Due to the lack of available annotated medical images, accurate computer-assisted diagnosis requires intensive data augmentation (DA) techniques, such as geometric/intensity transformations of original images; however, those transformed images intrinsically have a similar distribution to the original ones, leading to limited performance improvement. To fill the data lack in the real image distribution, we synthesize brain contrast-enhanced magnetic resonance (MR) images—realistic but completely different from the original ones—using generative adversarial networks (GANs). This study exploits progressive growing of GANs (PGGANs), a multistage generative training method, to generate original-sized 256 × 256 MR images for convolutional neural network-based brain tumor detection, which is challenging via conventional GANs; difficulties arise due to unstable GAN training with high resolution and a variety of tumors in size, location, shape, and contrast. Our preliminary results show that this novel PGGAN-based DA method can achieve a promising performance improvement, when combined with classical DA, in tumor detection and also in other medical imaging tasks.

Han, C., Rundo, L., Araki, R., Furukawa, Y., Mauri, G., Nakayama, H., et al. (2020). Infinite Brain MR Images: PGGAN-Based Data Augmentation for Tumor Detection. In A. Esposito, M. Faundez-Zanuy, F. Morabito, E. Pasero (a cura di), Neural Approaches to Dynamics of Signal Exchanges (pp. 291-303). Springer Science and Business Media Deutschland GmbH [10.1007/978-981-13-8950-4_27].

Infinite Brain MR Images: PGGAN-Based Data Augmentation for Tumor Detection

Rundo L.
Co-primo
;
Mauri G.;
2020

Abstract

Due to the lack of available annotated medical images, accurate computer-assisted diagnosis requires intensive data augmentation (DA) techniques, such as geometric/intensity transformations of original images; however, those transformed images intrinsically have a similar distribution to the original ones, leading to limited performance improvement. To fill the data lack in the real image distribution, we synthesize brain contrast-enhanced magnetic resonance (MR) images—realistic but completely different from the original ones—using generative adversarial networks (GANs). This study exploits progressive growing of GANs (PGGANs), a multistage generative training method, to generate original-sized 256 × 256 MR images for convolutional neural network-based brain tumor detection, which is challenging via conventional GANs; difficulties arise due to unstable GAN training with high resolution and a variety of tumors in size, location, shape, and contrast. Our preliminary results show that this novel PGGAN-based DA method can achieve a promising performance improvement, when combined with classical DA, in tumor detection and also in other medical imaging tasks.
Capitolo o saggio
Brain MRI; Data augmentation; Generative adversarial networks; Synthetic medical image generation; Tumor detection
English
Neural Approaches to Dynamics of Signal Exchanges
Esposito, A.; Faundez-Zanuy, M.; Morabito, F.C.; Pasero, E.
2020
978-981-13-8949-8
151
Springer Science and Business Media Deutschland GmbH
291
303
Han, C., Rundo, L., Araki, R., Furukawa, Y., Mauri, G., Nakayama, H., et al. (2020). Infinite Brain MR Images: PGGAN-Based Data Augmentation for Tumor Detection. In A. Esposito, M. Faundez-Zanuy, F. Morabito, E. Pasero (a cura di), Neural Approaches to Dynamics of Signal Exchanges (pp. 291-303). Springer Science and Business Media Deutschland GmbH [10.1007/978-981-13-8950-4_27].
open
File in questo prodotto:
File Dimensione Formato  
L48-WIRN 2018 PGGAN.pdf

accesso aperto

Tipologia di allegato: Publisher’s Version (Version of Record, VoR)
Dimensione 557.99 kB
Formato Adobe PDF
557.99 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/255875
Citazioni
  • Scopus 63
  • ???jsp.display-item.citation.isi??? 45
Social impact