Automatic initialization and quality control of large-scale cardiac MRI segmentations
- Submitting institution
-
The University of Leeds
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- UOA11-3871
- Type
- D - Journal article
- DOI
-
10.1016/j.media.2017.10.001
- Title of journal
- Medical Image Analysis
- Article number
- -
- First page
- 129
- Volume
- 43
- Issue
- -
- ISSN
- 1361-8415
- Open access status
- Technical exception
- Month of publication
- October
- Year of publication
- 2017
- 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)
-
C - BMH (Applied Computing in Biology, Medicine and Health)
- Citation count
- 15
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The first pipeline quantifying biventricular CMR fully automatically on over 2,500 cardiac MRI studies from the MESA and DETERMINE clinical trials. Based on a combination of statistical shape models and machine learning, it adds two novel and significant components: (1) fully automatic initialisation; (2) automatic segmentation quality control. Quantitative cardiac function parameters are indistinguishable from expert manual quantification. Pipeline subsequently tested on 20,000 CMR datasets (doi.org/10.1016/j.media.2019.05.006), demonstrating scalability. Currently deployed on our MULTI-X (www.multi-x.org) framework and evaluated at the UK Biobank (UKB Deputy CEO, mark.effingham@ukbiobank.ac.uk). This approach has led to longstanding collaborations with QMUL (s.e.petersen@qmul.ac.uk) and Oxford (stefan.neubauer@cardiov.ox.ac.uk).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
