All time slots are expressed in CET.

Replays of each session will be available to anyone who registers to the Winter School as we understand that CET time can make livestream difficult for some candidates living in different timezones.

04/01/2020
9:45 - 10:30
Opening ceremony

Opening ceremony

9:45 - 10:30

Emmanuel Bacry - General chair AI4Health, Chief Scientific Officer, Health Data Hub

Stéphanie Allassonnière - Deputy Director, 3IA PRAIRIE

Nicholas Ayache - Scientific Director, 3IA Côte d'Azur 

Alexandre Moreau-Gaudry - Health representative on MIAI Scientific board, 3IA MIAI Grenoble

 

10:30 - 11:30
Adrian Weller - “Trustworthy AI”

Adrian Weller - “Trustworthy AI”

10:30 - 11:30

12:00 - 13:00
Pearse Keane - “Transforming healthcare with ar...

Pearse Keane - “Transforming healthcare with artificial intelligence - lessons from ophthalmology“, part 1

12:00 - 13:00

Abstract: Ophthalmology is among the most technology-driven of all the medical specialties, with treatments utilizing high-spec medical lasers and advanced microsurgical techniques, and diagnostics involving ultra-high resolution imaging. Ophthalmology is also at the forefront of many trailblazing research areas in healthcare, such as stem cell therapy, gene therapy, and - most recently - artificial intelligence. In July 2016, Moorfields announced a formal collaboration with the world’s leading artificial intelligence company, DeepMind. This collaboration involves the sharing of >1,000,000 anonymised retinal scans with DeepMind to allow for the automated diagnosis of diseases such as age-related macular degeneration (AMD) and diabetic retinopathy (DR). In my presentation, I will describe the motivation - and urgent need - to apply deep learning to ophthalmology, the processes required to establish a research collaboration between the NHS and a company like DeepMind, the initial results of our research, and finally, why I believe that ophthalmology could be first branch of medicine to be fundamentally reinvented through the application of artificial intelligence.

14:00 - 15:00
Pearse Keane - “Transforming healthcare with ar...

Pearse Keane - “Transforming healthcare with artificial intelligence - lessons from ophthalmology“, part 2

14:00 - 15:00

Abstract: Ophthalmology is among the most technology-driven of all the medical specialties, with treatments utilizing high-spec medical lasers and advanced microsurgical techniques, and diagnostics involving ultra-high resolution imaging. Ophthalmology is also at the forefront of many trailblazing research areas in healthcare, such as stem cell therapy, gene therapy, and - most recently - artificial intelligence. In July 2016, Moorfields announced a formal collaboration with the world’s leading artificial intelligence company, DeepMind. This collaboration involves the sharing of >1,000,000 anonymised retinal scans with DeepMind to allow for the automated diagnosis of diseases such as age-related macular degeneration (AMD) and diabetic retinopathy (DR). In my presentation, I will describe the motivation - and urgent need - to apply deep learning to ophthalmology, the processes required to establish a research collaboration between the NHS and a company like DeepMind, the initial results of our research, and finally, why I believe that ophthalmology could be first branch of medicine to be fundamentally reinvented through the application of artificial intelligence.

15:30 - 17:00
Dorin Comaniciu - “Artificial Intelligence for ...

Dorin Comaniciu - “Artificial Intelligence for Healthcare: From Hype to Value“

15:30 - 17:00

Abstract: We present the evolution of artificial intelligence (AI) from a hyped technology to value creation for healthcare. After the introduction of theoretical concepts, we define four hierarchical levels of healthcare data generation and processing of increasing complexity. At the imaging scanner and instrument level, AI aims at improving, simplifying, and standardizing data acquisition and preparation. We present examples of systems for AI-driven automatic patient iso-centering before a computed tomography scan, deep learning-based image reconstruction, and creation of optimized and standardized visualizations, for example, automatic rib-unfolding. At the reading and reporting level, AI focuses on the detection and characterization of abnormalities and on automatic measurements in images. We introduce multiple AI systems for the brain, heart, lung, prostate, and musculoskeletal disease and define trustable AI. How can we estimate the uncertainty of an AI system? What are its sources of errors? The third level is exemplified by the integrated nature of the clinical data in a patient-specific manner. The AI algorithms at this level focus on risk prediction and stratification, as opposed to merely detecting, measuring, and quantifying images. An AI-based approach for individualizing radiation dose in lung stereotactic body radiotherapy is discussed. Furthermore, digital twin is presented as a concept of individualized computational modeling of human physiology. Finally, at the cohort and population analysis levels, the focus of AI shifts from clinical decision-making to operational decisions and process optimization.

