UMC and is chair elect of the data science section of the European Society of Intensive Care Medicine. Paul Elbers has extensive (research) experience in co-developing state-of-the-art AI solution at his department, and will present his work on implementing an antibiotic treatment decision support tool and other clinical applications of AI in his field.

Drs. Bas Becker , PhD candidate Organizational Science at Vrije Universiteit Amsterdam

Numerous studies highlight the potential advantages of clinical decision support systems (CDSS) in enhancing medical decision-making in laboratory settings. However, when CDSS are implemented in real-world care environments, such systems struggle to deliver on their intended promises, likely due to flawed expectations of clinical reality used to inform CDSS design. Such expectations are fuelled by notions that medical decision-making processes unfold in a structured, linear, and predictable fashion, a perspective incongruent with the dynamic nature of real-world medical decision-making. In the talk, I share insights about decision-making in ICU care after spending 26 months observing and interviewing physicians and nurses performing their daily work. By offering a renewed understanding of the ‘anatomy’ of medical decision-making in the ICU, we will conclude by discussing its implications for CDSS and AI technology design.

Dr. Tanja Alderliesten, Associate Professor at Leiden University Medical Center, department of Radiation Oncology 

In this presentation, I will share our multidisciplinary journey that started with the identification of challenges in the clinical practice of internal radiation therapy that we wanted to overcome, and next took us from obtaining funding, performing the required research, and finally to implementation in clinical practice.

At the basis of this journey lies a collaboration between the national research institute for mathematics and computer science in the Netherlands (CWI: Centrum Wiskunde & Informatica), the Amsterdam UMC – location AMC, and industry partner Elekta. With financial support from the Dutch Research Council (NWO: Nederlandse Organisatie voor Wetenschappelijk Onderzoek) and Elekta, we successfully developed an AI-based treatment planning approach for internal radiation therapy for prostate cancer.

In a follow-up project, funded by the Dutch Cancer Society and Elekta, we are further developing and extending our AI-based approach to cervical cancer treatment. Clinical translation potential will be studied through nation-wide validation studies.

Simone Cammel, Cloud Customer Engineer Data Analytics and AI @ Google Cloud-Innovation for Good

With a background in the LUMC, bringing AI to the patient bedside, and now a Data & AI Specialist for Google, Simone will take us through the existing challenges and solutions of bringing clinical AI into the real world, outside of a research setting. Taking the best practices from Google’s own AI methodology and applying those to the healthcare domain. We will also look at the future of AI in medicine, with the rise of the Large Language Models such as ChatGPT that bring a new technical disruption.