Speaker

Title

IHI and the expectations of IHI towards the EHDS

About Dr. Niklas Blomberg

Dr. Niklas Blomberg joined IHI as Executive Director in January 2024. He brings to the role extensive experience in both research and leadership roles in the life sciences.

A Swedish national, he has a bachelor’s degree in chemistry from the University of Gothenburg, and a PhD in structural biology and bioinformatics from the European Molecular Biology Laboratory (EMBL) in Germany.

He worked as a research scientist for AstraZeneca in Sweden for 14 years, taking on increasingly senior roles in the company, and leading the establishment of a team for data driven drug discovery in the respiratory and inflammation fields. During this time, he was industry co-lead of Open PHACTS, a project funded by the Innovative Medicines Initiative (IMI), the forerunner to IHI.

In 2013, he joined the fledgling research infrastructure ELIXIR. As its founding director, he oversaw the final negotiations between the member states to formally establish and launch ELIXIR. Under his leadership, the organisation built up its operational and project management capacity, grew to include 23 member states, and secured funding from both the member states and an extensive portfolio of EU-funded projects, including projects from IMI2, the European Open Science Cloud and the European Genomic Data Infrastructure.

Title

Federated Learning in Healthcare: A New Era of AI-Driven Scientific Discovery

Abstract

Federated learning (FL) is revolutionizing healthcare research by enabling collaborative research across institutions where algorithms are shared rather than data, ensuring patient privacy and security. By breaking down data silos, FL greatly facilitates access to diverse, real-world datasets, enhancing robustness and generalizability of AI-driven insights. Effective federated research requires the seamless integration of a scalable technology, collaborative science and data and agile algorithm development. This symbiotic trifecta will greatly enable accelerated scientific discovery and healthcare innovation, opening new frontiers in medical science while preserving individual contributors’ data control. In this session, we will explore how aligning these elements can drive impactful research and transform healthcare delivery.

About Prof. Dr. Frederik Buijs

Dr. Frederik Buijs is a senior global medical director at Roche where he leads initiatives leveraging federated and GenAI approaches to enable data-driven research unraveling the complexities of neurological diseases and improving healthcare delivery. He holds a PhD in neuroscience and an MD specializing in neurology. He worked in healthtech developing real world data platforms and studies, focussing on peri-launch evidence generation strategies, and continued that work at Roche.

Title

Federated Computing as a Catalyst for Scientific Research

Abstract

In this talk, we will explore how the Rhino Federated Computing Platform (FCP) is revolutionizing scientific research by enabling secure, collaborative data analysis across decentralized datasets. Traditional data centralization poses challenges related to privacy, security, and compliance, particularly in sensitive fields like healthcare and finance. Rhino FCP addresses these issues by leveraging federated learning and edge computing technologies, allowing institutions to perform advanced analytics and AI model training without transferring raw data. This approach not only preserves data sovereignty but also accelerates research by facilitating seamless collaboration among diverse stakeholders. We will discuss real-world applications where Rhino FCP has been instrumental, such as in drug development and outcome research, highlighting its impact on data-driven innovation.

About Dr. Ittai Dayan

Dr. Ittai Dayan is the Co-founder and CEO of Rhino FCP, leading the development of federated computing for regulated industries. Rhino’s Federated Computing Platform enables hospitals, payers, biopharma, and financial institutions to collaborate on AI-driven insights without compromising data privacy, using edge computing, federated learning, and advanced privacy methods.

A pioneer in privacy-preserving AI, Dr. Dayan led a global consortium in 2020 that published a landmark Nature Medicine study on federated learning in healthcare. Previously, he served as Executive Director at Mass General Brigham’s Center for Data Science, integrating AI into clinical workflows, and held leadership roles at BCG, focusing on AI applications in healthcare.

Dr. Dayan earned his MD and BSc from The Hebrew University of Jerusalem and an MPH from Johns Hopkins. He serves on the Nature Digital Medicine Editorial Board.