17:00 - 18:00
Poster session

Poster session

17:00 - 18:00

05/01/2021
9:30 - 11:30
Michael Bronstein - “Geometric deep learning on...

Michael Bronstein - “Geometric deep learning on graphs and manifolds: going beyond Euclidean data”, part 1

9:30 - 11:30

Abstract: In the past decade, deep learning methods have achieved unprecedented performance on a broad range of problems in various fields from computer vision to speech recognition. So far research has mainly focused on developing deep learning methods for Euclidean-structured data. However, many important applications have to deal with non-Euclidean structured data, such as graphs and manifolds. Such data are becoming increasingly important in computer graphics and 3D vision, sensor networks, drug design, biomedicine, high energy physics, recommendation systems, and social media analysis. The adoption of deep learning in these fields has been lagging behind until recently, primarily since the non-Euclidean nature of objects dealt with makes the very definition of basic operations used in deep networks rather elusive. In this talk, I will introduce the emerging field of geometric deep learning on graphs and manifolds, overview existing solutions and outline the key difficulties and future research directions. As examples of applications, I will show problems from the domains of computer vision, graphics, medical imaging, and protein science.

12:00 - 13:00
Michael Bronstein - “Geometric deep learning on...

Michael Bronstein - “Geometric deep learning on graphs and manifolds: going beyond Euclidean data”, part 2

12:00 - 13:00

Abstract: In the past decade, deep learning methods have achieved unprecedented performance on a broad range of problems in various fields from computer vision to speech recognition. So far research has mainly focused on developing deep learning methods for Euclidean-structured data. However, many important applications have to deal with non-Euclidean structured data, such as graphs and manifolds. Such data are becoming increasingly important in computer graphics and 3D vision, sensor networks, drug design, biomedicine, high energy physics, recommendation systems, and social media analysis. The adoption of deep learning in these fields has been lagging behind until recently, primarily since the non-Euclidean nature of objects dealt with makes the very definition of basic operations used in deep networks rather elusive. In this talk, I will introduce the emerging field of geometric deep learning on graphs and manifolds, overview existing solutions and outline the key difficulties and future research directions. As examples of applications, I will show problems from the domains of computer vision, graphics, medical imaging, and protein science.

14:00 - 16:00
Barbara Engelhardt - “Machine learning to impr...

Barbara Engelhardt - “Machine learning to improve clinical care: What exists, and what is left to do”

14:00 - 16:00

16:30 - 17:30
Partners' roundtable

Partners' roundtable

16:30 - 17:30

06/01/2021
9:30 - 11:30
Sophia Ananiadou - “Biomedical Text Mining: met...

Sophia Ananiadou - “Biomedical Text Mining: methods, tools and applications”, part 1

9:30 - 11:30

Abstract: In the last decade, Text Mining (TM) methods have been used for the effective, timely identification and extraction of knowledge for biomedicine. The adaptation of statistical, machine learning (ML) and more recently deep learning (DL) methods on text has boosted performance in several BioNLP tasks such as biomedical named entity recognition (NER), relation extraction (RE) and event extraction (EE). Such methods are vastly used in downstream applications, including evidence-based medicine, automated curation of pathway models, and literature-based discovery. Seeking to answer questions by applying a systematic and transparent methodology on textual evidence in medical resources (e.g. clinical trials, scientific literature, EHRs), text mining is able not only to identify information of interest but also to provide confidence measures for linking and ranking information, to generate new hypotheses and ultimately, to assist experimental design. The usefulness and robustness of these approaches has led to the development of an abundance of biomedical tools, applying AI technology to assist health practitioners in various ways. This talk will be focused on key biomedical TM methods and how these can be used into real-world applications. We will focus on methods developed in the National Centre for Text Mining (www.nactem.ac.uk), few of which include automating the process of systematic reviews (RobotAnalyst), annotation tools using active and proactive learning (APLenty), semantic faceted search engines (Thalia) and time-sensitive search for history of medicine texts (HoM).

