Hyperbolic Wavelet-Fisz denoising for a model arising in Ultrasound Imaging

Younes Farouj (1) , Jean-Marc Freyermuth (2, 3) , Laurent Navarro (4) , Marianne Clausel (5) , Philippe Delachartre (1)

1 Imagerie Ultrasonore
2 CAM - University of Cambridge [UK]
3 ORSTAT - Operations Research and Business Statistics
4 LINA-ENSMSE - Laboratoire Interdisciplinaire d'Etude des Nanoparticules Aérosolisées
5 DAO - Données, Apprentissage et Optimisation

We present an algorithm and its fully data-driven extension for noise reduction in ultrasound imaging. Our proposed method computes the hyperbolic wavelet transform of the image, before applying a multiscale variance stabilization technique, via a Fisz transformation. This adapts the wavelet coefficients statistics to the wavelet thresholding paradigm. The aim of the hyperbolic setting is to recover the image while respecting the anisotropic nature of structural details. The data-driven extension removes the need for any prior knowledge of the noise model parameters by estimating the noise variance using an isotonic Nadaraya-Watson estimator. Experiments on synthetic and real data, and comparisons with other noise reduction methods demonstrate the potential of our method at recovering ultrasound images while preserving tissue details. Finally, we emphasize the noise model we consider by applying our variance estimation procedure on real images.

Mots clés: Fisz transformation; Hyperbolic wavelets; Variance stabilization; Ultrasound imaging; Gaussianization; Data-driven denoising

Domaines: Traitement du signal et de l'image [eess.SP]

Younes Farouj, Jean-Marc Freyermuth, Laurent Navarro, Marianne Clausel, Philippe Delachartre. Hyperbolic Wavelet-Fisz denoising for a model arising in Ultrasound Imaging. IEEE Transactions on Computational Imaging, 2017, 3 (1), pp.1-10. ⟨10.1109/TCI.2016.2625740⟩. ⟨hal-01322246⟩