Restoration of ultrasonic images using non-linear system identification and deconvolution - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

Restoration of ultrasonic images using non-linear system identification and deconvolution

Résumé

This paper studies a new ultrasound image restoration method based on a non-linear forward model. A Hammerstein polynomial-based non-linear image formation model is considered to identify the system impulse response in baseband and around the second harmonic. The identification process is followed by a joint deconvolution technique minimizing an appropriate cost function. This cost function is constructed from two data fidelity terms exploiting the linear and non-linear model components, penalized by an additive-norm regularization enforcing sparsity of the solution. An alternating optimization approach is considered to minimize the penalized cost function, allowing the tissue reflectivity function to be estimated. Results on synthetic ultrasound images are finally presented to evaluate the algorithm performance.
Fichier principal
Vignette du fichier
hourani_24698.pdf (932.21 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03635405 , version 1 (08-04-2022)

Identifiants

Citer

Mohamad Hourani, Adrian Basarab, Denis Kouamé, Jean-Marc Girault, Jean-Yves Tourneret. Restoration of ultrasonic images using non-linear system identification and deconvolution. 15th IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI 2018), Apr 2018, Washington DC, United States. pp.1166-1169, ⟨10.1109/ISBI.2018.8363778⟩. ⟨hal-03635405⟩
26 Consultations
10 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More