Auto -Organiser Neural Network Application for Ultrasound Computed Tomographic Image Classification - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Auto -Organiser Neural Network Application for Ultrasound Computed Tomographic Image Classification

Marwa Fradi
  • Fonction : Auteur
  • PersonId : 1054282
Philippe Lasaygues

Résumé

Image classification still be a crucial topic of interest in the field of medical imaging diagnostic. In this paper, we have used a new Ultrasonic Computed Tomography (USCT) prototype giving us USCT bone images. However, given images are difficult to interpret and to classify. As a solution, an Auto Organiser Neural Network has been proposed to classify them automatically into pathologic bone images and healthy images. As a result, we have achieved an accuracy of 94.2% and an excellent value of gradient descent. Obtained results outperforms related works.
Fichier non déposé

Dates et versions

hal-02290156 , version 1 (17-09-2019)

Identifiants

  • HAL Id : hal-02290156 , version 1

Citer

Marwa Fradi, Philippe Lasaygues, Mohsen Machhout. Auto -Organiser Neural Network Application for Ultrasound Computed Tomographic Image Classification. 2019 19th International Conference on Sciences and Techniques of Automatic Control and Computer Engineering (STA), Mar 2019, Sousse, Tunisia. pp.19-23. ⟨hal-02290156⟩
34 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More