Automatic MRI Acquisition Parameters Optimization Using Perceptual Criteria

INTRODUCTION: The visualization of structures in MRI highly depends on many user defined scan parameters. The selection of them is always done heuristically and requires a vast experience from the operator. Furthermore, sometimes it is not simple to predict the effect on the visibility of the structures of interest when a parameter is modified. We propose a methodology based on an automatic optimization to find the MRI acquisition parameters that maximize the visibility of a desired structure. The objective function of our optimization is computed from Visibility Maps (VM) that are designed to measure the visibility of structures according to two perceptual criteria: sensitivity to contrasts and to spatial frequencies. METHODS: Since the Human Visual System (HVS) has been adapted to detect specific ranges of contrast and spatial frequencies, we developed Visibility Maps (VM) (Fig. 1d) that mimic these non–linear sensitivities. Our VM give as an output how visible is each voxel of an image, and they are constructed by calculating the pointwise product between two maps: a Contrast Map (Fig. 1b) and a Relevant Spatial Frequency (RSF) Map (Fig. 1c). To model the visibility of a structure according to its local contrast, we propose the creation of Contrast Maps (Fig. 1b), where the intensity of each pixel is the probability of having a given local intensity difference computed from the Contrast Sensitivity Function (CSF) [1]: