if you are at the Alzheimer's Association International Conference 2019 in L.A. this July, make sure you say hello to Nonie and Maxine that are presenting their research! Maxine will be presenting recent work on a) identifying prodromal features of dementia patients using primary care electronic health records, and b) using data-driven approaches to identify variation in dementia recognition as a result of national policies. Nonie will be presenting her work on identifying clinically-meaningul subtypes of Alzheimer's using unsupervised learning.
Spiros Denaxas has been successful in his application for promotion and has now been awarded the title of Professor of Biomedical Informatics at UCL. “I am particularly pleased for Spiros as he is one of the few people in the UK to lead the developments underpinning methodologies related to data science at scale. His career path I hope will serve as an important addition to the role models for early-career researchers.”
Harry Hemingway, Director of the UCL Institute of Heath Informatics. Congratulations to the many other successful candidates at this year's academic promotions round!
Our research was featured in the CERN Enlight Highlights summer issue!
Congratulations to Nonie Alexander who successfully upgraded to a full PhD!! Nonie's project "Identifying subtypes of Alzheimer’s disease in Electronic Health Records" focuses on applying unsupervised machine learning algorithms to identify disease subtypes at a phenotypic and molecular level. Onwards and upwards!
Using electronic health records from the CALIBER platform, we have created the first chronological map of human disease tracking over 300 physical and mental morbidities from birth to death. The research paper, published in the Lancet Digital Health, shows the chronology of human health for approx 4 million patients registered in the UK public healthcare system. Reproducibility was a key aspect of our research: all phenotyping algorithms we created are curated in the open-access CALIBER Portal and can be reproduced by the research community.
Homelessness has increased by 165% since 2010, homeless deaths by 24% over the past five years. Our research published in Wellcome Open Research examined >13,000 hospital admissions in electronic health records and found that a third of deaths were from a condition that our current medical knowledge and technology could have prevented, highlighting the importance of causes such as coronary heart disease, respiratory disease and cancer. There is a clear and urgent need to identify individuals at risk earlier, and develop models of care that enable them to engage with interventions proven to either prevent or improve outcomes for early onset chronic disease. The Bureau of Investigative Journalism has published a relevant story.
Maxine and Spiros attended the Alan Turing Health programme conference in Manchester last week. The objective was to promote collaboration across university partners and highlight research strengths in the diversity of applications and analytic fields that Turing researchers are engaged in, as well as highlight new opportunities for impact. The Alan Turing Institute health programme is co-lead by Chris Holmes, Sebastian Vollmer and Catherine Lawrence.
Spiros will take up the role of Deputy Director of Research as of the 1st of February at the Institute of Health Informatics. Spiros will work alongside with Paul Taylor and take on new challenges in 2019 and the years ahead as the Institute seeks to grow.
Maria, Maxine and Nonie attended a methodology workshop on using clustering approaches and unsupervised learning for identifying subtypes of disease at the Usher Institute in Edinburgh. The workshop was funded by Asthma UK Centre for Applied Research and organized by Elsie Horne (@ElsieHorne) which has written a great summary of the event.
We are looking for a data manager to join the CALIBER team. The postholder will contribute significantly to the management and data processing requirements of a large linked electronic health record data platform - SQL/Python and stats required. Help us tackle over nine billion rows of data and improve human health and healthcare.
Our paper on predicting heart failure admissions from large-scale electronic health records by learning clinical concept embeddings (with GloVE) was accepted for the NIPS machine learning for health (ML4H) workshop!
Spiros was invited to speak in the UK Parliament as part of the All-Party Parliamentary Group on Artificial Intelligence , chaired by Stephen Metcalfe MP. The Group brought together policymakers, industry representatives, academics, philanthropists, and members of the public to discuss next steps in our journey towards an AI and data-driven world. Spiros provided evidence on how a data-driven, AI-powered healthcare system could look like in the future.
Arturo, Vaclav, Maria and Spiros have been invited to give several keynote speeches, presentations and tutorials at the Big Data Science conference at the Federal University of Bahia (UFBA) in Brazil! Members of the team will be presenting on data harmonization approaches using OMOP, machine learning for discovering obstructuve disease subtypes from electronic health records and building risk prediction tools to predict cardiovascular disease.
Spiros attended the the sixth Royal Academy of Engineering Frontiers of Engineering for Development symposium in Ho Chi Minh City, Vietnam. The event was made up of three multidisciplinary and interactive workshop-style sessions that explore the overall theme of Engineers as Healtchare Practitioners.
Congratulations to Maria Pikoula who was awarded a Dr Joseph Footitt research grant in chronic obstructive pulmonary disease research by the British Lung Foundation. Maria will apply and evaluate novel unsupervised machine learning methods in order to create of robust classifications which accurately reflect the complex and longitudinal nature of COPD and provide insight into the underlying respiratory disease prognostic factors.
Arturo and Collin ran a successful session on the uses of big health data for drug discovery and high resolution epidemiology in epilepsy research at the 13th European Congress on Epileptology in Vienna. The session was identified as a highlight of the conference in the closing keynote!
Congratulations to Kenan for successfully obtaining a UCL Arena Associate Fellowship. UCL Arena is UCL's professional development pathway for teaching: a scheme of awards accredited by the Higher Education Academy giving teaching and support staff nationally recognised fellowships. Kenan was made an Associate Fellowship in recognition of his high-quality contributions to teaching as part of our MSc in Health Data Science and short courses.
