van der Schaar Lab

Revolutionizing Healthcare

What? Revolutionizing Healthcare is a series of engagement sessions for clinicians, aiming to share ideas and discuss topics that will define the future of machine learning in healthcare. These events are for members of the healthcare community and focus on challenges and opportunities in clinical application of machine learning. We now have roughly 450 clinicians from around the world registered to participate in these sessions.

Why? As a lab, our purpose is to create new and powerful machine learning techniques and methods that can revolutionise healthcare. This doesn’t happen in a vacuum. At inception, we are inspired by ideas and discussions; in implementation, we need connections, trust, and partnership to make a real difference.

Who? While you can learn about our work at major conferences in machine learning or in our papers, we think it’s a better idea to create a community and keep these conversations going. We’re also aware that many people—both in healthcare and machine learning—have questions about what we do, and how they can contribute. We have chosen to restrict access to the actual discussion in the Revolutionizing Healthcare sessions to practicing clinicians so that we can maintain a focus on the clinical issues. 

The full videos of all the previous discussions and presentations are available to anyone and can be found below or on our Youtube channel

We also have dedicated engagement sessions for ML researchers and practitioners called Inspiration Exchange, which focuses on the development of new methods, approaches and techniques.

Thank you for your interest.

When? For more information about Revolutionizing Healthcare—and to sign up to join in—please have a look at the sections below, and keep checking for new updates.

Revolutionizing Healthcare

Themed discussion sessions specifically for healthcare professionals (primarily clinicians).

We would like to:
– introduce machine learning concepts as they relate to healthcare
– spark new projects and collaborations
– demonstrate the real-world impact of machine learning in clinical settings
– discuss institutional barriers preventing wider adoption
– develop a shared vision for the future of machine learning in healthcare.

Standard session format:
– brief introductory presentation
– roundtable discussion featuring clinicians
– open Q&A

Next session in December!

Our next session will take place in December, and will run for just over 1 hour. This session’s content will be revealed in due time!

If you’d like to join us for this or future sessions, please sign up using the form below, and check this page regularly.

Cambridge, United Kingdom Mon, 21 Nov 2022 at 16:00 GMT
Paris, France Mon, 21 Nov 2022 at 17:00 CET
New York, USA Mon, 21 Nov 2022 at 11:00 EST
San Francisco, USA Mon, 21 Nov 2022 at 08:00 PST
Beijing, China Mon, 21 Nov 2022 at 23:00 CST

More time zones can be found here.

If you are not a practicing clinician/do not have an MD, PLEASE DO NOT SIGN UP FOR THESE SESSIONS. If you are a machine learning researcher, please sign up for our Inspiration Exchange engagement sessions.

Session 1: what machine learning can offer healthcare

Ahead of this session, Mihaela published a piece of content on machine learning and the future of healthcare, entitled Revolutionizing healthcare: an invitation to clinical professionals everywhere.

Session 2: a framework for ML for healthcare

Ahead of this session, Mihaela published a piece of content entitled Machine learning for healthcare: Towards a unifying framework.

Session 3: tools for acute care

The focus of the session was on addressing real-world problems in the acute care setting by matching them to formalisms.

During the session, we introduced the lab’s Hub for Healthcare, which contains a classification of some medical problems and associated examples, and then provides formalisms and methods by which they can be solved.

Session 4: ML tools for cancer (risks, screening, diagnosis)

The focus of the session was on addressing real-world problems in the cancer domain (with an emphasis on the pathway up to the point of diagnosis) by matching them to formalisms.

Session 5: ML tools for cancer (post-diagnosis care)

The focus of the session was on addressing real-world problems in the cancer domain (with an emphasis on post-diagnosis treatment) by matching them to formalisms.

Session 6: roundtable on interpretability

Our sixth session was a roundtable on the topic of interpretability.

Following a presentation by Mihaela van der Schaar, a panel of four clinicians discussed various definitions and types of interpretability, as well as real-world needs and contexts in healthcare settings.

