UNIGE document Scientific Article
previous document  unige:31709  next document
add to browser collection
Title

Multiscale lung texture signature learning using the Riesz transform

Authors
Foncubierta-Rodriguez, Antonio
Published in Lecture Notes in Computer Science. 2012, vol. 7512, no. Pt 3, p. 517-24
Abstract Texture-based computerized analysis of high-resolution computed tomography images from patients with interstitial lung diseases is introduced to assist radiologists in image interpretation. The cornerstone of our approach is to learn lung texture signatures using a linear combination of N-th order Riesz templates at multiple scales. The weights of the linear combination are derived from one-versus-all support vector machines. Steerability and multiscale properties of Riesz wavelets allow for scale and rotation covariance of the texture descriptors with infinitesimal precision. Orientations are normalized among texture instances by locally aligning the Riesz templates, which is carried out analytically. The proposed approach is compared with state-of-the-art texture attributes and shows significant improvement in classification performance with an average area under receiver operating characteristic curves of 0.94 for five lung tissue classes. The derived lung texture signatures illustrate optimal class wise discriminative properties.
Keywords AlgorithmsArtificial IntelligenceHumansLung/radiographyLung Diseases/radiographyPattern Recognition, Automated/methodsRadiographic Image Enhancement/methodsRadiographic Image Interpretation, Computer-Assisted/methodsReproducibility of ResultsSensitivity and SpecificityTomography, X-Ray Computed/methods
Identifiers
PMID: 23286170
Full text
Article (Published version) (1.3 MB) - document accessible for UNIGE members only Limited access to UNIGE
Other version: http://link.springer.com/bookseries/558
Structures
Research groups Traitement d'images médicales (893)
Groupe Geissbuhler Antoine (informatique médicale) (222)
Citation
(ISO format)
DEPEURSINGE, Adrien et al. Multiscale lung texture signature learning using the Riesz transform. In: Lecture Notes in Computer Science, 2012, vol. 7512, n° Pt 3, p. 517-24. https://archive-ouverte.unige.ch/unige:31709

185 hits

0 download

Update

Deposited on : 2013-12-04

Export document
Format :
Citation style :