Title

Biotech’s Open Rebellion: How AI Agents Are Breaking the Patent Chains and are Building a Personalized Future

Abstract

Beginning 2025, DeepSeek stunned the world, vaporizing billion-dollar bets on Big Tech’s closed AI models, an open-source thunderclap that laid bare the power of shared innovation. Biotech’s own DeepSeek moment is closer than we think, driven by AI’s relentless push to democratize knowledge and data. Open-source models and networked AI agents, autonomous, collaborative, and fiercely adaptive, are breaking the chokehold of patented exclusivity, making tools like Evo 2 a game-changer. This bio-medical model doesn’t just accelerate discovery; it rewires control, empowering "patients-like-me" communities to turn data into a shared resource, much like a community library where insights are co-created. While models such as EVO-2 won’t alter legal ownership of patient data, it’s giving patients unprecedented agency, mirroring the open-source LLM revolution, where models like DeepSeek fueled collaborative ecosystems. Patients could license their derived insights under open-source terms, accelerating innovation and eroding privacy-by-exclusivity. From real-time mRNA vaccines to bespoke CAR-T therapies and gene edits on demand, value shifts from IP to proximity. In 5 to 10 years, AI agents could orchestrate decentralized labs, syncing researchers, clinicians, and patient networks to deliver personalized medicine at scale, sidestepping Big Pharma’s gatekeepers. This keynote will challenge the audience to envision a biotech landscape where collaboration trumps competition, adaptability outpaces exclusivity, and the tools of tomorrow are already in everyone’s hands.

About Bart de Witte

Bart de Witte is a visionary leader in digital health and a staunch advocate for open-source innovation. With over 20 years of experience in healthcare technology, including pivotal roles at IBM and SAP, he is recognized as one of Europe’s foremost experts on AI-driven transformation in medicine. De Witte founded the HIPPO AI Foundation, a Berlin-based nonprofit dedicated to democratizing medical AI through open-source frameworks, earning honors such as the 2021 German AI Prize and a finalist position at the 2020 Falling Walls conference. He recently co-founded Isaree AI, a company focused on the safe and secure deployment of AI agents and assistants in healthcare, further advancing his mission to revolutionize the industry. A sought-after lecturer at universities across Germany, Belgium, Switzerland, Austria, and China, De Witte challenges conventional models, pushing for a future where technology drives equitable, sustainable healthcare for all.

Title

TBA

Abstract

TBA

About Dr. Marc Jacobs

Dr. Marc Jacobs is the Head of the Group Software and Scientific Computing at the Fraunhofer Institute for Algorithms and Scientific Computing (SCAI) since 2019. He has been a Group Leader in Scientific Software Development and a Principal Scientist at the same institute since 2000. Prior to this, he worked as a Research Scientist in the Computer Science Department and lectured on Algorithms and Cheminformatics at the University of Bonn, Germany, from 1998 to 2000.

Dr. Jacobs holds a PhD in Computer Science from the University of Bonn, obtained in 2004, and a Diploma in Computer Science from the same institution, which he completed in 1998. He possesses strong leadership and project management skills, with over 10 years of experience in machine learning and artificial intelligence.

Additionally, he has more than 10 years of expertise in scientific software development, high-performance computing (HPC), and cloud computing. Dr. Jacobs has coordinated work packages in international research projects and has been involved in project coordination within German national funding schemes.

Title

Using AI for streamlining health data discovery, access application and release processes for EHDS regulated research

Abstract

Access to data is the fuel for AI development. The European Health Data Space (EHDS2) regulation aims to facilitate the secondary use of un-consented healthcare data for cross-border research within the EU, under strict protocols for data request approvals, access applications, and releases. Due to EHDS2, number of data queries coming from other EU countries are expected to grow significantly, while mandating timely responses. Meanwhile, AI – advancing at an unprecedented pace – has the potential to optimize and accelerate these processes.  In this presentation, we model EHDS2 processes, introduce a reference architecture, identify potential bottlenecks, and explore how AI can be leveraged to streamline data access workflows both now and in the future.

About Timo Kanninen

Timo Kanninen is CSO and founder of BC Platforms AG. He is the visionary behind BC Platforms’ data management systems. He has long-term experience in software development, genetic epidemiology, statistical genetics, and clinical statistics. Mr. Kanninen has also worked with hospital and occupational health IT systems.

