Imperial College London
Portrait Picture

Dr Ben Glocker

Faculty of Engineering, Department of Computing

Professor in Machine Learning for Imaging
 
 
 
//

Contact

 

+44 (0)20 7594 8334b.glocker Website CV

 
 
//

Location

 

377Huxley BuildingSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@article{Ellis:2022,
author = {Ellis, S and Manzanera, OEM and Baltatzis, V and Nawaz, I and Nair, A and Folgoc, LL and Desai, S and Glocker, B and Schnabel, JA},
journal = {The Journal of Machine Learning for Biomedical Imaging},
pages = {1--36},
title = {Evaluation of 3D GANs for lung tissue modelling in pulmonary CT},
url = {http://arxiv.org/abs/2208.08184v1},
volume = {1},
year = {2022}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - GANs are able to model accurately the distribution of complex,high-dimensional datasets, e.g. images. This makes high-quality GANs useful forunsupervised anomaly detection in medical imaging. However, differences intraining datasets such as output image dimensionality and appearance ofsemantically meaningful features mean that GAN models from the natural imagedomain may not work `out-of-the-box' for medical imaging, necessitatingre-implementation and re-evaluation. In this work we adapt and evaluate threeGAN models to the task of modelling 3D healthy image patches for pulmonary CT.To the best of our knowledge, this is the first time that such an evaluationhas been performed. The DCGAN, styleGAN and the bigGAN architectures wereinvestigated due to their ubiquity and high performance in natural imageprocessing. We train different variants of these methods and assess theirperformance using the FID score. In addition, the quality of the generatedimages was evaluated by a human observer study, the ability of the networks tomodel 3D domain-specific features was investigated, and the structure of theGAN latent spaces was analysed. Results show that the 3D styleGAN producesrealistic-looking images with meaningful 3D structure, but suffer from modecollapse which must be addressed during training to obtain samples diversity.Conversely, the 3D DCGAN models show a greater capacity for image variability,but at the cost of poor-quality images. The 3D bigGAN models provide anintermediate level of image quality, but most accurately model the distributionof selected semantically meaningful features. The results suggest that futuredevelopment is required to realise a 3D GAN with sufficient capacity forpatch-based lung CT anomaly detection and we offer recommendations for futureareas of research, such as experimenting with other architectures andincorporation of position-encoding.
AU  - Ellis,S
AU  - Manzanera,OEM
AU  - Baltatzis,V
AU  - Nawaz,I
AU  - Nair,A
AU  - Folgoc,LL
AU  - Desai,S
AU  - Glocker,B
AU  - Schnabel,JA
EP  - 36
PY  - 2022///
SP  - 1
TI  - Evaluation of 3D GANs for lung tissue modelling in pulmonary CT
T2  - The Journal of Machine Learning for Biomedical Imaging
UR  - http://arxiv.org/abs/2208.08184v1
UR  - https://www.melba-journal.org/papers/2022:024.html
UR  - http://hdl.handle.net/10044/1/102055
VL  - 1
ER  -
Download