The Effective 3D MRI Reconstruction Method Driven by the Fusion Strategy in NSST Domain

The 3D reconstruction of medical images plays an important role in modern clinical diagnosis. Although the analytic-based, iterative-based and deep learning-based methods have been popularly used, there are still many problems to deal with. analysis-based not accurate enough, iteration-based computationally intensive, learning based heavily dependent on training data. To solve default that only single scan sequence is included traditional methods, a method driven by non-subsampled shearlet transform (NSST) algebraic technique (ART) proposed. Firstly, multiple magnetic resonance imaging (MRI) sequences decomposed into high-frequency low-frequency components NSST. Secondly, parts fused with weighted average fusion scheme coefficient guided regional gradient energy. Finally, performed using ART algorithm. Compared proposed able capture more information from MRI sequences, which makes results much clearer accurate. By comparing single-sequence model without fusion, experiments fully prove accuracy effectiveness.