Mr. Kanninen is the founder and CEO/CTO of Statwell Oy (1990‐2002). The company provided statistical consultation and developed software for data collection and statistical analyses of personnel and occupational health questionnaires. The company was merged into BC Platforms in 2002.

Mr. Kanninen studied information technology in production and statistics. He is a co-author on 16 scientific articles published in high-level journals.

Title

IDERHA: Driving Secure Health Data Access and Enabling Federated Ai Solutions aligned with the EHDS Requirements

Abstract

Opportunities and learnings from the first 2 years of the IDERHA Innovative Heatlh Initiative enabled Public Private Partnership consortium. In IDERHA a federated disease agnostic, federated health data space will be implemented to enable the connectivity, access, use and reuse of health data incl. the possibility to develop federated Ai-Solutions. The first use cases will focus onto support more efficient and accurate risk profiling, malignancy risk prediction, diagnosis, and prognosis on lung cancer.

About Dr. Christian Muehlendyck

Christian Muehlendyck is the Scientific Partnerships Lead for Johnson&Johnson MedTech in the Europe Middle East and Africa region, focusing on the Innovative Health Initiative (IHI) activities of J&J MedTech. This includes the establishment and industry leadership of IDERHA, one of the first IHI public-private partnerships. Furthermore, he is a member of the IHI Science & Innovation Panel and Co-Chair of the MedTech Europe Research & Innovation Committee.

Within Johnson & Johnson he has held a series of local, regional and global roles of increasing responsibility spanning from Medical Training to commercial positions and Health Economics & Market Access across the surgical and orthopedic spectrum. Within his roles he worked closely with physicians, patients & health authorities and acquired broad expertise in health care innovation and research incl. healthcare digitalization and robotic surgery.

He began his career as a physician in Orthopedic and General Surgery and obtained his MD/PhD co-leading the development of an innovative Medical Device. He subsequently transitioned into the MedTech industry to further his passion for medical innovation.

Title

Federated Learning in Healthcare: A New Era of AI-Driven Scientific Discovery

Abstract

Federated learning (FL) is revolutionizing healthcare research by enabling collaborative research across institutions where algorithms are shared rather than data, ensuring patient privacy and security. By breaking down data silos, FL greatly facilitates access to diverse, real-world datasets, enhancing robustness and generalizability of AI-driven insights. Effective federated research requires the seamless integration of a scalable technology, collaborative science and data and agile algorithm development. This symbiotic trifecta will greatly enable accelerated scientific discovery and healthcare innovation, opening new frontiers in medical science while preserving individual contributors’ data control. In this session, we will explore how aligning these elements can drive impactful research and transform healthcare delivery.

About Robin Röhm

CEO & Co-founder of Apheris – Robin studied medicine, philosophy and mathematics and was trained in global banking at UBS. In one of his previous start-ups, he lost multiple customers as data couldn’t be centralized due to regulatory constraints. He is driving the vision, strategy, and culture of Apheris.​

Title

Ethical & legal hurdles for medical data collaboration

Abstract

The analysis of large data sets has the potential to significantly advance medical research and to create new foundations for therapeutic options. This applies not only to the evaluation of laboratory data or routine patient data, but also and especially to image data from radiology or pathology. From a purely medical point of view, the creation of databases that combine large amounts of medical data from all medical disciplines would be desirable. However, data protection law sets high hurdles for such aggregated data analysis. This is because European data protection law prohibits data processing in principle and only allows it in cases where the law provides for it. One solution would be to use data trustees to intermediate between the individual players.