Maxine Mackintosh has been offered a place on the Alan Turing Institute PhD Enrichment Scheme. The Enrichment Scheme offers students, currently enrolled on a doctoral programme, the opportunity to join the ATI for up to 12 months to boost their skills, grow their network and work alongside Turing researchers. Maxine will be starting the scheme in October 2018 for nine months but will still remain an active member of the lab!
Congratulations to Marina Daskalopoulou who had two research abstracts accepted to the British HIV Association 2018 conference: a) Disclosure of HIV-serostatus to new sex partners and sexual behaviours among HIV-diagnosed MSM in the UK: results from the ASTRA study and b) HIV as a risk factor in the initial presentation of a range of cardiovascular, coronary, cerebrovascular, and peripheral arterial diseases: a linked electronic health records study of 8 million adults in the UK.
More good news from Maxine Mackintosh who was successful in securitng two research grants: a) from the UCL Grand Challenges Society’s Response to Ageing initiative and b) seed funding from the Doctoral Skills Development Programme Researcher-Led Initiative Award for establishing a Student HealthTech Hub. Well done!
Congratulations to Maxine Mackintosh who kicked off the new year with a successful PhD upgrade seminar titled How can Primary Care Electronic Health Records facilitate earlier diagnosis of dementia in the prodromal period?. Onwards and upwards!
Congratulations to Michalis Katsoulis who was awarded a four-year BHF Immediate PostDoctoral Basic Science Research Fellowship on “Weight change and the onset and progression of cardiovascular diseases in large scale electronic health records”. A great way to finish the year!
Our paper Comparing and Contrasting A Priori and A Posteriori Generalizability Assessment of Clinical Trials on Type 2 Diabetes Mellitus won one of the Distinguished Paper Awards of the American Medical Informatics Association 2017 Annual Symposium. Congratulations to Arturo for his great work!.
Many congratulations to Vaclav Papez who successfully defended his thesis and was awarded his PhD on: Archetype-based approach for modelling of electroencephalographic/event-related potentials data and metadata.
Many congratulations to Marina Daskalopoulou who successfully defended her thesis and was awarded her PhD on: Sexual behaviour of HIV-diagnosed men who have sex with men in the current era of effective antiretroviral therapy in the UK.
The ability of external investigators to reproduce published scientific findings is critical for the evaluation and validation of biomedical research by the wider community. However, a substantial proportion of health research using electronic health records (EHR), data collected and generated during clinical care, is potentially not reproducible mainly due to the fact that the implementation details of most data preprocessing, cleaning, phenotyping and analysis approaches are not systematically made available or shared. Our work outlines a collection of methods and tools from adjunct scientific disciplines that can potentially enable the reproducibility of EHR research findings.
Congratulations to Lab members Ghazaleh Fatemifar and Tom Lumbers for securing a prestigious AHA Data Fellowship to investigate heart failure subtypes using machine learning in heterogenous EHR data.
Many congratulations to Christiana McMahon who successfully defended her thesis and was awarded her PhD on: The evaluation and harmonisation of disparate information metamodels in support of epidemiological and public health research.
Our research using linked primary care and hospital electronic health records from 1.9m patients to study the association of clinically recorded alcohol consumption and the initial presentation of 12 CVDs has just been published in the BMJ and featured in The Guardian and New Scientist. While moderate drinking was associated with a protective effect in terms of some CVD presentations (such as heart attack), the associations observed were overall heterogeneous and varying. This has implications for counselling patients, public health communication, and clinical research, suggesting a more nuanced approach to the role of alcohol in prevention of cardiovascular disease is necessary.
Spiros represented the lab at this year's IEEE BHI conference, hosted in sunny Orlando, Florida. Abstracts from the tea, were presented on a) utilizing phenotyping algorithms derived from electronic health records in order to perform genome wide association studies and b) developing and evaluating a novel probabilistic methods for linking large administrative data from the Brazilian 100 million cohort. Hearty congratulations are due to Clicia and Marcos for winning the best Rapid Fire Presentation / Poster award!
Researchers from the UCL Institute of Health Informatics and Cardiovascular Science in collaboration with other UK and European institutes, were recently awarded €20 million research funding from the European Innovative Medicines Initiative. The research programme, will integrate healthcare data, activity monitors (wearables), state-of-the-art ‘-omics’ profiles, information about patients' lifestyles and health and their own reporting of symptoms, to better understand the causes of heart attack, atrial fibrillation and heart failure. Our lab is co-leading the health informatics work-stream. More details on UCL Press Release.
Congratulations to Maria, Clinical Data Scientist, who has been awarded her PhD from the University of Oxford entitled, "Multiscale modeling of the endothelial glycocalyx in physiological flow: A molecular dynamics approach". (abstract).
The report on the five-year NIHR project that founded CALIBER has been published in the NIHR Programme Grants for Applied Research (PGfAR) journal. The report describes the research undertaken in 33 studies using nationwide 'big health data' from linked electronic health records to help improve outcomes in cardiovascular diseases using methods from clinical epidemiology, biomedical informatics, health economics and social science as part of the CALIBER programme. Learn more by visiting the NIHR Journals Library website.