Automated Triaging of Adult Chest Radiographs with Deep Artificial Neural Networks
- Submitting institution
-
King's College London
- Unit of assessment
- 12 - Engineering
- Output identifier
- 111468898
- Type
- D - Journal article
- DOI
-
10.1148/radiol.2018180921
- Title of journal
- Radiology
- Article number
- -
- First page
- 196
- Volume
- 291
- Issue
- 1
- ISSN
- 1527-1315
- Open access status
- Not compliant
- Month of publication
- January
- Year of publication
- 2019
- 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
- Yes
- Number of additional authors
-
5
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Prompt reporting of radiographs is challenging for the NHS with 20% unfilled posts and rising demand, leading to substantial diagnostic delays (https://www.rcr.ac.uk/clinical-radiology/service-delivery/rcr-radiology-workforce-census). Using our underpinning natural language processing technology, we developed an AI system incorporating 2 deep neural networks, on 470,388 chest radiographs, that identifies key findings. By reducing NHS reporting delays for critical/urgent radiographs, this is potentially transformative. This generated international news interest (32 items e.g. https://www.thetimes.co.uk/article/hospitals-will-use-ai-to-check-x-rays-9g6nmxp0g; https://www.bbc.co.uk/programmes/p06yqysy); Wellcome funding (210839/Z/18/Z) for a prospective clinical study; £26M Medical Imaging&AI Centre funding (UKRI 10469) for pathway evaluation; and industry collaboration for international commercialisation (Soliton IT, Leigh Edwards ).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -