Deep architectures for high-resolution multi-organ chest X-ray image segmentation - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Neural Computing and Applications Année : 2019

Deep architectures for high-resolution multi-organ chest X-ray image segmentation

Oscar Gómez
  • Fonction : Auteur
  • PersonId : 1062005
Pablo Mesejo
Oscar Ibáñez
  • Fonction : Auteur
Andrea Valsecchi
  • Fonction : Auteur
  • PersonId : 1062006
Oscar Cordón
  • Fonction : Auteur

Résumé

Chest X-ray images (CXRs) are the most common radiological examination tool for screening and diagnosis of cardiac and pulmonary diseases. The automatic segmentation of anatomical structures in CXRs is critical for many clinical applications. However, existing deep models work on severely down-sampled images (commonly 256 × 256 pixels), reducing the quality of the contours of the resulting segmentation and negatively affecting the possibilities of such methods to be effectively used in a real environment. In this paper, we study multi-organ (clavicles, lungs, and hearts) segmentation, one of the most important problems in semantic understanding of CXRs. We completely avoid down-sampling in images up to 1024 × 1024 (as in the JSRT dataset) and we diminish its impact in higher resolutions via network architecture simplification without a significant loss in the accuracy. To do so, we propose four different convolutional models by introducing structural changes to the baselines employed (U-Net and InvertedNet) as well as by integrating several techniques barely used by CXRs segmentation algorithms, such as instance normalization and atrous convolution. We also compare single-class and multi-class strategies to elucidate which approach is the most convenient for this problem. Our best proposal, X-Net+, outperforms 9 state-of-the-art methods on clavicles and lungs obtaining a Dice similarity coefficient of 0.938 and 0.978, respectively, employing a 10-fold cross validation protocol. The same architecture yields comparable results to the state-of-the-art in heart segmentation with a Dice value of 0.938. Finally, its reduced version, RX-Net+, obtains similar results but with a significant reduction in memory usage and training time.
Fichier principal
Vignette du fichier
NCAA.pdf (3.15 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02424739 , version 1 (28-12-2019)

Identifiants

  • HAL Id : hal-02424739 , version 1

Citer

Oscar Gómez, Pablo Mesejo, Oscar Ibáñez, Andrea Valsecchi, Oscar Cordón. Deep architectures for high-resolution multi-organ chest X-ray image segmentation. Neural Computing and Applications, 2019. ⟨hal-02424739⟩
83 Consultations
272 Téléchargements

Partager

Gmail Facebook X LinkedIn More