Robust and interpretable machine learning for biomedicine and healthcare
Lead Research Organisation:
University of Oxford
Abstract
Context & Potential Impact: Deep learning methods excel at spotting patterns in large datasets. The increasing availability of medical datasets has thus led to many reports of deep learning models performing on par with human physicians in prediction and diagnosis tasks. In my DPhil research, I will develop and apply machine learning models in the context of biomedicine and healthcare, with a particular focus on two topics: interpretability and robustness.
Interpretability: Deep learning models are notoriously opaque: Even if they give highly accurate predictions, it's not easy to figure out how they do so. There would be two major advantages in improving our ability to interpret, explain, and understand neural networks. First, black-box algorithms are unlikely to inspire the required trust in medical decisionmakers and thus unlikely to find their way into clinical practice. Second, the ability to interpret deep neural networks could help us to capitalise on the implicit knowledge captured in these trained models, potentially elucidating the mechanisms that play a role in disease development and progression.
Robustness: Robustness to dataset shift is a central topic in many areas of machine learning. In the medical context, robustness is critical as covariate shifts are ubiquitous. Most notably, different healthcare providers serve different populations, with different demographics, baseline prevalence levels, behaviours and needs. If we ever hope to apply machine learning models in healthcare settings, they need to generalise to these different populations, or at least 'know' when they should be uncertain and be readily adaptable to new populations.
Aims and Objectives:
To develop novel methodology for interpretable and robust machine learning in the medical setting.
To apply these methods to generate new insights about health and disease.
Novelty of the research methodology: While there is a significant body of work on interpretability and robustness, much still needs to be done, for example in defining metrics of interpretability that are valid and useful in the biomedical context. We will improve upon existing methods by drawing from various related subfields in machine learning, such as causality, adversarial robustness, out-of-distribution detection, and Bayesian machine learning). We will work closely with domain experts to identify the most relevant research questions in biomedicine and healthcare.
Alignment to EPSRC's strategies and research areas. This research fits within the following EPSRC research areas: Artificial Intelligence Technologies, Image and Vision Computing, Information Systems, Mathematical Biology, Medical Imaging (including Medical Image and Vision Computing), Digital Signal Processing and Statistics and Applied Probability.
Companies or collaborators involved: None so far
Interpretability: Deep learning models are notoriously opaque: Even if they give highly accurate predictions, it's not easy to figure out how they do so. There would be two major advantages in improving our ability to interpret, explain, and understand neural networks. First, black-box algorithms are unlikely to inspire the required trust in medical decisionmakers and thus unlikely to find their way into clinical practice. Second, the ability to interpret deep neural networks could help us to capitalise on the implicit knowledge captured in these trained models, potentially elucidating the mechanisms that play a role in disease development and progression.
Robustness: Robustness to dataset shift is a central topic in many areas of machine learning. In the medical context, robustness is critical as covariate shifts are ubiquitous. Most notably, different healthcare providers serve different populations, with different demographics, baseline prevalence levels, behaviours and needs. If we ever hope to apply machine learning models in healthcare settings, they need to generalise to these different populations, or at least 'know' when they should be uncertain and be readily adaptable to new populations.
Aims and Objectives:
To develop novel methodology for interpretable and robust machine learning in the medical setting.
To apply these methods to generate new insights about health and disease.
Novelty of the research methodology: While there is a significant body of work on interpretability and robustness, much still needs to be done, for example in defining metrics of interpretability that are valid and useful in the biomedical context. We will improve upon existing methods by drawing from various related subfields in machine learning, such as causality, adversarial robustness, out-of-distribution detection, and Bayesian machine learning). We will work closely with domain experts to identify the most relevant research questions in biomedicine and healthcare.
Alignment to EPSRC's strategies and research areas. This research fits within the following EPSRC research areas: Artificial Intelligence Technologies, Image and Vision Computing, Information Systems, Mathematical Biology, Medical Imaging (including Medical Image and Vision Computing), Digital Signal Processing and Statistics and Applied Probability.
Companies or collaborators involved: None so far
Planned Impact
AIMS's impact will be felt across domains of acute need within the UK. We expect AIMS to benefit: UK economic performance, through start-up creation; existing UK firms, both through research and addressing skills needs; UK health, by contributing to cancer research, and quality of life, through the delivery of autonomous vehicles; UK public understanding of and policy related to the transformational societal change engendered by autonomous systems.
