Revolutionizing Healthcare: returning on September 29

Following a short break over the summer, the van der Schaar Lab’s Revolutionizing Healthcare engagement series for clinicians will resume on September 29, with a session on time series AI and machine learning tools.

Revolutionizing Healthcare: looking back on a year of discussions, insights, and partnerships

In September 2020, Mihaela van der Schaar created Revolutionizing Healthcare, inviting clinicians around the world to forge new partnerships with the van der Schaar lab.

The aim of the series was to learn to speak a common language by fostering understanding of, and trust in, AI and machine learning among clinicians—while simultaneously guiding the creation of new tools that clinicians and other healthcare decision-makers would find genuinely trustworthy and useful.

Clinicians have taken center stage in each session since the start—even more so since the series switched to a roundtable format. With 9 sessions in the books, the series has explored a wide range of topics, from the broad and conceptual to the specific.

Links to all 9 sessions (and supporting materials) are provided below:
Session 1: what machine learning can offer healthcare [written companion piece]
Session 2: a framework for ML for healthcare [written companion piece]
Session 3: tools for acute care
Session 4: ML tools for cancer (risks, screening, diagnosis) [related content]
Session 5: ML tools for cancer (post-diagnosis care) [related content]
Session 6: roundtable on interpretability [written companion piece]
Session 7: second roundtable on interpretability [written companion piece]
Session 8: roundtable on personalized therapeutics [related content]
Session 9: roundtable on AI/ML decision-support tools [related content]

September 29 session on time series AI/machine learning tools

The theme of the upcoming September 29 Revolutionizing Healthcare session will be time series AI and machine learning tools—an absolutely vital topic that receives insufficient consideration.

The transformation of healthcare through machine learning depends heavily on the successful application of time series data to model longitudinal trajectories for health and disease. This is a complex, hard-to-define undertaking that touches every other area of machine learning for healthcare. Importantly, we must ensure that the tools that are developed are ones that decision-makers find trustworthy, relevant, and interpretable.

This topic has been explored extensively in a recent write-up.

In the Revolutionizing Healthcare session on September 29, Mihaela will showcase a working demonstrator of Clairvoyance, the lab’s pipeline toolkit for time series. A panel of clinicians will explore some patient case studies within the Clairvoyance demonstrator, and will share their thoughts on its strengths and weaknesses. They will then discuss the role such tools could play in supporting decision-making, as well as some of the key considerations these tools must take into account in order to be truly useful.