About Dr. med. Theodor Rüber

Priv.-Doz. Dr. med. Theodor Rüber is a neurologist and clinician scientist at the department of neuroradiology at the University of Bonn Medical Center, where he leads the translational neuroimaging research group. Furthermore, he acts as managing director of the Center for Medical Data Usability and Translation at Bonn University under its director Professor Alexander Radbruch. Dr. Rüber studied medicine, philosophy, and catholic theology at the University of Bonn, philosophy at Harvard University and medicine at the Universidad del Valle in Cali/Colombia. He completed his doctoral thesis in the Neuroimaging & Stroke Recovery Lab at Harvard Medical School and has received several awards for his research, including the NRW Young Scientist Award. Furthermore, Dr. Rüber is founder of the NGDO CASA HOGAR, which is committed to women empowerment and capacity building of women in Colombia.

Title

"The Devil is in the Details" of Clinical and Omics Data Classes in Large Scale Platforms

Abstract

Innovation failure involves emotional aspects such as grief, shame, and blame games. It can result in traumatic experiences or become a source of embarrassment. Nevertheless, a growing group of researchers focuses on the positive connotation of innovation failure when they describe failure as necessary for learning and growth. A certain number of failures could be essential for developing an optimal innovation strategy.  One of the fundamental problems in science is the understanding of discovery and innovation process to explain (and hopefully) scientific and technological change in society. So-called market-oriented sciences tend to have a higher probability of generating inventions and innovations. In particular, the field of medicine sees a significant number of inventions, driven by the opportunistic behavior of its practitioners. In general, major causes of innovation failure seemed to be located in planning and design, execution, and market orientation and are associated with goal difficulty, environmental complexity, and uncertainty. Research shows that organizations that embrace and learn from innovation failure gain a competitive edge. Companies that are prepared to acknowledge, analyze, and adapt after setbacks become stronger, smarter, and more strategic over time. But here’s the catch: Organizations that avoid failure or ignore its lessons struggle to develop effective innovation strategies - putting their competitive future at risk.

About Rudi Schmidt

Rudi Schmidt serves as CEO at the Amanahyat Institute in Abu Dhabi (AIAD). AIAD evaluates biomedical organizations and companies, their assets and business models in the context of financing rounds, M&A and due diligence reviews.

Before AIAD, Rudi served as COO at the Institute for Healthier Living Abu Dhabi and as CEO at !mmunetrue, a company that analyzes immunological patterns across diseases and treatments.

Rudi Schmidt also spent 15 years at Asklepios Hospitals, the 2nd largest hospital chain in Europe, the last 7 years as EVP Precision Medicine, and as CEO of several Asklepios affiliates, including a metagenome company and a provider of Real-World Data in precision medicine for commercial R&D.

In previous leadership positions, Rudi was responsible for Corporate Communications and Investor Relations at Europe's largest listed IT system house (Bechtle AG) and at debitel AG, a joint venture between DaimlerChrysler, Metro and EP, which was later acquired by Swisscom AG.

Title

Data visiting in medicine, a shift towards real-world collaboration

Abstract

The recent increase of multi-model medical data is mind-blowing. Many new data types – particularly the multi-omics data – are not directly addressable by humans as we do not have the right senses to identify patterns in these data. Consequently, to make use of them in medical applications and to achieve precision health, machine learning and artificial intelligence (AI) are prerequisites for the medical transformation to precision health as we need machines to identify patterns in high-dimensional medical data. Current principles for data access are based on data sharing with central data repositories and data access platforms playing a major role in current healthcare ecosystems. However, as medical data are inherently produced in a decentralized fashion, these approaches are not sustainable considering the expected data avalanche. We therefore propose shifting from data sharing to data visiting and insight sharing principles, also considering sustainability issues. We propose an ecosystem based on a technical, legal, organizational and strategical framework that allows the development of large research communities (hospitals, research organizations, companies invested in the health care sector) working as collaborative networks enabling access to larger medical data in a fully decentralized fashion thereby increasing the dataspaces for the development of scalable and generalizable AI applications. Further, our envisioned networks enable full compliance with current AI regulations (e.g. European AI Act) or data privacy regulations (e.g. GDPR).

