Ultrasound image reconstruction from compressed measurements using approximate message passing

Jong-Hoon Kim (1) , Adrian Basarab (2, 3) , Paul Hill (1) , David Bull (1) , Denis Kouamé (2) , Alin Marian Achim (1)

1 University of Bristol [Bristol]
2 IRIT-TCI - Traitement et Compréhension d’Images
3 UT3 - Université Toulouse III - Paul Sabatier

In this paper we propose a novel framework for compressive sampling reconstruction of biomedical ultrasonic images based on the Approximate Message Passing (AMP) algorithm. AMP is an iterative algorithm that performs image reconstruction through image denoising within a compressive sampling framework. In this work, our aim is to evaluate the merits of several combinations of a denoiser and a transform domain, which are the two main factors that determine the recovery performance. In particular, we investigate reconstruction performance in the spatial, DCT, and wavelet domains. We compare the results with existing reconstruction algorithms already used in ultrasound imaging and quantify the performance improvement.

Mots clés: Nonconvex optimization; Compressive sampling; Ultrasonic images; IRLS; AMP; Image denoising

Domaines: Synthèse d'image et réalité virtuelle [cs.GR]; Traitement du signal et de l'image [eess.SP]; Traitement des images [eess.IV]; Vision par ordinateur et reconnaissance de formes [cs.CV]; Intelligence artificielle [cs.AI]

Jong-Hoon Kim, Adrian Basarab, Paul Hill, David Bull, Denis Kouamé, et al.. Ultrasound image reconstruction from compressed measurements using approximate message passing. 24th European Signal and Image Processing Conference (EUSIPCO 2016), Aug 2016, Budapest, Hungary. pp. 557-561. ⟨hal-01517386⟩