De-smokeGCN: Generative Cooperative Networks for Joint Surgical Smoke Detection and Removal
- Submitting institution
-
The University of Bradford
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 57
- Type
- D - Journal article
- DOI
-
10.1109/TMI.2019.2953717
- Title of journal
- IEEE Transactions on Medical Imaging
- Article number
- -
- First page
- 1615
- Volume
- 39
- Issue
- 5
- ISSN
- 0278-0062
- Open access status
- Compliant
- Month of publication
- -
- Year of publication
- 2019
- URL
-
https://ieeexplore.ieee.org/document/8902171
- 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
-
4
- Research group(s)
-
-
- Citation count
- 2
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Surgical-smoke poses significant challenges in medical imaging tasks and presents hazards to surgeons. This paper, in the world-leading journal in medical image, presents a novel deep-learning training framework for surgical-smoke removal during minimally-invasive surgery. This is significant because previous work has limited success in improving intra-operative image quality. The framework doesn’t require large hand-labelled datasets, learns from computer-augmented images. This bridges a substantial gap between conventional and learning based methods, outperforms the state-of-the-art approaches. It won Braford’s best publication Jan/2020, led to funding proposal (GK6207071341), an invited talk (XIBEI University 2020), a PhD student gaining a prestigious job (Bluevision).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