About Prof. Dr. Joachim Schultze

Joachim Schultze is the Director of Systems Medicine at the German Center for Neurodegenerative Diseases (DZNE) and Professor at the University of Bonn. He went to Medical School in Tübingen, spent 10 years at DFCI, Harvard Medical School, in Boston before he returned to Germany with a Sofia Kovalevskaya Award of the Humboldt Foundation in 2001. Since 2019, he is the coordinator of the German DFG-funded NGS competence centers in Germany, speaker of the West German Genome Center, and board member of the Excellence Cluster ImmunoSensation2. He is directing the large EU consortium NEUROCOV and is partner in several other EU consortia. Since 2019, he has been a highly cited researcher. He is an expert in macrophage biology working at the interphase between immunology, neurosciences, genomics and the computational sciences including AI research. He is one of the inventors of memory-driven computing in medicine and Swarm Learning as a new technology enabling data visiting principles.

Title

Data are health, and digital twins are the future of AI in modern treatments

Abstract

Data are crucial for the improvement of healthcare and the implementation of artificial intelligence (AI) in the medical field. With the increasing availability of health data, the potential for improving patient outcomes and healthcare delivery is immense. Digital twins, which are virtual representations of individual patients, have emerged as a promising tool for personalized medicine and predictive analytics. By integrating patient data with digital twin technology, healthcare providers can gain valuable insights as decision support into disease progression, treatment response, and overall patient health. This presentation explores the significance of data in healthcare and the potential of digital twins as the future of AI in medicine. It discusses the opportunities and challenges associated with leveraging health data for AI applications and therefore for support in treatment and diagnostics, as well as the implications for the software as a medical device. The integration of data and digital twins has the potential to revolutionize the way healthcare is delivered, leading to more personalized and effective treatments for patients. However, there are also technical, ethical, regulatory and privacy considerations that need to be addressed to ensure the responsible use of health data. Overall, the intersection of data, health, and digital twins presents exciting opportunities for advancing AI in healthcare and improving patient outcomes.

About Dr. Christian Stephan

With his focus on bioinformatics, PD Dr Christian Stephan combines the increasingly converging worlds of biomedical research and applied health IT. In the course of his career, he conducted research in the Department of Neurology at Heinrich Heine University Düsseldorf for five years before taking over as head of the Bioinformatics/Biostatistics working group at the Medical Proteome Centre at Ruhr University Bochum (RUB) for more than eight years. As a graduate private lecturer, he continues to teach at the RUB Faculty of Medicine.  Since 2012, Christian Stephan has been a partner and managing director of KAIROS GmbH, which has been part of IQVIA since 2021. Thanks to his vast academic and industrial experience in national and international research and industry projects, PD Dr Christian Stephan speaks the language of treating and researching physicians as well as that of researchers and software experts.

Title

LLM use-cases in Drug Development

Abstract

Large Language Models find many efficiency increasing use-cases in Drug Development Industry from Advanced Multi-modal Search over internal program documents to writing regulatory documents for IND-enabling submission or NDA registration. This requires developing technologies in house which build on LLMs such as Retrieval Augmented Generation (RAG), Advanced Prompt Engineering and  Reasoning and Acting (ReAct) Agents. In my talk, I will give a couple of examples of these how these  technologies can be built collaboratively with Data Scientists and Data Engineers and how this work differs from the traditional AI/ML workflow of model development, deployment and operation.

About Dr. Ashar Ahmad

Dr. Ashar Ahmad has a multidisciplinary background with studies in Chemical Engineering and Computer Science. Between 2014 and 2018 he worked at b-it in Prof. Dr. Holger Fröhlich's lab on Statistical Machine Learning methodologies and contributing to translational research projects at the University Medical Centre in Bonn. After receiving his PhD, he joined UCB Pharma as a Post Doctoral Scientist working in the Translational Medicine department. Since 2021, he has been working as a Data Scientist, Associate Director at Grünenthal GmbH in the Drug Development department driving AI and Data Science use-cases across various functions in Global R&D.