Session 7: second roundtable on interpretability

This session was the second roundtable in a double-header focusing on interpretability in ML/AI for healthcare.

Following a quick introduction by Mihaela van der Schaar, a panel of four clinicians and the audience of clinicians discussed a range of complex issues surrounding interpretability, including whether or not current expectations among the clinical community are realistic.

Session 8: roundtable on personalized therapeutics

This session was a roundtable focusing on personalized therapeutics and individualized treatment effects (ITE).

Following a short presentation by Mihaela van der Schaar, a panel of four clinicians and the audience discussed a range of topics related to personalized therapeutics, including the limitations of existing clinical guidelines, and the potential for machine learning and AI to aid the development of more personalized and useful guidelines in healthcare.

Session 9: roundtable on AI/ML decision-support tools

This session featured an international panel of 5 clinicians who discussed AI and machine learning decision support tools for diseases such as breast cancer. One such tool (developed by the van der Schaar lab) is Adjutorium, which was the focus of an extensive study published in Nature Machine Intelligence.

The session started with presentations on Adjutorium from Mihaela van der Schaar and postdoc Ahmed Alaa, which formed the basis of the subsequent clinician roundtable. During the roundtable, the panelists explored a range of topics related to the clinical application of tools like Adjutorium—such as the kinds of information that would be useful for clinicians and patients, and how this information can be displayed (among a range of other topics).

Session 10: getting ML-powered tools in the hands of clinicians (part 1)

In this session, Mihaela van der Schaar and a panel of 4 international clinicians discussed the importance of bridging the gap between ideas and implementation in machine learning for healthcare.

In the first part of the session, Mihaela and the clinicians showed how machine learning can help clinicians estimate individualized treatment effects, using a working demonstrator to show real-world case studies from the ICU setting. The latter part of the session featured a clinician roundtable, during which the panelists discussed different approaches to ensuring that clinicians can make full use of machine learning models.

Session 11: getting ML-powered tools in the hands of clinicians (part 2)

In this session, Mihaela van der Schaar and a panel of 3 international clinicians expanded on the agenda outlined in the previous session: the importance of bridging the gap between ideas and implementation in machine learning for healthcare.

In the first part of the session, Mihaela and the clinicians showed how machine learning can assist dynamic time-to-event analysis and temporal phenotyping, using a working demonstrator to show real-world case studies from prostate cancer. The latter part of the session featured a clinician roundtable, during which the panelists discussed different approaches to ensuring that clinicians can make full use of machine learning models.

Session 12: how can AI/ML transform organ transplantation?

In this session, Mihaela van der Schaar and a panel of 4 international clinicians explored possible AI and machine learning approaches to organ transplantation.

In the first part of the session, Dr. Alexander Gimson, a Cambridge-based transplant hepatologist, outlined the importance, complexities, and unique challenges of the organ transplantation setting, while also highlighting some recent collaborative projects using machine learning for donor-recipient matchmaking and survival prediction. The latter part of the session featured a clinician roundtable comprising an expert panel of four transplantation specialists. Our panelists answered an array of questions (most of which came from the audience) and discussed the path forward for AI and machine learning for organ transplantation.

Session 13: AI and ML for early detection and diagnosis (1/2)

In this session, Mihaela van der Schaar and an international panel of clinical experts explored how AI and machine learning can help transform early detection and diagnosis (ED&D). This session was the first in a double-header on this important topic.

The session started with a presentation by Mihaela van der Schaar, who highlighted the importance of ED&D as one of healthcare’s “holy grails” and introduced a wide range of areas where machine learning and AI can have a positive impact. The latter part of the session featured a clinician roundtable comprising an expert panel of five specialists in the area of diagnosis and detection (primarily related to cancer). Our panelists answered an array of questions posed by Mihaela, and discussed the path forward for AI and machine learning for ED&D.

Session 14: AI and ML for early detection and diagnosis (2/2)

This was the second instalment in a double-header focusing on the absolutely crucial topic of early detection and diagnosis (ED&D). The session included a roundtable consisting of Mihaela van der Schaar and an international panel of three expert clinicians exploring how ED&D can be transformed by AI and machine learning.