Autonomous systems are acknowledged by essentially all stakeholders as important to the future UK economy. PwC claim that there is a £232 billion opportunity offered by AI to the UK economy by 2030 (10% of GDP). AIMS has an excellent track record of leadership in spinout creation, and will continue to foster the commercial projects of its students, through the provision of training in IP, licensing and entrepreneurship. With the help of Oxford Science Innovation (investment fund) and Oxford University Innovation (technology transfer office), student projects will be evaluated for commercial potential.
AIMS will also concretely contribute to UK economic competitiveness by meeting the UK's needs for experts in autonomous systems. To meet this need, AIMS will train cohorts with advanced skills that span the breadth of AI, machine learning, robotics, verification and sensor systems. The relevance of the training to the needs of industry will be ensured by the industrial partnerships at the heart of AIMS. These partnerships will also ensure that AIMS will produce research that directly targets UK industrial needs. Our partners span a wide range of UK sectors, including energy, transport, infrastructure, factory automation, finance, health, space and other extreme environments.
The autonomous systems that AIMS will enable also offer the prospect of epochal change in the UK's quality of life and health. As put by former Digital Secretary Matt Hancock, "whether it's improving travel, making banking easier or helping people live longer, AI is already revolutionising our economy and our society." AIMS will help to realise this potential through its delivery of trained experts and targeted research. In particular, two of the four Grand Challenge missions in the UK Industrial Strategy highlight the positive societal impact underpinned by autonomous systems. The "Artificial Intelligence and data" challenge has as its mission to "Use data, Artificial Intelligence and innovation to transform the prevention, early diagnosis and treatment of chronic diseases by 2030". To this mission, AIMS will contribute the outputs of its research pillar on cancer research. The "Future of mobility" challenge highlights the importance the autonomous vehicles will have in making transport "safer, cleaner and better connected." To this challenge, AIMS offers the world-leading research of its robotic systems research pillar.
AIMS will further promote the positive realisation of autonomous technologies through direct influence on policy. The world-leading academics amongst AIMS's supervisory pool are well-connected to policy formation e.g. Prof Osborne serving as a Commissioner on the Independent Commission on the Future of Work. Further, Dr Dan Mawson, Head of the Economy Unit; Economy and Strategic Analysis Team at BEIS will serve as an advisor to AIMS, ensuring bidirectional influence between policy objectives and AIMS research and training.
Broad understanding of autonomous systems is crucial in making a society robust to the transformations they will engender. AIMS will foster such understanding through its provision of opportunities for AIMS students to directly engage with the public. Given the broad societal importance of getting autonomous systems right, AIMS will deliver core training on the ethical, governance, economic and societal implications of autonomous systems.
Autonomous systems are acknowledged by essentially all stakeholders as important to the future UK economy. PwC claim that there is a £232 billion opportunity offered by AI to the UK economy by 2030 (10% of GDP). AIMS has an excellent track record of leadership in spinout creation, and will continue to foster the commercial projects of its students, through the provision of training in IP, licensing and entrepreneurship. With the help of Oxford Science Innovation (investment fund) and Oxford University Innovation (technology transfer office), student projects will be evaluated for commercial potential.
AIMS will also concretely contribute to UK economic competitiveness by meeting the UK's needs for experts in autonomous systems. To meet this need, AIMS will train cohorts with advanced skills that span the breadth of AI, machine learning, robotics, verification and sensor systems. The relevance of the training to the needs of industry will be ensured by the industrial partnerships at the heart of AIMS. These partnerships will also ensure that AIMS will produce research that directly targets UK industrial needs. Our partners span a wide range of UK sectors, including energy, transport, infrastructure, factory automation, finance, health, space and other extreme environments.
