Medical image matching using variable randomized undersampling probability pattern in data acquisition

Abstract

This paper proposes a randomized variable probability pattern in under-sampling acquisition for medical image matching which is a method that can perform the quantitative analysis of tissue parameters. For high-speed estimation of tissue parameters, random under-sampling with less than the Nyquist rate in k-space is required. This study presents an accurate parameter mapping method for under-sampled data by using various randomized probability pattern. In comparison to the fixed probability pattern, the proposed method shows improved estimation results with reduced artifacts such as ghosting effects due to the undersampling scheme.