Generating high-fidelity synthetic patient data for assessing machine learning healthcare software
- Submitting institution
-
Brunel University London
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 095-227235-5379
- Type
- D - Journal article
- DOI
-
10.1038/s41746-020-00353-9
- Title of journal
- npj Digital Medicine
- Article number
- 147
- First page
- -
- Volume
- 3
- Issue
- 1
- ISSN
- 2398-6352
- Open access status
- Compliant
- Month of publication
- November
- Year of publication
- 2020
- URL
-
-
- Supplementary information
-
https://static-content.springer.com/esm/art%3A10.1038%2Fs41746-020-00353-9/MediaObjects/41746_2020_353_MOESM1_ESM.pdf
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
3
- Research group(s)
-
1 - Artificial Intelligence (AI)
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This work explores the use of probabilistic graphical models with latent variables to model real-world primary care data so that synthetic data can be generated and used in its place to remove issues of privacy. It has been tested extensively on UK data and has led to extended work funded by NHS X. As a result of this work, the Medicine Health Regulatory Authority has introduced a new synthetic data generation service. Two new synthetic datasets have been released in 2020 based on cardiovascular disease and Covid19, enabling the development of new diagnostic tools: https://www.gov.uk/government/news/new-synthetic-datasets-to-assist-covid-19-and-cardiovascular-research
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