Title

Cardiac magnetic resonance imaging for target identification and precision medicine

Abstract

Cardiac magnetic resonance imaging (CMRI) has evolved as a pivotal tool for diagnostic and therapeutic workup for cardiovascular and other disease. It has raised increasing interest for precision medicine and target discovery approaches enabling patient specific clustering and investigation of potential disease causing mechanisms in larger cohorts.  We present applications of UKBiobank’s cardiac CMRI measurements for early target discovery. In a first use case we analyse longitudinal and radial peak diastolic strain rates and left-atrial volume from roughly 40000 participants. We uncover associations to various demographic, hemodynamic, and cardiovascular risk factors and identify gene associations from the underlying genetic architecture of diastolic function traits.  In a second use case we predict chronological age using more than 100 derived CMRI and also 12-lead electrocardiogram features. We derive cardiac age to chronological age difference as bespoke phenotypes, revealing that cardiometabolic risk factors can accelerate cardiac ageing. Notably, we find that age deltas correlate with genes implicated in the cardiovascular ageing process.  These examples show how complex phenotypic characteristics from CMRI can be integrated with genetic and health information to drive target identification for future cardiovascular therapeutic applications.

About Dr. Christian Bender

After my studies in Bioinformatics I worked at the German Cancer Research Center in Heidelberg on statistical methods for the analysis of high-throughput proteomics and transcriptomics data with focus on cancer signaling pathways and biomarker identification. After I received my PhD in 2011 from University of Heidelberg I worked as scientist at the Translational Oncology gGmbH in Mainz focusing on identification of novel targets for personalized immune therapy from RNA sequencing data. In 2014 I started as computational scientist at the Bayer AG working on data management and mining applications in Biologics Research, before I took up my position as senior scientist for human genetics of cardiovascular disease in 2020. Here, I focus mainly on analysing phenotypic data from cardiac magnetic resonance imaging data and integrate these with genotype, electronic health record and diverse Omics data to derive novel target hypothesis for cardiovascular indications.

Title

An editorial perspective on AI and publishing

Abstract

AI technologies are a double edged sword in scientific publishing. Powerful new tools may speed up the writing and reviewing of manuscripts, but also enable new and exciting forms of academic misconduct. This talk will offer an editorial perspective on the use and abuse of AI tools in publishing, as well as views on best practices in publishing research involving AI methods.

About Dr. Andrew Bissette

Andrew is a chemist by training and Editor in Chief of Cell Reports Physical Science. He received his DPhil and post-doctoral training at the University of Oxford. His editorial career began in 2017 when he moved to Communications Chemistry as one of the launch team, handling submissions across chemical biology and organic chemistry. He joined Cell Reports Physical Science in December 2021, and became Editor in Chief in September 2024.

Title

Value of real-world data in clinical use – case Parkinson’s disease

Abstract

Parkinson’s disease presents a significant burden to patients and a complex challenge for clinicians, especially in assessing treatment efficacy. Symptom fluctuations occur hourly, making it difficult to monitor a patient’s condition outside clinical settings. By collecting accurate and sensitive, real-world data from daily life and analysing it with modern AI technologies, we gain valuable insights into symptom patterns and treatment impact over time. This data-driven approach is set to transform Parkinson’s care, enabling more personalized and effective treatment strategies.

About Paulus Carpelan

Paulus Carpelan has over 30 years of experience in developing high-tech solutions in global corporations as well as in technology startups, including 10 years with medical device development. As a CEO and co-founder of Adamant Health, the resent focus has been on bringing the data driven AI/ML solution to the market to better understand the motor symptoms of people with Parkinson’s disease and myoclonus epilepsy. This insight helps the physicians to better understand their patients and to optimise their care. The second focus area is pharma and research projects, where we can provide clear visibility on how the patients are reacting to a new treatment in a clinical trial or in a post market study.

Title

Generative Models for Drug Discovery

Abstract

Generative models such as diffusion models have seen great success and made many headlines. More subtly, conditional generative models also have the potential to solve very complex inverse problems, e.g., problems in which we only indirectly observe the quantity of interest.  In this talk, we will first discuss recent generative models and then illustrate how they can be used to solve inverse problems in drug discovery. We will discuss some interesting applications, and give an outlook how they can be further applied to structure-based drug design_

About Dr. Djork-Arné Clevert

Dr. Djork-Arné Clevert has a background in computer science and received his doctorate on machine learning for computational biology. In 2022, he accepted a global role as VP, Head of Machine Learning Research within Pfizer R&D. His prior experience includes a seven-year tenure at Bayer AG in Berlin, where he ascended to the position of Director, Head of Machine Learning Research.