The roundtable followed an introductory presentation by Mihaela van der Schaar, who built on the explorations of the previous session in this double-header and set the stage for the panel. After our panelists engaged in their enlightening discussion about possible paths forward for AI and machine learning for ED&D, Dr Eoin McKinney presented a newly-developed machine learning demonstrator specifically designed for ED&D.

Session 15: Leading clinical voices on the future of AI/ML for healthcare

Over the course of three insightful, energetic, and often provocative discussions, Profs. Euan Ashley, Geraint Rees, and Eric Topol shared their personal views on the state of AI and machine learning in healthcare, as well as their respective visions for AI-powered healthcare systems of the future.

Following these discussions, we opened the session up to our audience of clinicians, who shared their own views on the topics Mihaela had discussed with our three leading clinical voices.

Session 16: Synthetic Data in Healthcare

The session started with a presentation by Mihaela, who set the stage with and introduction to synthetic data and its value for modern healthcare. This was followed up with the presentation of a synthetic data demonstrator by Dr Jem Rashbass, MD, developed and designed by members of the van der Schaar lab.

The latter part of the session featured an engaging clinician roundtable comprising an expert panel of three specialists in the area of synthetic data. Our panelists answered a number of questions posed by Mihaela and the audience.

Session 17: Using Machine Learning to power Clinical Trials I

The session started with short presentations by our panellists framing the problems clinicians are currently facing in regards to clinical trials, defining important terms, and giving examples of current challenges and opportunities. 

The latter part of the session featured an engaging clinician roundtable comprising an expert panel of four specialists in the area of clinical trials. Our panellists answered a number of questions posed by Mihaela and the audience.

Session 18: Using Machine Learning to power Clinical Trials II

The session started with a presentation by Mihaela, summarising the previous session, key opportunities, and discussing the main machine learning solutions to the established challenges.

This was followed up by a short, presented discussion of Mihaela and Dr Eoin McKinney about machine learning solutions for estimating heterogenous treatment effects, and SyncTwin to emulate clinical trials. A second conversation between Mihaela and Prof Richard Peck about personalised dose response using machine learning then set the stage for our roundtable.

Session 19: Machine Learning and Cystic Fibrosis

The session started with a presentation by Mihaela and Andres, introducing to the complex challenges surrounding cystic fibrosis, and presenting what ML methods have been developed already.

This was followed up by a panel discussion and questions from the audience.

We thank Prof Damian Downey, Dr Jamie Duckers, Dr Robert Gray, Dr Charles Haworth, Prof Alexander Horsley, and Caroline Cartellieri Karlsen for their participation.

Session 20: AutoPrognosis: Using the next generation of ML tools

The session started with a presentation by Mihaela, introducing AutoPrognosis to the guests and audience. This was followed by a talk by Dr Eoin McKinney, presenting a clinical perspective on the potential of AutoPrognosis. Dr Tom Callender then presented an example of AutoPrognosis used in a real-world case-study.

More info about AutoPrognosis here: https://www.autoprognosis.vanderschaar-lab.com/

This was followed up by a panel discussion and questions from the audience.

We thank Prof Donald E. Ingber (Harvard), Prof Suetonia Palmer (Otago), and Dr Paul Goldsmith (NHS) for their participation.

Session 21: AutoPrognosis 2.0

This session started with a short demonstration by Dr Thomas Callender, MSc MRCP, followed by a short talk by Mihaela about AutoPrognosis in a cardiovascular setting. Prof Eoin McKinney, MD, rounded up the introduction by giving some additional comments from the clinical perspective on AutoPrognosis.

More info about AutoPrognosis here: https://www.autoprognosis.vanderschaar-lab.com/

This was followed up by a panel discussion and questions from the audience.

We thank Dr Anthony Philippakis, MD, PhD (MIT and Harvard) and Dr Sarah Blake, MRCP, PhD (NHS) for their participation.

To learn more about our team of researchers, click here. You can find our publications here.