12:00 - 13:00
Sophia Ananiadou - “Biomedical Text Mining: met...

Sophia Ananiadou - “Biomedical Text Mining: methods, tools and applications”, part 2

12:00 - 13:00

Abstract: In the last decade, Text Mining (TM) methods have been used for the effective, timely identification and extraction of knowledge for biomedicine. The adaptation of statistical, machine learning (ML) and more recently deep learning (DL) methods on text has boosted performance in several BioNLP tasks such as biomedical named entity recognition (NER), relation extraction (RE) and event extraction (EE). Such methods are vastly used in downstream applications, including evidence-based medicine, automated curation of pathway models, and literature-based discovery. Seeking to answer questions by applying a systematic and transparent methodology on textual evidence in medical resources (e.g. clinical trials, scientific literature, EHRs), text mining is able not only to identify information of interest but also to provide confidence measures for linking and ranking information, to generate new hypotheses and ultimately, to assist experimental design. The usefulness and robustness of these approaches has led to the development of an abundance of biomedical tools, applying AI technology to assist health practitioners in various ways. This talk will be focused on key biomedical TM methods and how these can be used into real-world applications. We will focus on methods developed in the National Centre for Text Mining (www.nactem.ac.uk), few of which include automating the process of systematic reviews (RobotAnalyst), annotation tools using active and proactive learning (APLenty), semantic faceted search engines (Thalia) and time-sensitive search for history of medicine texts (HoM).

14:00 - 16:00
Susan Murphy - “Challenges in Developing Online...

Susan Murphy - “Challenges in Developing Online Learning and Experimentation Algorithms in Mobile Health”, part 1

14:00 - 16:00

Abstract: Mobile health provides a great testbed with unique challenges to existing reinforcement learning (RL) algorithms. In mobile health, smart devices collect the user's state via sensors/self-report, deliver treatment actions, and collect subsequent reward signals. Challenges to RL include (1) high noise in both the reward and well as state transitions, (2) domain science needs to use data resulting from the use of RL, in subsequent batch causal and off-policy analyses, (3) non-stationarity and (4) domain science needs for assessments of how well the RL algorithm personalized the mobile health app to the user. We will discuss generalizations to RL to meet the challenges as well as results from a mobile health clinical trial using a RL algorithm.

16:30 - 17:30
Susan Murphy - “Challenges in Developing Online...

Susan Murphy - “Challenges in Developing Online Learning and Experimentation Algorithms in Mobile Health”, part 2

16:30 - 17:30

Abstract: Mobile health provides a great testbed with unique challenges to existing reinforcement learning (RL) algorithms. In mobile health, smart devices collect the user's state via sensors/self-report, deliver treatment actions, and collect subsequent reward signals. Challenges to RL include (1) high noise in both the reward and well as state transitions, (2) domain science needs to use data resulting from the use of RL, in subsequent batch causal and off-policy analyses, (3) non-stationarity and (4) domain science needs for assessments of how well the RL algorithm personalized the mobile health app to the user. We will discuss generalizations to RL to meet the challenges as well as results from a mobile health clinical trial using a RL algorithm.

17:30 - 18:30
Networking event

Networking event

17:30 - 18:30

07/01/2021
9:30 - 11:00
Practical sessions

Practical sessions

9:30 - 11:00

11:30 - 13:00
Practical sessions

Practical sessions

11:30 - 13:00

14:00 - 16:00
Practical sessions

Practical sessions

14:00 - 16:00

16:30 - 18:00
Practical sessions

Practical sessions

16:30 - 18:00

08/01/2021
9:30 - 11:00
Practical sessions

Practical sessions

9:30 - 11:00

11:30 - 13:00
Practical sessions

Practical sessions

11:30 - 13:00

14:00 - 16:00
Practical sessions

Practical sessions

14:00 - 16:00

16:30 - 18:00
Practical sessions

Practical sessions

16:30 - 18:00

Download the program below

Organized by the French Health Data Hub, the unique gateway to access health data for public interest research in France, and some of France’s leading AI research institutions: MIAI Grenoble, 3IA Cote d’Azur (Nice) and PRAIRIE (Paris),

Supported by the French Association for Medical Informatics, the Paris Brain Institute – ICM and the University of Paris