Thank you to all of those who attended our ICML tutorial on Machine Learning for Personalised Health. The slides for the tutorial are now available here.
Goal and Motivation
Machine learning advances are opening new routes to more precise healthcare, from the discovery of disease subtypes for patient stratification to the development of personalised interactions and interventions. As medicine pivots from treating diagnoses to treating mechanisms, there is an increasing need for personalised health through more intelligent feature extraction and phenotyping. This offers an exciting opportunity for machine learning techniques to impact healthcare in a meaningful way, by putting patients at the centre of research. Health presents some of the most challenging and under-investigated domains of machine learning research. This tutorial presents a timely opportunity to engage the machine learning community with the unique challenges presented within the healthcare domain as well as to provide motivation for meaningful collaborations within this domain.
- Introduction: What are the drivers of machine learning in healthcare?
- Wellness and self-care personalisation: patient perspective
- Precision drug discovery, development and therapeutics: industry perspective
- Population data-driven healthcare: policy perspective
- Data protection and connected care: provider and regulator perspectives
- Critical evaluation of what has been done previously within data-driven healthcare
- Traditional biostatistical and clinical epidemiological approaches
- Current applications of machine learning in health sciences and clinical practice
- A contextual evaluation of problem-led modelling frameworks
- Machine learning strategies for healthcare personalisation
- Probabilistic modelling to uncover and understand disease subtypes
- Predictive modelling framework
- Causal modelling framework
- Interpretability of personalisation
- Examples of Fairness and bias in ML algorithms for health
- Future challenges of ML in healthcare: A unified framework for integrating multiple data types to understand causality and refine personalisation
This tutorial will be targeted towards a broad machine learning audience with various skill sets, some of whom may not have encountered practical applications. The main goal is to transmit inter- as well as intra- disciplinary thinking, to evaluate problems across disciplines as well as to raise awareness of context-driven solutions which can draw strength from using multiple areas of critique within the machine learning discipline. No background in Healthcare or Medicine is needed.
Lamiae Azizi is an Assistant Professor at the School of Mathematics and statistics and the Research director for health at the centre for translational data science at the University of Sydney. Her research interests are developing and applying statistical machine learning methods to complex real life applications in particular biomedical data. She is particularly interested in developing probabilistic models for: data fusion/integration from multiple sources of information, data privacy and AI ethical issues in healthcare; and for building effective recommender systems ready for use by medical practitioners in their daily tasks.
Konstantina Palla is a Machine Learning Researcher in the Healthcare AI Division at Microsoft Research Cambridge. Her research is focusing on the construction and application of Bayesian probabilistic models for discovering latent structure in data. Recently, she has been particularly interested in the application of probabilistic modelling in the Healthcare domain as a means to understand disease subtypes and patients’ subgroups. In her PhD, she developed nonparametric models for relational data with a focus on time evolving settings.
Danielle Belgrave is a Machine Learning Researcher in the Healthcare AI Division at Microsoft Research Cambridge. She also has a (tenured) Research Fellowship at Imperial College London and received a Medical Research Council Career Development Award in Biostatistics (2015 – 2018). Her research focuses on integrating expert scientific knowledge to develop statistical machine learning models to understand disease progression over time, with the goal of identifying personalized disease management strategies. She has experience of applied machine learning for personalized health both within the pharmaceutical industry and academia.