From 2007 to 2015, Dr. Clevert was a senior scientist at the renowned Sepp Hochreiter Lab in  Austria. He has been Co-Principal Investigator on several significant projects with the pharmaceutical industry, notably in transcriptome analysis and statistical genetics with J&J and Merck Serono Geneva.

Dr. Clevert's research trajectory began with an emphasis on microarray data and later pivoted towards predicting the biological effects of compounds using advanced techniques such as deep neural networks. A notable achievement was the development of the Exponential Linear Units (ELUs), now a widely accepted standard in deep learning.

He has authored over 100 research articles, scientific reviews, and book chapters, amassing more than 14,000 citations and released open-source software packages, including FARMS, cn.FARMS, RFN, and Img2Mol.

Title

Q1.6 - Predicting Patient Behavior to Improve Clinical Trial Retention

Abstract

Patient attrition remains a major challenge in clinical trials, driving up costs, causing delays, and compromising data integrity. At Q1.6, we are exploring how deep learning models, integrated with real-time patient data capture (ePRO), can predict patient behavior and proactively address retention issues. By leveraging AI-driven insights, we can detect early indicators of dropout risk and implement targeted interventions to improve patient engagement. Through real-world applications and data-driven strategies, this approach aims to enhance clinical trial efficiency, minimize attrition rates, and improve overall study outcomes.

About Dr. Guillaume de Decker

Guillaume de Decker has nearly two decades of experience in leadership roles within the pharmaceutical and medical industries. His career has been shaped by a strong focus on innovation, operational excellence, and advancing healthcare technologies. Since 2023, Guillaume has led Q1.6, defining its overall direction, vision, and strategy. Under his leadership, Q1.6 has embraced deep learning models to enhance efficiency in clinical trials and remote patient monitoring. His commitment to leveraging technology for meaningful improvements in healthcare continues to drive the company’s evolution.

Title

Patient representation learning from real-world data: from target discovery to clinical development

About Dr. Johann de Jong

Johann de Jong is a biomedical data scientist experienced in identifying and implementing opportunities for machine learning across the pharmaceutical R&D pipeline, ranging from target discovery to clinical development, and from oncology to neurological disorders. He obtained his PhD in Computational Cancer Biology from Delft University of Technology (the Netherlands), developing machine learning models for studying gene regulation and cancer gene discovery. After a number of years in academia and industry, including at the Netherlands Cancer Institute, BASF, UCB Pharma and Boehringer Ingelheim, he re-joined UCB Pharma in October 2024 as a Data Science Lead in the Integrated Insights & Advanced Analytics department centered around the use of real-world patient data for supporting the progression of assets through preclinical and clinical development.

Title

Applications of AI in drug development: the Catch-22 of simplification and generalizability

Abstract

Biomedical machine learning has shown great promise for the analysis of high-dimensional (-omics, medical images, digital pathology) data from clinical trials. However, creating a ML model that performs well on in-the-bag test data is not sufficient for subsequent adoption in clinical trials. In this talk, we will explore the importance of context of use, generalizability, simplification and application using examples from realistic scenarios.  We will explore the challenges of using advanced methods in a heavily regulated industry, and discuss how simplification and validation are key to improve the credibility, interpretability and applicability of models in the clinical trial setting.

About Dr. Gunther Jansen

Gunther Jansen leads the Multimodal Data & Analytics organization in the Development arm of Novartis Pharma. With his team, he is responsible for analyzing high-dimensional data from clinical trials. Their primary focus is on applying models and generating insights to inform portfolio decision making rather than de novo discovery. Gunther has a particular interest in early fusion multimodal AI architectures, federated learning, applications of biomedical foundation models and domain adaptation/generalizability within regulated environments. Prior to joining Novartis, Gunther was Global Head of Personalized Healthcare Analytics at Roche and has also worked in smaller biotechs as well as academia.  He has a background in computational biology and genetics and is deeply familiar with RWD.

