Biomedical Images Enhancement

Biomedical images enhancement using optimized Schrödinger operator-based method


                              ​​Click here to watch the Abderrazak's presentation on Youtube.

The main of challenge of Magnetic Resonance Imaging (MRI) is dealing with high levels of noise which may corrupt the image especially since the noise is almost correlated with the image details. In this regard, we propose a new MRI enhancement method to overcome this limitation. The proposed MRI enhancement method relies on square sub-images enhancement depending on the noise level in each position using spatial adaptation of the Semi-Classical Signal Analysis (SCSA) method.

 The new method is based on an adaptive selection of a filtering parameter by a grid segmentation of the noisy input image (Fig.1). The filtering parameter, called h, will follow an appropriate distribution along the different sub-images allowing the adaptation of its value to the spatial variation of noise and responded efficiently to the denoising objectives.
 The method has been applied to a synthetic dataset from BrainWeb (Fig.2) and real MRI images (Fig.3) show the effectiveness of the proposed approach compared to the standard case with one fixed parameter.
These preliminary results provide very encouraging insight and show good potential in MRI noise removal while preserving the small details.


​​Hacene Serrai, Eric Achten​ from Gent University, Belguim.


​​Abderrazak Chahid, Hacene Serrai, Eric Achten, Taous-Meriem Laleg-Kirati, “Adaptive Method for MRI Enhancement Using Squared Eigenfunctions of the Schrödinger Operator," accepted in BIOCAS2017, Italy, 2017.​