The autonomous systems that AIMS will enable also offer the prospect of epochal change in the UK's quality of life and health. As put by former Digital Secretary Matt Hancock, "whether it's improving travel, making banking easier or helping people live longer, AI is already revolutionising our economy and our society." AIMS will help to realise this potential through its delivery of trained experts and targeted research. In particular, two of the four Grand Challenge missions in the UK Industrial Strategy highlight the positive societal impact underpinned by autonomous systems. The "Artificial Intelligence and data" challenge has as its mission to "Use data, Artificial Intelligence and innovation to transform the prevention, early diagnosis and treatment of chronic diseases by 2030". To this mission, AIMS will contribute the outputs of its research pillar on cancer research. The "Future of mobility" challenge highlights the importance the autonomous vehicles will have in making transport "safer, cleaner and better connected." To this challenge, AIMS offers the world-leading research of its robotic systems research pillar.
AIMS will further promote the positive realisation of autonomous technologies through direct influence on policy. The world-leading academics amongst AIMS's supervisory pool are well-connected to policy formation e.g. Prof Osborne serving as a Commissioner on the Independent Commission on the Future of Work. Further, Dr Dan Mawson, Head of the Economy Unit; Economy and Strategic Analysis Team at BEIS will serve as an advisor to AIMS, ensuring bidirectional influence between policy objectives and AIMS research and training.
Broad understanding of autonomous systems is crucial in making a society robust to the transformations they will engender. AIMS will foster such understanding through its provision of opportunities for AIMS students to directly engage with the public. Given the broad societal importance of getting autonomous systems right, AIMS will deliver core training on the ethical, governance, economic and societal implications of autonomous systems.
People |
ORCID iD |
Yarin Gal (Primary Supervisor) | |
Jan Brauner (Student) |
Publications
Altman G
(2022)
A dataset of non-pharmaceutical interventions on SARS-CoV-2 in Europe.
in Scientific data
Brauner JM
(2021)
Inferring the effectiveness of government interventions against COVID-19.
in Science (New York, N.Y.)
Gavenciak T
(2022)
Seasonal variation in SARS-CoV-2 transmission in temperate climates: A Bayesian modelling study in 143 European regions.
in PLoS computational biology
Gavenciak T
(2021)
Seasonal variation in SARS-CoV-2 transmission in temperate climates
Leech G
(2022)
Mask wearing in community settings reduces SARS-CoV-2 transmission.
in Proceedings of the National Academy of Sciences of the United States of America
Lison A
(2023)
Effectiveness assessment of non-pharmaceutical interventions: lessons learned from the COVID-19 pandemic
in The Lancet Public Health
Meyerowitz-Katz G
(2021)
Is the cure really worse than the disease? The health impacts of lockdowns during COVID-19.
in BMJ global health
Mishra S
(2021)
Changing composition of SARS-CoV-2 lineages and rise of Delta variant in England.
in EClinicalMedicine
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/S024050/1 | 30/09/2019 | 30/03/2028 | |||
2242671 | Studentship | EP/S024050/1 | 30/09/2019 | 31/03/2024 | Jan Brauner |
Description | We analysed the effectiveness of various government interventions against the spread of COVID-19. |
Exploitation Route | The research outputs continue to guide COVID policies and can serve as a solid fundament to build on, in any potential future pandemics. |
Sectors | Other |
Description | My findings have been used to guide policy-making on COVID-19 interventions around the globe. |
First Year Of Impact | 2020 |
Sector | Other |
Impact Types | Societal Policy & public services |
Description | Gave input on UK's national COVID response |
Geographic Reach | National |
Policy Influence Type | Contribution to a national consultation/review |
Impact | * Early awareness and monitoring of the rise of Delta variant in UK. * Informed UK's Plan B |
URL | https://web.archive.org/web/20210726231234/https:/www.gov.uk/government/publications/spi-m-o-consens... |
Description | Research cited in federal German law |
Geographic Reach | National |
Policy Influence Type | Citation in other policy documents |
Impact | Directly impacted the countermeasures implemented against the third wave of COVID in Germany. |
URL | https://dip21.bundestag.de/dip21/btd/19/284/1928444.pdf |
