A Deep Neural Network Application for Improved Prediction of HbA1c in Type 1 Diabetes
- Submitting institution
-
The University of Sheffield
- Unit of assessment
- 12 - Engineering
- Output identifier
- 5885
- Type
- D - Journal article
- DOI
-
10.1109/jbhi.2020.2967546
- Title of journal
- IEEE Journal of Biomedical and Health Informatics
- Article number
- -
- First page
- 2932
- Volume
- 24
- Issue
- 10
- ISSN
- 2168-2194
- Open access status
- Compliant
- Month of publication
- January
- Year of publication
- 2020
- URL
-
-
- Supplementary information
-
-
- 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
-
5
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- As a primary marker of long-term average blood glucose, the Haemoglobin A1C (HbA1c) test is essential for controlling type-1 diabetes. The paper presents a novel data-driven HbA1c prediction model based on deep learning and convolutional neural networks. Benchmark comparisons on a large number of patients demonstrated that the model was more practical and robust that existing approaches in the presence of imperfect data and outliers. Since 2018, the model has been used by NHS clinicians and patients in the NIHR funded DAFNE Plus trial (https://www.isrctn.com/ISRCTN42908016, Programme Grant RP-PG-0514-20013) and helped secure further funding via the UKRI grant (ES/V009796/1).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -