||Magnetic resonance imaging (MRI) techniques are widely applied in various disease diagnoses and scientific research projects as noninvasive methods. However, lower signal-to-noise ratio (SNR), B1 inhomogeneity, motion-related artifact, susceptibility artifact, chemical shift artifact and Gibbs ring still play a negative role in image quality improvement. Various techniques and methods were developed to minimize and remove the degradation of image quality originating from artifacts. In the first part of this dissertation, a motion artifact reduction technique based on a novel real time self-gated pulse sequence is presented. Diffusion weighted and diffusion tensor magnetic resonance imaging techniques are generally performed with signal averaging of multiple measurements to improve the signal-to-noise ratio and the accuracy of diffusion measurement. Any discrepancy in images between different averages causes errors that reduce the accuracy of diffusion MRI measurements. The new scheme is capable of detecting a subject's motion and reacquiring motion-corrupted data in real time and helps to improve the accuracy of diffusion MRI measurements. In the second part of this dissertation, a rapid T1 mapping technique (two dimensional singleshot spin echo stimulated echo planar image--2D ss-SESTEPI), which is an EPI-based singleshot imaging technique that simultaneously acquires a spin-EPI (SEPI) and a stimulated-EPI (STEPI) after a single RF excitation, is discussed. The magnitudes of SEPI and STEPI differ by T1 decay for perfect 90o RF pulses and can be used to rapidly measure the T1 relaxation time. However, the spatial variation of B1 amplitude induces uneven splitting of the transverse magnetization for SEPI and STEPI within the imaging FOV. Therefore, correction for B1 inhomogeneity is critical for 2D ss-SESTEPI to be used for T1 measurement. In general, the EPI-based pulse sequence suffers from geometric distortion around the boundary of air-tissue or bone tissue. In the third part of this dissertation, a novel pulse sequence is discussed, which was developed based on three dimensional singleshot diffusion weighted stimulated echo planar imaging (3D ss-DWSTEPI). A parallel imaging technique was combined with 3D ss-DWSTEPI to reduce the image distortion, and the secondary spin echo formed by three RF pulses (900-1800-900) is used to improve the SNR. Image quality is improved.