Enhancing MRI Imaging Efficiency: A Hybrid Under-Sampling Strategy for k-Space Data Acquisition

Compressed Sensing (CS) offers a promising solution to reduce MRI acquisition times, addressing challenges of prolonged scans and patient discomfort.This paper presents a new method for compressing and reconstructing MRI images using k-space gradients.A hybrid under-sampling approach allocates 80% of measurements to random sampling and 20% to deterministic sampling near the k-space center.Additionally, it explores the impact of reducing kx samples by 15%, 25%, and 50% on image quality.Reconstruction uses a nonlinear conjugate gradient method, with image quality assessed via a similarity index Q.Results show the proposed CS approach effectively compresses MRI data while preserving essential image quality, optimizing protocols and reducing scan times.