Suggestions
Spiros Denaxas
Professor of Biomedical Informatics
Spiros Denaxas is a Professor of Biomedical Informatics at University College London (UCL) and the Institute of Health Informatics.14 He is also a Visiting Professor at the National & Kapodistrian University of Athens, a position he has held since January 2024.5
Denaxas is a prominent health data scientist whose research focuses on using electronic health records, clinical data, genetic information, and artificial intelligence to improve human health and healthcare.5 He leads the design and implementation of scalable machine learning approaches using large datasets from electronic health records to generate clinically-actionable insights for disease prevention, diagnosis, and treatment.5
Some of his key roles and achievements include:
- Associate Director at the British Heart Foundation Data Science Centre, leading the Computational Phenotyping theme.2
- Leader of Health Data UK's (HDR UK) National Phenotype Library project, which provides an open access library of electronic health record phenotyping algorithms.1
- Deputy Research Director of the Institute of Health Informatics, University College London.3
- Author of over 290 research papers in the field of health informatics.3
Denaxas's work spans various areas of biomedical informatics, including creating and evaluating algorithms for phenotyping, risk prediction, data modeling, and disease subtype discovery in structured electronic health records, clinical and omics data.12
Highlights
Very excited to share the culmination of 2+ years of research with fantastic colleagues across England, Scotland and Wales as part of the BHF Data Science Centre (British Heart Foundation) and led by Chris Tomlinson and Johan Thygesen
Our aim: create a reproducible framework to identify how and when people with COVID-19 are diagnoses, how they move through the healthcare system, the types of treatment they receive, and their outcomes.
We combined eight national data sources and analyzed electronic health records from 57 million people and billions of data points. We defined 10 COVID-19 phenotypes combining evidence from PCR testing, hospitalization/ICU admissions and treatment, primary care diagnoses and specialized COVID registries.
Paper in Lancet Digital Health The Lancet: https://lnkd.in/gm4g4khB
Code and algorithms under open source license: https://lnkd.in/gC7JrzQJ
Grateful for the amazing data made available through NHS and it's patients and the help from many great colleagues: Sam Hollings Mehrdad A. Mizani Jennifer Cooper Ashley Akbari Amitava Banerjee Alvina L. Kezhi (Ken) Li Bilal Mateen Navid Sataar reecha sofat Ana Torralbo Honghan Wu Angela Wood Jonathan Sterne Christina Pagel Will Whiteley Cathie Sudlow Harry Hemingway FFPH FRCP FMedSci
Great news - our Institute is leading the clinical data science theme.