Title

Utilization of AI/ML in Clinical Pharmacology to improve clinical decision making

Abstract

In recent years, the integration of Artificial Intelligence (AI) and Machine Learning (ML) into clinical pharmacology has transformed the landscape of pharmacometrics. This presentation will explore how these advanced computational methodologies enhance clinical decision-making processes, offering a more precise and efficient approach to patient care. We will discuss the utilization of AI/ML techniques in pharmacokinetic and pharmacodynamic modeling, highlighting the ability to analyze complex datasets that predict drug behavior and optimize dosing strategies.

About Dr. sc. hum. Sven Mensing

Senior Director, Senior Research Fellow Head of Pharmacometrics and Quantitative Systems Pharmacology Clinical Pharmacology, R&D, Ludwigshafen, Germany

Sven leads a group of 40 modeling and data experts supporting the most challenging modeling & simulation activities in Clinical Pharmacology at AbbVie. Using mechanistic mathematical models and advanced statistical tools, his pharmacometrics team delivers data driven assessments to optimize AbbVie’s clinical development strategies for speed, size, and insight. Sven joined Abbott/AbbVie in 2008 where he contributed to the success of numerous assets (including Humira, Mavyret, Venetoclax and many more) striving towards replacing the need to observe with the ability to predict by using science, math, and IT. Sven holds a PhD in Medical Informatics from the University of Heidelberg and a master’s degree in Biomathematics from the University of Greifswald. Sven was named Senior Research Fellow in 2023.

Title

Stopping people from becoming patients – Harnessing the power of real-world data and omics to predict and preempt disease

About Dr. Lykke Pedersen

Currently, I am a Product Creator in the Transformational Prevention Unit (TPU) at Novo Nordisk. Our mission in the TPU is to develop science-based and scalable commercial solutions to prevent obesity and its consequences. My focus is to create data-driven solutions that predict obesity, applying a precision medicine approach to tailor interventions to individual’s specific needs and optimize health outcomes.

With an academic background in biophysics from the University of Copenhagen, I am driven to explore biological systems and derive insights from data that can be applicable across various disciplines and scientific fields.

Following my academic carrier at the University of Copenhagen, my role as a principal bioinformatician at Roche allowed me to serve as the lead antisense oligonucleotide designer on several drug discovery projects involving drug candidates that advanced to clinical trials.

Jumping from big pharma to start-up world, I spent more than three years as the Chief Pharma Officer and Head of RNA Therapeutics at a Danish/Spanish AI start-up, where I focused on disease understanding and RNA therapeutics discovery. This experience provided me with extensive knowledge on how different pharma and biotech companies utilize data to gain insights into disease mechanisms, the outcome of clinical trials or insights into drug properties.

Title

AI Innovations: Transforming Medical Imaging and Shaping the Future of Healthcare

Abstract

Artificial intelligence is driving a new era in healthcare, transforming diagnostic accuracy and efficiency across various domains. This talk will explore how AI-powered technologies are revolutionizing medical imaging, paving the way for a future where healthcare is more precise, personalized, and proactive. Significant progress includes the development of intelligent systems that improve image quality and shorten examination times, enabling quicker and more accurate diagnoses. AI-driven imaging is reshaping modern medicine by offering personalized insights and advancing health outcomes. The integration of AI into medical imaging workflows not only streamlines routine tasks but also empowers healthcare professionals to focus on complex interpretations and patient care.

About Edina Timkó

Edina is a bioengineer turned AI professional who found her true calling in healthcare and data science. With a passion for exploring the intersection of biology and technology, she pursued a Master's degree in info-bionics engineering. Since joining GE Healthcare as an intern during her final years at university, Edina has steadily advanced to her current role as a Staff Data Scientist. She has extensive experience in developing AI for medical imaging, including AI product development for ultrasound images. Currently, Edina is focused on biomarker extraction from MRI images for the early prediction of Alzheimer's disease as part of the PREDICTOM consortia. Her work is dedicated to advancing healthcare through innovative AI-driven solutions.