Description | Research cited in several reviews and meta-analyses on the effect of government interventions against COVID-19 |
Geographic Reach | Europe |
Policy Influence Type | Citation in systematic reviews |
Title | Large dataset on COVID-19 interventions in Europe |
Description | We collected a custom NPI dataset for this modelling study, as existing datasets do not provide sufficient geographical resolution to model the second wave (Table 1). Further advantages of our dataset are NPI definitions tailored towards the second wave and high data quality through extensive validation. All data necessary for the replication of our results are publicly available on https://github.com/MrinankSharma/COVID19NPISecondWave/tree/main/data |
Type Of Material | Database/Collection of data |
Year Produced | 2021 |
Provided To Others? | Yes |
Impact | THis is the first subnational COVID-19 intervention dataset for Europe, the resulting paper has been cited 33 times as of 1 March 2022. |
URL | https://github.com/MrinankSharma/COVID19NPISecondWave/tree/main/data |
Title | epidemics/COVIDNPIs: Inferring the effectiveness of government interventions against COVID-19 |
Description | Pre release |
Type Of Material | Database/Collection of data |
Year Produced | 2020 |
Provided To Others? | Yes |
Impact | Can estimate the effect of government interventions against COVID-19 with it. Other papers have used it, e.g. Banholzer et al., 2021. |
URL | https://zenodo.org/record/4268433 |
Description | Many interviews with international news outlets |
Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Media appearances and work featured in major newspapers, radio stations, TV stations, e.g. Forbes, The Guardian, Vox. Examples: Prominent news outlets featuring my work: • International: The Guardian, Forbes, Vox • German: Tagesschau, Focus, Spiegel, Spektrum der Wissenschaft, Christian Drosten's Podcast (1,2) Interviews with online and print media: • 10/10/2020: Interview by rbb24 on the effective of countermeasures against COVID (link, link) • 20/11/2020: Interview with major German Newspaper (Süddeutsche Zeitung) on the effectiveness of government interventions against COVID-19 (link) • 18/02/2021: Interview with Estonian Public Broadcasting (link, link) • 03/04/2021: Interview with rbb24 (link) • 09/12/2021: Interview with the Guardian Interviews with radio stations: • 25/08/2020: Interview by MDR (Major German radio station) on the effective of countermeasures against COVID (link, link) • 11/01/2021: Live radio interview with Deutschland-Funk, a major German science radio show (link, link, link) • 25/03/2021: Interview by Bayern 2 Radiowelt (Major German radio station) on the effective of countermeasures against COVID Interviews with TV stations: • 19/01/2021: Live TV interview with Deutsche Welle TV, on the effectiveness of government measures against COVID-19. (link) • 21/04/2021: Interview by ARD (main German public TV channel) on countermeasures against COVID, used for TV-Show "Monitor" (link, paper discussed extensively, but interview not shown in the show) |
Year(s) Of Engagement Activity | 2020,2021 |
URL | https://www.theguardian.com/world/2021/dec/20/covid-omicron-england-rules-being-considered |
Description | Many interviews with national news |
Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Many interviews with newspapers, radio stations, and TV channels, on government interventions against COVID-19 transmission. Some examples. https://www.rbb24.de/panorama/thema/2020/coronavirus/beitraege_neu/2020/10/experten-massnahmen-rat-tracing-kontakte.html https://www.mdr.de/wissen/mensch-alltag/corona-massnahmen-sinnvoll-oder-nicht-100.html https://projekte.sueddeutsche.de/artikel/wissen/coronavirus-was-kommt-nach-dem-lockdown-e862488/ https://www.ardaudiothek.de/forschung-aktuell/vertrauter-feind-6-6-wie-bekaempft-man-sars-cov-2-am-effizientesten/85079556 |
Year(s) Of Engagement Activity | 2020,2021 |
URL | https://www.rbb24.de/panorama/thema/2020/coronavirus/beitraege_neu/2020/10/massnahmen-eindaemmung-st... |
Description | My paper was reported on widely, e.g. in TIME and on the front-page of the Guardian |
Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | My paper was reported on widely, e.g. in TIME and on the front-page of the Guardian |
Year(s) Of Engagement Activity | 2023 |
URL | https://www.theguardian.com/technology/2023/oct/24/ai-firms-must-be-held-responsible-for-harm-they-c... |
Description | Op-ed in TIME |
Form Of Engagement Activity | A magazine, newsletter or online publication |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Op-ed in TIME |
Year(s) Of Engagement Activity | 2023 |
URL | https://time.com/6303127/ai-future-danger-present-harms/ |