Ultrasound medical image deconvolution using CLEAN algorithm

Liviu-Teodor Chira , Jean-Marc Girault (1) , Tony Mateo , Sébastien Ménigot , Corneliu Rusu

1 iBraiN - Imaging, Brain & Neuropsychiatry

The problem of reconstruction of ultrasound medical images using blind deconvolution algorithm has been recognized as one of the most important aspect in ultrasound images. The image resolution is deteriorated by many parameters such as the diffusive effect in tissues, which produce the speckle noise. We intend to implement a nonlinear algorithm based on joint use of the well known CLEAN method and the Hybrid Parametric Inverse Filtering method. This method suppose an iterative process for extracting the “brightness” small objects presents in the image using a “dirty beam” Point Spread Function. The PSF is obtained with HYPIF algorithm, a blind technique for ultrasound medical images. The technique is applied for the 1D signals extracted from RF ultrasound images (simulated and experimental). The results are compared with Wiener filter.

Mots clés: Clean; blind deconvolution; ultrasound; point spread function

Domaines: Acoustique [physics.class-ph]

Liviu-Teodor Chira, Jean-Marc Girault, Tony Mateo, Sébastien Ménigot, Corneliu Rusu. Ultrasound medical image deconvolution using CLEAN algorithm. Acoustics 2012, Apr 2012, Nantes, France. ⟨hal-00810817⟩