This project aims to create the capability for a near real-time analytics of unstructured and structured records to support research, trial recruitment and value-based healthcare in radiology decision making.
Building on the expertise developed in SAIL, UKSeRP, and deployment of MedGate NLP in NHS data appliances, the team will develop a system to liberate millions of NHS images from multiple modalities and annotate these with sufficient exposure and outcome variables to support efficient development and evaluation of scalable AI/ML algorithms (by the global community) that can be adopted and implemented by the NHS.
Risk of blood clots remains for almost a year after COVID-19 infection, study suggests
20 September 2022
COVID-19 infection increases the risk of potentially life-threatening blood clots for at least 49 weeks, according to a new study of health records of 48 million unvaccinated adults from the first...
The HDR UK Phenotype Library celebrates reaching 1000 phenomes
30 August 2022
The milestone marks the Library as the largest national resource for information, tools and phenotyping algorithms to allow researchers to harness data held in Electronic Health Records.
Global digital health leaders propose framework to enhance research using real-world data
29 August 2022
Researchers from HDR UK Midlands and BHF Data Science Centre part of international group proposing new framework to improve the integrity and quality of studies using healthcare data, and boost...